Designing_a_treatment_system_for_..._4237536.

Designing_a_treatment_system_for_..._4237536.
01 December 2014
Designing a treatment system for the underground effluent at Loulo
Gold mine (Mali, West Africa)
Msc Thesis Report
Author
Student nbr
:
:
Hilaire B Diarra
4237536
Supervisor (s)
:
Mike Buxton
Charles Wells
Graham Trusler
Jennifer Molwantwa
Associate Professor Resource
Engineering, TuDelft
Group Sustainability Manager, Randgold
Resources Ltd
Chief Executive Officer, Digby Wells
Environmental
Research Manager, Water Research
Commission (South Africa)
Address
:
Section for Resource Engineering
Department of Applied Earth Sciences
Delft University of Technology
P.O. Box 5028
The Netherlands
Telephone
:
Fax
:
(31) 15 2781328
(secretary)
(31) 15 2781189
Copyright ©2012 Section for Resource Engineering
All rights reserved. No parts of this publication may be reproduced, Stored in a retrieval system, or transmitted, In
any form or by any means, electronic, Mechanical, photocopying, recording, or otherwise, Without the prior
written permission of the Section for Resource Engineering
Abstract
Water monitoring and analysis results have shown that the current underground water
discharge into the adjacent Faleme river is substandard, indicating that the existing settling
ponds are not suitable to bring the effluent into compliance with the Malian discharge limits
1
prior to discharge. Since the identification of the problem, the mine has undertaken a number
of studies to find the best possible solution to the discharge quality issue.
This study puts all the different studies into perspectives and uses the available data to
assess the impact of the effluent on the receiving environment and to develop a constructed
wetland which is believed to be a more adequate treatment solution than the existing settling
ponds. The study is divided into three main parts.
The first part assesses the impact of the mine effluent on the Faleme river by characterizing
the discharge water quality since mining commenced (i.e. from 2010) up to date; then by
evaluating the current ecological status of the Faleme river and finally by attempting to draw
conclusions as to the potential sources of the issues identified in the Faleme with the mine
discharge quality and other activities in and along the river (e.g. artisanal mining activities,
agriculture, and others). Even though the quality of the mine discharge is substandard, the
current issues associated with the Faleme river ecosystem are caused by the artisanal
mining activities and aggregate mining by the clandestine sand miners.
The second part of the report deals with the designs of a thickening system to remove the
solids from the effluent prior to entering a constructed wetland which design also is proposed
to remove the nitrate from the effluent. A 10 m diameter thickener was designed assuming a
160m3/h maximum flow rate, a 5%m solid concentration in the feed and a 2m/h critical rise
to achieve a settling rate of 20 t0 30 m/h. the overflow of the thickener will enter a
subsurface horizontal flow type wetland with an inlet, macrophyte zone and an outlet. Due to
the complex nature of the processes in the wetland and the difficulty in modelling it, a
number of design options was proposed to be trial tested. Since the final wetland should be
at least 100 m wide and 300m long, the proposed pilot scale size is 10m x 30m. Cyperus
papyrus and Typha will be tested as the plants to be used.
If the uncertainties around the design of the wetland can be reduced and a model developed
through the pilot scale test, this can be used in other mines in Mali and even in other
countries with similar climatic conditions and similar problems.
Acknowledgement
Sincere thanks go to my hierarchy in Randgold Resources Ltd (Haladou M Manirou, Rodney
Quick and Charles Wells) to have allowed me to pursue a Master degree. Thanks to
Randgold Resources Ltd to have funded this study. I would like to thank also Dr Mike
Buxton, M. Graham Trusler, Dr. Jennifer Molwantwa & Charles Wells for reviewing the report
and providing valuable inputs.
Special thanks go to Mrs. Jikste van der Laan for her constant assistance all along my
studies at Delft (registration, visa and accommodation arrangement). Prof Hans de Ruiter
has always been available to help the EGEC student – thank you.
2
Contents
Abstract ..................................................................................................................... 1
Acknowledgement ..................................................................................................... 2
3
Introduction ......................................................................................................... 9
3.1.
Motivation ................................................................................................... 10
3.2.
Research Questions ................................................................................... 11
3.3.
Aim and Objectives .................................................................................... 12
3.4.
Methodology and the structure of the report ............................................... 13
3.5.
Research Scope ......................................................................................... 14
4
Background information..................................................................................... 15
4.1.
Regional Climate ........................................................................................ 15
4.2.
Geology and Mining ................................................................................... 19
5
Regulatory framework ....................................................................................... 26
5.1.
Mining code ................................................................................................ 26
5.2.
Convention ................................................................................................. 26
5.3.
Water code ................................................................................................. 27
5.4.
Pollution and Nuissance prevention Laws .................................................. 27
6
Characterization of the Underground Effluent .................................................... 29
6.1.
Flow volume ............................................................................................... 29
6.2.
Water quality .............................................................................................. 38
7
Faleme River Ecosystem Assessment .............................................................. 43
7.1.
Information on the Faleme .......................................................................... 43
7.2.
Methodologies ............................................................................................ 44
7.3.
Results ....................................................................................................... 48
8
Gara Underground Effluent treatment ................................................................ 67
8.1.
Proposed Solutions .................................................................................... 67
8.2.
Constructed Wetland – conceptual design ................................................. 69
9
Conclusions & Discussions ............................................................................. 105
9.1.
Impact of the Underground Effluent on the Faleme river .......................... 105
9.2.
Design of the Gara Effluent Treatment System ........................................ 113
10
Recommendations ....................................................................................... 122
3
11
References .................................................................................................. 139
Nomenclature
CW
Constructed Wetland
DnS
Downstream
DWA
Digby Wells and Associates
DWE
Digby Wells Environmental
EMS
Environmental Management System
EPT
Ephemeroptera (Mayflies), Plecoptera (Stoneflies) and Trichoptera (Caddisflies)
ET
Evapotranspiration
FAO
Food and Agriculture Organisation
FAST
Faculté des Sciences et Techniques
FL1
Faleme Downstream of Loulo Operation
FL2
Faleme Upstream of Loulo Operation
FWS
Free Water Surface
GK
Gounkoto
GK1
Faleme Downstream of Gounkoto Operation
GK2
Faleme Upstream of Gounkoto Operation
GR1
In the lower reaches of the Garra Dam adjacent to the Loulo camp
GR2
At the Garra River bridge downstream of the local orpailleurs (artisanal miners)
HECRAS Hydrologic Engineering Centers River Analysis System
HF
Horizontal subsurface Flow
HSSF
Horizontal Subsurface Flow
ICP-OES
Inductively Coupled Plasma Optical Emission Spectrophotometry
ICP-MS
Inductively Coupled Plasma Mass spectrophotometry
IFC
International Finance Corporation
IHAI
Intermediate Habitat Assessment Integrity
4
ISO
International Organization for Standardization
MEC
Midpoint Effect Concentration
OMVS
Organisation pour la Mise en Valeur du Fleuve Sénégal
PC-UP
Immediately Upstream of the discharge area into the Falémé River
PC-DWN
Immediately Downstream of the discharge area into the Falémé River
PEC
Probable Effect Concentration
PSD
Particle Size Distribution
P&C
Patterson &Cooke
SASS5
South African Scoring System (SASS), Version 5
SSF
Subsurface Flow
TEC
Threshold Effect Concentration
UpS
Upstream
USEPA
United States Environmental Protection Agency
VF
Vertical subsurface Flow
List of Tables
Table 1. Scope of the research ................................................................................ 15
Table 2. Water quality standards and guidelines ..................................................... 27
Table 3. Descriptive statistics of the data set (value in m3) ...................................... 29
Table 4. Steady state calibrated hydraulic conductivity (K) values ........................... 37
Table 5. Effluents Solids analyses results for heavy metals ..................................... 39
Table 6. Quality data of the discharge before settling (2010-2014) .......................... 42
Table 7. Quality data of the discharge after existing settlers .................................... 43
Table 8. Summary of water quantity of the two catchments (Source: DWA, 2009) .. 43
Table 9. Sampling points and description (Modified from DWA, 2012) .................... 44
Table 10. Description of IHAS scores with the respective percentage category ....... 46
Table 11. Defined thresholds for metal concentrations in sediments (DWA, 2012) .. 46
Table 12. IHIA for instream Habitat within the Faleme River (Source: Tate, 2014) .. 52
Table 13. IHIA for Riparian Habitat within the Faleme River (Source: Tate, 2014) ... 52
Table 14. IHIA for instream Habitat within the Gara River (Source: Tate, 2014) ...... 53
Table 15. IHIA for Riparian Habitat within the Gara River (Source: Tate, 2014) ....... 53
Table 16. Water quality results obtained during the 2014 July survey ...................... 53
Table 17. Chemical analysis of water from the sites considered in the 2014 survey 55
5
Table 18. Chemical analysis of sediment samples collected ................................... 56
Table 19. Integrated Habitat Assessment System (IHAS) results for the 2014 ........ 57
Table 20. Macroinvertebrates sampled during the July 2014 survey. ....................... 57
Table 21. Families recorded during the 2014 (Source: Tate, 2014). ........................ 61
Table 22. Effluent characteristics (SGS, 2011) ........................................................ 83
Table 23. Dynamic Thickening Test Settings and Data ............................................ 90
Table 24. Thickener Operating Parameters ............................................................. 92
Table 25. Thickener Sizing Parameters ................................................................... 92
Table 26. Minimum Thickening Area (P&C, 2011) ................................................... 93
Table 27. Recommended Thickener Diameter (P&C, 2011) .................................... 93
Table 28. Recommended thickener size (P&C, 2011) ........................................... 123
List of Figures
Figure 1. Location map (Source: RRL annual reports 2012/2013) ............................. 9
Figure 2. Loulo gold mine with the location of the underground discharge point ...... 11
Figure 3. Objectives (left) that answer the research questions (right) ..................... 13
Figure 4. Methodology ............................................................................................. 14
Figure 5. Average monthly temperatures (Source: Loulo Baseline report, 2007) ..... 16
Figure 6. Average monthly rainfalls (Source: Loulo Baseline report, 2007) .............. 17
Figure 7. Average monthly evaporation (Source: Loulo Baseline report, 2007) ........ 17
Figure 8. Average and maximum wind speed and direction ..................................... 18
Figure 9. Average monthly sunlight (Source: Loulo Baseline report, 2007) .............. 19
Figure 10. Geological Map of the Kenieba window (source: baseline report, 2007) . 20
Figure 11. Geology in the study area (Courtesy from Holliday, 2014) ...................... 21
Figure 12. Grab and drill core samples of the Gara vein stockwork.. ....................... 23
Figure 13. Back-scattered electron (BSE) images of the major and minor sulphide
phases present at Gara.) .................................................................................................... 24
Figure 14. Gara underground layout showing the portal from the gara pit into
undergrounds through spirals declines ................................................................................ 25
Figure 15. Underground water reticulation (Source: Randgold, 2011) ..................... 26
Figure 16. Monthly flow data from January 2011 to January 2014 ........................... 29
Figure 17. Boreholes locality map............................................................................ 31
Figure 18. Damage zone with fault intersection - north wall of Gara pit.................... 32
Figure 19. Model setup (not all layers are shown) ................................................... 34
Figure 20. Gara Underground mine plan (rotated 110̊ clockwise to show decline) ... 34
Figure 21. Calculated vs. observed groundwater levels ........................................... 37
Figure 22. Steady state water balance..................................................................... 38
Figure 23. Variations of total suspended solids (TSS) in the final discharge ............ 39
Figure 24. Variations of Aluminium (Al) in the final discharge .................................. 40
Figure 25. Variations of total Iron (Fe-Total) in the final discharge ........................... 41
6
Figure 26. Variations of Nitrate (NO3) in the final discharge ..................................... 41
Figure 27. Variations of Sulfate (SO4) in the final discharge .................................... 42
Figure 28. Arial map showing the sampling points (Source: DWA, 2012) ................ 45
Figure 29. Photography of the sampling points at Faleme and Garra ...................... 45
Figure 30. Aerial Image of the Falémé River, depicting meandering nature and
alluvial deposition (Google Earth)........................................................................................ 49
Figure 31. Cobble beds in the Falémé River. A: Natural cobble bed; B: Modified
cobble bed; C: Modified cobble bed. ................................................................................... 50
Figure 32. Image of intact riparian habitat (Farandie). ............................................. 50
Figure 33. Image of riparian habitat (Cobble Bed) ................................................... 51
Figure 34. Longitudinal profile of the Falémé River (Google Earth™) ...................... 52
Figure 35. Percentage contribution of EPT at the sites assessed in the 2014 survey.
........................................................................................................................................... 59
Figure 36. Margalef’s Diversity Index results for the July 2014 survey ..................... 60
Figure 37. Results of Shannon’s Diversity Index during the July 2014 surveys. ....... 60
Figure 38. Results of Pielou’s Evenness Index during the July 2014 surveys. ......... 61
Figure 39. Photographs of fishes from the Alestid family (Source: Tate, 2014) ........ 62
Figure 40. Photographs of fishes from the Cichlid family (Source: Tate, 2014) ........ 63
Figure 41. Bioaccumulation of metals in the Faleme River (Source: Tate, 2014) ..... 65
Figure 42. Bioaccumulation of metals in the Gara River (Source: Tate, 2014) ......... 65
Figure 43. Comparison of metal bioaccumulation between the Gara and Faleme
River (Source: Tate, 2014) .................................................................................................. 66
Figure 44. Surface settling dams showing large amount of suspended solids.......... 68
Figure 45. Gabion system conceptual design (DWA, 2013) ..................................... 69
Figure 46. Gabions (left) and gabions meshed cages (right) (DWA, 2013) .............. 69
Figure 47. Free Surface flow (FSW) constructed wetland – section view (Source:
Wetland International, 2003) ............................................................................................... 70
Figure 48. Sub-Surface flow (SSF) constructed wetland (Source: Wetland
International, 2003) ............................................................................................................. 70
Figure 49. Nitrogen cycle ........................................................................................ 73
Figure 50. Nitrogen transformations in a constructed wetland treatment system
(Source: Cooper et al., 1996) .............................................................................................. 76
Figure 51. Extensive root system of plants (source: modified from Cooper et al.,
1996) .................................................................................................................................. 78
Figure 52. Bench-top dynamic thickening rig (P&C, 2011) ....................................... 83
Figure 53. Particle size distribution of the solids in effluent (P&C, 2011).................. 84
Figure 54. Flocculant screening results (SGS, 2011) ............................................... 85
Figure 55 : Thickener Feed Effluent Solids Conc. (settling rate) .............................. 86
Figure 56 : Thickener Feed Effluent Solids Conc. (clarity) ....................................... 86
Figure 57 : Flocculant Dose (settling rate) ............................................................... 87
7
Figure 58 : Flocculant Dose (clarity)Error! Bookmark not defined............................................. 88
Figure 59 : Static Settling Rate ................................................................................ 89
Figure 60 : Static Settling Rate (P&C,2011) ............................................................. 89
Figure 61 : Static Mud Bed Compaction .................................................................. 90
Figure 62 : Solids Flux Curves ................................................................................. 91
Figure 63 : Hydraulic Flux Curves............................................................................ 91
Figure 64. A 3D representation of the proposed thickener design (TWP, 2011) ....... 94
Figure 65. Illustration of the constructed wetland layout (Source: DWA, 2014) ........ 97
Figure 66. Illustration of the cross sectional through the constructed wetland layout.
Inlet zone (0m – 250m), Macrophyte zone (250m – 650m) and Outlet zone (650m - 800m) 98
Figure 67. Darcy’s flow equation for saturated flow through the macrophyte zone ... 99
Figure 68. Schematic of the root zone and components of the tensiometers ......... 103
Figure 69. Photos of the above ground Cyperus papyrus in the water body .......... 104
Figure 70. Artisanal gold processing chemicals. A: Mercury and other metals used in
gold processing. B: Processing chemicals left adjacent the river ....................................... 109
Figure 71. Possible negative impacts at Farandie. A: Dredging, B: Hydrocarbon
pollution. ........................................................................................................................... 109
Figure 72. Heavily impacted faleme cobble bed due to sand mining at 13.001471°; 11.405753° ....................................................................................................................... 110
Figure 73. Photographs illustrating urban pollutants. A: washing of vehicles and
clothing; B: urban runoff/solid waste. ................................................................................. 110
Figure 74. Extensive artisanal activities in the Gara River ..................................... 111
Figure 75. Proposed surface treatment system (Source: DWA, 2014) ................... 117
Figure 76. Examples of the trials that may be undertaken as iterative process for
optimisation of the constructed wetland (Source: DWA, 2014) .......................................... 121
Figure 77. Examples of pilot scale constructed wetlands (Source: Vymazal, 2009) 124
8
3 Introduction
Loulo Gold Mine is a Randgold Resources Ltd subsidiary and has been in operation
since June 2005. The Mine is located in the Western part of Mali, 350 kilometres west of
Bamako and 220 kilometres south of Kayes and is adjacent to the Faleme River which forms
the border between Mali and Senegal. The river originates in Guinea, approximately 250km
upstream of the study area and joins another large river, the Senegal river 270 km
downstream of the mine, whereafter it flows 590 km before entering the Atlantic Ocean as
the Senegal River. The river is managed by a Basin Management Authority (OMVS –
Organisation for utilizing the Senegal River) with the Governments of Guinea, Mali, Senegal
and Mauritania having input. The river is also a vital lifeline to communities living along it and
is the only regional water supply in the area during drought conditions. Managing the effect
of the mine on this vital water resource is thus essential. The tributary streams of the Faleme
River that flow in close proximity to the mine include the Gara and Dande. The other likely
receptors of the stream water include a few villages (communities), whose sole source of
drinking water is groundwater, but utilize the stream water for other domestic and agricultural
uses. See below a map showing the location of the Loulo mine.
Figure 1. Location map (Source: RRL annual report 2012/2013)
Loulo currently consists of two underground operations (opencast operations stopped
at the beginning of 2013) – Yalea and Gara. Although dewatering is taking at both
operations, this report will focus on the gara underground dewatering. This thesis
characterizes the ecological health of the main receptor, i.e. the Faleme river adjacent to the
Loulo mine to understand the existing/potential impact of the substandard mine discharge on
it and then develops a feasible treatment solution to bring the mine effluent into compliance
with the country water discharge limits. A number of expert advice has been seeked in the
past as to the most effective way of treating the effluent prior to discharge. The last two
consultations were in January 2011 and in October 2013 when Digby Wells Environmental
9
undertook a site visit to evaluate the issues and to recommend measures. Proposed
measures were to settle the solids in a mechanical settler and construct a wetland/reedbeds
to deal with the nutrient rich effluent.
3.1.
Motivation
The Gara underground project started in 2010 and will be in operation until 2024.
Dewatering of the mine has been taking place since the beginning of the mine and the
current infrastructures in place to improve the quality of the water before discharge have
proven to be inadequate causing substandard effluent to be discharged into the adjacent
Faleme river. Looking at the sensitivity of the Faleme river in terms ownership (the faleme is
an international river) and use by the downstream communities, it was of paramount
importance to determine if the discharge was impacted on the river and to design a solution
to bring the effluent into compliance with the acceptable Malian water discharge limit and
IFC discharge guidelines in order to preserve the river ecosystem and to preserve the mine
image vis-à-vis of its stakeholders. The findings of this study could help resolve the long
standing issue related with finding a feasible cost effective technology for the treatment of
mine water effluent in the country and in other developing countries where the cost of the
sophisticated water treatment system make their implementation impossible, consequently
putting the environment at risk. The findings could be adapted to other mines of the
Randgold Resources Ltd as well.
Figure 2 below shows the Loulo gold mine and the discharge point and route of the
Gara underground effluent towards the faleme.
10
Faleme
River
Gara Settling
ponds
Discharge
route
MALI
SENEGAL
Figure 2. Loulo gold mine with the location of the Gara underground discharge point (source: 2012
satellite image of Loulo gold mine modified in arcGIS 10.1)
3.2.
Research Questions
It is evident from the above that the substandard mine effluent has been discharged
for a while and may have resulted in a possible environmental damage which could, if it is
proven to occur, have an impact on the company image and reputation and on the people
using the resources of the water not to mention the high cost that would be involved for the
remediation. Continuing to discharge substandard water into the river is a legal
noncompliance that may result in lawsuit from the regulatory bodies (OMVS and Mali
Government) and potentially fines.
11
It is therefore of paramount importance to answer the following question –

Has/Is the mine effluent caused/causing any impact on the receiving aquatic
environment (Faleme river)? If yes, to what extent? If no, what are the risks?
Independently of the first question, two second questions are to be answered as it is a legal
requirement to treat substandard effluent before discharge –

Which type of settling will work on the gara effluent? Passive or dynamic? With
flocculant or without flocculant?

Which type of wetland will work better considering the site specific parameters?
3.3.
Aim and Objectives
This study aims at characterizing the impact of the Gara mine effluent into a river
body and developing an integrated cost effective solution to treat the effluent before
discharge. The findings will contribute at developing the mine water treatment technologies
industry in a developing country such as Mali and for Randgold Resources Ltd. There are
two objectives as described below:
The determination of the impact of the gara effluent discharge on the Faleme river
ecosystems based on:




The Malian regulatory framework;
The characteristics of the mine effluent discharge
The characteristics of other non-mining related activities along the river
The current ecological health status of the Faleme river upstream,
downstream and at the mine discharge area.
The design of integrated and adapted cost effective physical settling infrastructure
and wetland to treat the mine effluent, based on:


The characteristics of the mine effluent discharge;
The site specific parameters such as climate, soil, vegetation, and land
availability
Figure 3 shows the research objectives (left) and questions (right)
12
•Has/Is the mine effluent caused/causing any impact on the
receiving aquatic environment (Faleme river)? If yes, to what
Impact of the Gara extent? If no, what are the risks?
effluent on the
Faleme river
ecosystems
•Which type of settling will work on the gara effluent? Passive or
dynamic? With flocculant or without flocculant?
Design of
integrated and
adapted water
treatment system
•Which type of wetland will work better considering the site
specific parameters?
Figure 3. Objectives (left) that answer the research questions (right)
3.4.
Methodology and the structure of the report
Figure 4 summarizes the methodology used to achieve each objective and to answer
the research questions. First, the available monitoring data will be used to characterize (flow
and quality) the gara discharge and an assessment will be undertaken to evaluate the
ecological health of the receiving aquatic system (Faleme) to establish any link between the
discharge quality and the health thereof; second, a literature review will be undertaken to
describe the state of the constructed wetlands in the water treatment industry and to review
the most critical parameters for a design in order to propose some design options specific to
our study; and third the designs of the treatment system will be proposed. Results will be
presented and discussed and recommendations made for the treatment of the gara effluent
and the way forward.
13
• Litterature review of the malian legislation related to wastewater discharge
• Characterizing the gara effluent in terms of pollutants and volume discharged
using the available 4-year data (2010-2014)
• Faleme water quality sampling and lab analysis
• Faleme sediment sampling and lab analysis
Impact of the Gara
• Faleme habitat assessment
effluent on the Faleme • Faleme benthic macroinvertebrates sampling and characterization
river ecosystems
• Faleme ichthyofauna characterisation and bioaccumulation assessment
Design of integrated
and adapted water
treatment system
• Litterature review - constructed wetlands and site characteristics
• Design of the constructed wetland
• Settling test on the gara effluent
• Design of the settler
Figure 4. Methodology
The report will consist of three parts. Part 1 will consist of describing the site in terms
of climate, geology and the mining; part 2 will consist of reviewing the legal requirements
related to wastewater discharge and evaluating the impact the mine discharge has had so
far on the Faleme river, part 3 will cover the design of the treatment options and part 4 will
deal with the discussions, conclusions and the recommendations.
3.5.
Research Scope
This research will lead to the development of conceptual designs for the treatment of
the mine effluent that eventually may need to be tested and fine-tuned through a pilot scale
field experiment before a large scale implementation. Table 1 tabulates the aspects included
in this research and those excluded. The study did not evaluate the impact of the volume of
the discharge on the state of the faleme river as it was assumed to have negligible effect
compared to the quality of the discharge. A detailed assessment of the other activities (e..g.
artisanal mining activities etc.) along the Faleme and in its catchment to determine the other
potential sources of pollutions was not undertaken even though there is a broad knowledge
and mention thereof. A literature review on the different wastewater technologies was not
needed as experts recommended the construction of a thickener to remove solids and the
constructed wetland to remove the nitrate and heavy metals. This report determines the
design requirements to achieve the final water objective. Unfortunately due to time constraint
and the complex nature of water treatment systems – the research stop short of providing a
fully tested concept to be directly used. The report includes aspects related to the quality
impact of the discharge and its impact on the faleme river ecosystem that is the habitats,
ichthyofauna, the sediments and the water quality. The sizing of the settler is limited to
14
technical tests data, analyses and recommendations provided by Patterson and Cooke. The
study focuses on the design requirements of a subsurface flow constructed wetland type.
Table 1. Scope of the research
Included
Excluded
Impact of the gara effluent Effluent quality monitoring Quantity impact
on the Faleme ecosystem
data over 4 years and during
Detailed assessment of the
the survey
non-mining activities and
Aquatic assessment results potential impact
of the Faleme adjacent to
site
Limited or broad information
on the artisanal mining
activities
Design of the constructed The sub surface flow wetland Literature
on
water
wetlands
treatment technologies in
system
use in the mining industry
Conceptual models to be
behind
the
tested based on literature Rationales
selection of a wetland to
research
remove nitrate
Testing of
parameters
Design of the settler
the
design
behind
the
settling Rationales
selection of a thickener to
remove the solids
Sizing results of the settler
Results
testwork
of
the
4 Background information
4.1.
Regional Climate
The climate at Loulo is strongly influenced by the northward and southward
movement of the Inter Tropical Convergence Zone (ITCZ), which creates distinctive wet and
dry seasons. The site falls within the Sahelian transition zone between the Sahara Desert in
the North and the tropical climate in the South. The low altitude of the site and the absence
of any intervening mountains mean that the humidity is directly conveyed to the area by the
wind. Climate data was obtained from the Kéniéba weather station (12.85°N, 11.20°W; 132
m above mean sea level) located 35km south of Loulo Mine and from data collected on site.
15
4.1.1. Temperature
Western Mali is extremely hot for most of the year. The average maximum
temperature varies from 30.6ºC in August to 40.4ºC in April. The average minimum
temperature varies from 15.8º in December to 27.1º in May. Figure 5 below indicates the
monthly average temperatures taken over a 31 year period from the Kéniéba weather
station.
Figure 5. Average monthly temperatures (Source: DWA, 2007)
4.1.2. Rainfall and Evaporation
Rainfall data was obtained from the Kéniéba weather station for the period 1969 to
2005. The results captured indicate that there is a distinctive rainy season from May to
October with little or no rainfall from December to April (Figure 6). The highest rainfall levels
(above 400mm) were recorded during the month of August in the years 1969, 1973, 1981,
1982, 1999 and 2000. During the time period 1969 to 2005 the months of December,
January and February; negligible or no rainfall was recorded.
Evaporation data was obtained from the Kéniéba weather station for the 1970 to
1996 period. The trend for evaporation was opposite to rainfall with higher evaporation over
the dry, sunny months of November to May where evaporation exceeded rainfall and lower
evaporation for the overcast and rainy season from June to October where rainfall exceeded
evaporation (Figure 7). During this period, February 1976 showed highest total evaporation
(331.5 mm) and August 1971 showed lowest total evaporation (49 mm).
16
Figure 6. Average monthly rainfalls (Source: DWA, 2007)
Figure 7. Average monthly evaporation (Source: DWA, 2007)
4.1.3. Wind and sunlight hours
The north and southward movement of the intertropical convergence zone is a prime
determinant of the wind direction at the site. Wind speed and direction data was obtained
from the Kéniéba weather station for the 1970 to 1996 period (Figure 8). The general wind
direction changes according to the seasonal shifts of the intertropical front. During the dry
season the intertropical front is located to the south and the dominant wind is north to east
(Harmattan). In the rainy season the front is located to the north of Mali and the dominant
17
wind is the north to west (Monsoon). The Harmattan is responsible for carrying large
amounts of fine dust from the Sahara over large parts of Northern Africa. For a few months
of the year, this has a large impact on activities by reducing visibility. This generally clears
up after the first rains.
Figure 8. Average and maximum wind speed and direction (Source: DWA, 2007)
There are not many cloudy days at Loulo giving a high number of sunlight hours
throughout the year. The monthly averages have been obtained from cumulative monthly
totals for the 1970 to 1996 period from the Kéniéba weather station and are indicated in
Figure 9.
18
Figure 9. Average monthly sunlight (Source: DWA, 2007)
4.2.
Geology and Mining
4.2.1. Petrography
The regional geological setting is well described in the environmental baseline report
done in 2007 by Digby Wells Environmental. Figure 10 gives a geological map of the region.
The area is located within the Kéniéba inlier of Lower Proterozoic (2.1 Ga) Birimian
metasedimentary-volcanic sequences. The stratigraphic sequences of the inlier include from
west to east the volcanic and sedimentary units of the Kenieba Formation, the continental
shelf sedimentary sequences of the Saboussire Formation and the shallow to deep marine
sedimentary units of the Kofi formation. These formations have been subjected to a series of
deformation events of which the principle one is characterized by a N-NE trending schistosity
with associated structures. The most prominent of these is the Senegal – Malian shear zone,
a major first-order lineament trending N15°, which can be traced for over 200 km in Mali and
Senegal and forms the contact between Upper and Lower Proterozoic sequences in the
vicinity of the Sadiola deposit. The geology of the Kéniéba Inlier is also characterized by a
series of syn and post tectonic granitoid to dioritic intrusives associated with the major
structural zones.
The current geological model for the area of study involved the juxta positioning of
the various formations as a result of generation of a series of accretionary wedges related to
the development of back-arc, arc, shelf and basin successions generated by low angle
subducting tectonic plates. Crustal melting relating to this collision produced a whole suite of
evolving extrusive and intrusive bodies and the generation of hydrothermal gold and
sulphide-bearing fluids.
The Senegal – Mali shear marks a major break in the geology from shelf carbonates
with the Falémé ironstone unit in the west to the sedimentary sequences of the Kofi
formation in the east. East of the Senegal – Mali shear the sediments of the Kofi formation
are characterized by the following: limestone’s and shallow shelf sandstones followed by
cyclical succession of greywacke with intercalated mudstone, quartzite, marl and limestone
which progressively evolve eastward into a classic turbiditic sequence (see Figure 11
below). This sequence is transacted by three major N-NE orientated structures referred to as
the (1) Sakola shear zone, (2) Yalea shear zone and the (3) Senegal – Mali shear zone.
These major lineaments have been subjected to major strike slip movement with associated
antithetic structures and often resulted in the repetition of the stratigraphy. There is a
coincidence of major gold targets and regional soil anomalies associated with these major
shear zones and their antithetic structures.
19
Figure 10. Geological Map of the Kenieba window (source: DWA, 2007)
20
Figure 11. Geology in the study area (unpublished map from Joe holiday, 2014)
4.2.2. Mineralogy
Lawrence et al. (2013) described the mineralization of the Gara ore, our area of
concern – Figure 11 below shows the detailed Gara geology – he described that Gara forms
21
a carbonate-quartz-sulphide vein stockwork deposit. The Gara stockwork contains high
carbonate concentrations (60-70 vol.% of the veins). This is an atypical characteristic for
vein-hosted orogenic gold deposits, where carbonate concentrations are usually ≤5 to 15
vol.%. Mineralised veins are composed of Fe-Mg bearing carbonates (ankerites), while
carbonate constituents of the barren veins are mainly composed of calcite. Quartz is the
second major component of the Gara veins. Two separate styles of quartz generation are
observed. A broad milky quartz generation exists as a cogenetic phase with ankerite or
calcite, while a late grey quartz generation can be seen sealing brecciated ankerite veins
(reactivation phase; Figure 12. c). Sulphides mainly occupy 5-30 vol.% of the veins (see
Figure 13) and are situated in both carbonate and quartz vein material. Gangue minerals
include albite and minor chlorite. The most common accessory minerals are rutile and
apatite. Rutile exists as acicular, blocky and skeletal grains (20-80 μm) dispersed within
quartzcarbonate vein material. Apatite increases in abundance in carbonate-dominated
veins and contain micro-inclusions (<10 μm) of scheelite and monazite. In terms of sulphide
phases, Gara contains a fairly simple ore petrogenetic history with pyrite occurring as the
principal sulphide mineral (95-99% of total sulphides) and the only gold-bearing sulphide
phase. Chalcopyrite is the main accessory sulphide, occurring as a trace phase (<5%). Two
generations of chalcopyrite are observed: an early phase forming contemporaneously with
pyrite (chalcopyrite-I); and a later, more dominant, phase post-dating pyrite formation
(chalcopyrite-II). Other accessory sulphides, in order of decreasing abundance, include:
pyrrhotite, gersdorffite, pentlandite and arsenopyrite. No Sb, Pb or Zn-bearing sulphides are
observed at Gara. The Gara mineralised veins also contain common rare earth minerals with
similar concentration levels to chalcopyrite (monazite and xenotime coexist with pyrite).
Other accessory minerals include scheelite (a common ore phase at Loulo).
Pyrite is host to a range of trace Fe-Ni-As sulphides. Pyrrhotite and gersdorffite are
the most common phases.
22
Figure 12. Grab and drill core samples of the Gara vein stockwork. The photographs illustrate the
composition of the veins and the range of vein morphologies that occur in the tourmalinite
host. A) A pit sample from a high-grade zone showing the presence of ankerite-pyrite
veins. B) A sulphide-rich ankerite vein associated with high-high grade (90.3 g/t), showing
clearly the link between gold concentrations and carbonate-pyrite abundance. Limonite
alteration (brown material on the right of the image) occurs after pyrite (LOCP124, 552 m).
C) Sigmoidal mineralised vein showing reactivation and brecciation of an early carbonate
vein (white material) by later dull grey quartz, with both vein phases associated with
sulphide generation (LOCP97, 347 m). D) A strongly brecciated carbonate vein with the
vein structure no longer present (LOWDH23, 441 m). E) Straight barren (0.11 g/t) milky
quartz veins (minor carbonate). The low-grade can be attributed to the lack of the vein
carbonate and the low number of veinsets (2 vein directions, perpendicular to each other)
23
(LOCP49, 180.6m). F) A folded barren calcite vein (LOCP81, 676.55m). Folded veins
usually occur within the weakly-tourmalinised quartz-wacke (note the lighter colour of the
host rock compared to A to D) and pre-date mineralisation (D1 to D2 origin) (Source:
Lawrence, 2013).
Figure 13. Back-scattered electron (BSE) images of the major and minor sulphide phases present at
Gara. A) Highly altered pyrites replaced by late stage tourmaline (discussed in chapter 5)
(LOCP117, 338 4 ) B) V it l d b t li (LOCP117 338 4 ) C) Z d it t l 10 μm 20 μm 338.4 m).
Vuggy pyrite replaced by tourmaline LOCP117, 338.4 m). Zoned pyrite crystal showing
lighter zones of As-Ni bearing pyrite. Small inclusions of arsenopyrite are associated with
these As-Ni bearing zones (LOCP124, 557.6 m). D) Euhedral arsenopyrite intergrowth with
24
pyrite (LOCP59, 535.4 m). E) Anhedral chalcopyrite-I inclusions situated within the core of
a pyrite crystal (LOWDH23, 438.05 m). F) Late chalcopyrite-II sealing fractures within pyrite
or located along pyritegersdorffite contacts (LOWDH23, 438.05 m). Abbreviations- Carb =
carbonate; Tur = tourmaline; Py= pyrite; Aspy = arsenopyrite; Cpy = chalcopyrite; Gf =
gersdorffite. (Source: Lawrence, 2013)
4.2.3. Mining
The Loulo Underground mine uses conventional drill and blast to extract the ore from
the stopes and to develop the underground workings. The mining consists of drilling
blastholes, charging them up with the explosives (Rioflex, ANFO or emulsion), blasting,
washing down the broken rock, scaling, supporting, mucking and transporting the ore to
crusher and then to surface while waste is used to backfill open voids underground. Other
activities that happen concomitantly to the mining are the geological drilling (deep probing
holes) and dewatering. Total decline metres developed to the south bottom is 3414m and
vertical distance developed is 420m (survey department, 2014). Monthly averages of 181
tons of rioflex and 35 tons of ANFO (ammonium nitrate and fuel oil, diesel) are used in Gara
(2012-2013
explosive
record).
Figure 14. Gara underground layout showing the portal from the gara pit into undergrounds through
spirals declines (Source: gara ore reserve statement report, Dec 2012)
4.2.4. Dewatering
The dewatering in the underground is such that all waters (clean or dirty) are drained
to lowest points on every level and pump to the underground pumping station sump where it
is mixed and pumped into the surface settling ponds – this makes dewatering the main
pollutants pathway in the event of any pollution. Below is a figure showing the dewatering
layout from underground to surface dams.
25
Figure 15. Underground water reticulation
Potential pollutants sources
The potential contaminants to expect in the discharge from the underground
workings include chemical elements from explosives used, from the ore and to a lesser
extent from the host rock, hydrocarbons and the emissions from the fleet, and sediments
from the drilling and blasting activities. Release mechanisms into the effluent include
emissions and spills from the fleet (including drilling rigs), washing explosives materials not
fully detonated or spills during chargeup, and dust particles generated during drilling. Below
are the different pollutants, their sources and exacerbating factors.
5 Regulatory framework
5.1.
Mining code
The mining in Mali is regulated by the mine code (Law no. 2012-015/AN of 27
February 2012) which states in his article no. 142 that all owner of a mining permit must
comply with the Malian legal and regulatory requirements related to the Environment.
5.2.
Convention
The convention (1993) between the mine shareholders and the Government also
says in its article 20 (refer to art. 20 hereafter) that the mine must:

Art. 20.c. – At any time comply with the legal requirement related to
hazardous waste, Natural resources and the protection of the environment.

Art.20.f. – establish a treatment system for its wastewaters (the discharge
guidelines are defined in the document MN-03-02/002:2006)
26
5.3.
Water code
In Mali, the law that covers water related issues is the water code (Law no.02-006/AN
of 31 January 2002) and its implementation decree No. 04-183/P-RM of 11 June 2004. It
describes all the requirements related to water resource protection on the Malian territory –
of interest to us are the articles 14, 16 and 60 that stipulate that it is forbidden to discharge
pollutants into water courses which could threaten public, aquatic and flora health – anyone
whose activity can cause pollution of the water course should take the necessary measures
to prevent it. Polluters must pay for the cleanup costs.
5.4.
Pollution and Nuissance prevention Laws
Seeing above and the other legal requirements such as those related to pollution and
nuisance (Law no. 01-020/AN of 30 May 2011) and those related to the required measures
for the management of wastewaters in Mali (Decree no. 01-395/PRM of 06 September 2001)
– the mine must ensure the underground mining effluents comply with the acceptable
discharge limits before discharging it into the Faleme river which is under the regime of the
water charter (OMVS – Organisation for utilizing the Senegal River). OMVS has no guideline
for the discharge water quality which leaves us with the Malian and the International Finance
Corporation (IFC) guidelines to comply this. Below is a table showing the different standards
and guidelines used in this study.
Table 2. Water quality standards and guidelines
Guidelines
Drinking water
Effluent discharge
WHO Drinking water
guidelines (Gorchev &
Ozolins 2008)
IFC EHS Mining
guidelines (IFC 2007)
Malian Discharge Standards
≤1
Al
0.2
As (soluble)
0.01
As (total)
0.01
0.1
≤ 0.05
BOD
50
50
≤50
Cd
0.003
0.05
≤ 0.02
Cl
250
COD
250
150
≤150
EC (uS/cm)
-
-
≤2500
-
-
≤ 1200.0
Cr 3
2
≤ 0.2
Cr (hexavalent)
-
Cr (total)
0.05
0.1
Cu
2
0.3
≤ 0.1
CN (free)
0.07
0.1
-≤0.5
CN (total)
0.07
1
≤1
CN WAD
-
0.5
-
27
Guidelines
Drinking water
Effluent discharge
WHO Drinking water
guidelines (Gorchev &
Ozolins 2008)
IFC EHS Mining
guidelines (IFC 2007)
Malian Discharge Standards
Faecal Coliform
0
-
≤12000
Escherichia coli
0
-
-
Total Coliforms
0
-
≤ 20000
Fe
0.3
2
≤2
F
1.5
-
≤6
Hg
0.006
0.002
≤ 0.005
Ni
0.07
0.5
≤ 2.0
Mn
0.4
-
≤2
PO4
-
-
≤ 10
NO3
50
-
≤ 30
NO2
3
-
Oil andgrease
-
10
-≤5
Pb
0.01
0.2
≤ 0.2
pH
6.5-9.5
6-9
6.5-9.5
Phenols
200
0.5
≤ 0.5
Se
0.01
0.1
-
Sn
10
Na
200
-
-
SO4
250
-
≤ 1000
S
1
Sulphite
1
TDS
1000
Temp Increase
-
≤ 1000
<3
≤40
Temp. Total
Total Hardness
-
-
TSS
-
50
≤ 30
Turbidity (NTU)
5
-
-
Zn
-
0.5
-≤0.5
Ag
-
-
≤ 0.01
V
-
-
≤ 1.0
B
0.5
-
≤ 1.0
Dissolved oxygen
>6
Chlorine
≤0.2
Ammonia
≤15
Animal and veg. fat
≤20
Aromatic solvent
≤0.2
The elements of concerns, against the above table, in the discharge have been
described below in the section on the characterization of the underground effluent.
28
6 Characterization of the Underground Effluent
6.1.
Flow volume
6.1.1. Historical data
The records available cover three years (January 2011 to January 2014) of recording
using ultrasonic flowmeters (Innova-Sonic model 205 thermal energy) on the discharge lines.
Although there were instances where the flow volume was read over 24-hour period,
most of the times the flow data were read irregularly (48-hour, 72-hour and sometimes
monthly cycles) making it difficult to evaluate the peak and low flow 24-hour discharge
volumes and instantaneous variations. The data was averaged over the number of days
accumulated between two subsequent readings – this was found to provide a fair
representation of the reality which will allow the design of a treatment system capable of
handling the expected volumes and taking into account the fluctuations in volume. Also any
value below 100 m3 was removed from the data set. In total a data set of 754 out of 844
values was used. The Statistical description of the data is shown below in Table 3 together
with a bar graph showing the monthly volumes (see Figure 16) –
3
Table 3. Descriptive statistics of the data set (value in m )
Minimum
Maximum
Mean
Median
First Quartile
Second Quartile
Third Quartile
Fourth Quartile
102
9538
1671
1486
533
1485.6
2364.3
9538
140000.0
120000.0
100000.0
80000.0
60000.0
40000.0
20000.0
0.0
Jan-14
Nov-13
Sep-13
Jul-13
May-13
Mar-13
Jan-13
Nov-12
Sep-12
Jul-12
May-12
Mar-12
Jan-12
Nov-11
Sep-11
Jul-11
May-11
Mar-11
Jan-11
Figure 16. Monthly flow data from January 2011 to January 2014 (source: water monitoring database
of Loulo Gold mine)
29
The maximum daily flow data analysis returned 9,538 m3 as the highest daily flow
and was recorded in April 2012; the minimum and the average daily volumes were 102 m3
and 1,671 m3 respectively.
6.1.2. Future Flow prediction
Methodology
A geohydrological investigation was undertaken in 2012 around the gara
underground areas to determine the volume of water inflow underground life of mine. Digby
Wells Environmental was appointed to assist. The author participated in all data collection
and interpretation of the results. The scope of work has been conducted through several
tasks as defined in the final report submitted by Digby Wells, and included the following:

Assessment of available geological and hydrogeological data;

Ground geophysical survey around the pit and underground areas;

Extended drilling of candidate dewatering holes in the perimeter;

Refinement of the conceptual groundwater flow model;

Construction and calibration of a numerical flow model; and

Determination of the required dewatering rates and number of dewatering
boreholes to ensure workable conditions at Gara underground.
The aquifers in the vicinity of Gara were studied based on recent drilling and preexisting boreholes, in conjunction with the structural geology established at Gara. A total of
12 boreholes were sited and drilled at Gara. Details of all drilled boreholes are presented in
Appendix A. The figure below indicates the position of the different boreholes
30
Figure 17. Boreholes locality map (Source: DWA, 2012)
Previous conceptual hydrogeological models of Gara aquifer system suggested is a
two-aquifer system, consisting of a fractured aquifer overlain by a weathered aquifer, where
groundwater flow mimics surface topography (DWA, 2007). It is evident that gold
mineralisation in Gara was structurally controlled. The geomorphology of the local drainage
system is highly controlled by the fault architecture. Surface water flowed through and
eroded open fractures in exposed damage zones (zone of subsidiary structures surrounding
a fault).
The fractured aquifer consists of a rock matrix and a fracture network. In rocks with
low primary porosity (in the rock matrix), groundwater flow only occurs in the secondary
porosity provided by fracture networks and dissolution associated with fault zones. Within a
typical fractured aquifer the secondary porosity, i.e. fracture network, dominates the
groundwater flow, while the primary porosity, i.e. rock matrix, dominates groundwater
storage. Based on the current drilling and reassessment of historic geological and
hydrogeological data, the groundwater system cannot only be described in terms of an
elevation or stratigraphic units, as traditional aquifers are, but instead by the relationship with
the fault (and the nature of the fault architecture). The fault damage zones and fracture
networks in Gara study are so large and interconnected, that they allow substantial
groundwater systems to be hosted in a system almost entirely controlled by secondary
(fracture) permeability. However, some of the fault zones form an impermeable barrier
caused by the breccia re-cementation and gouge development. Breccia zones demonstrate
additional strength (resistance to erosion when fractures are cemented). This effectively
fuses cracks and heals fractures which creates barrier for fluid flow.
31
High fracture density area
Faults intersection bounding damaged zone
Figure 18. Damage zone with fault intersection - north wall of Gara pit (Source: DWA, 2012)
The fractured aquifer system around the pit is highly compartmentalised by faulting
both laterally and vertically, as depicted by the sudden change in hydraulic head over short
distances. A steep change in groundwater level occurs between boreholes 16-420N and 16540N. Groundwater levels changes from 84mamsl, at borehole 16-520N, to 19mamsl at
borehole 16-420N. These boreholes which are just 75m apart have a collar elevation
difference of 1m. A SE structure, SSE of Gara pit, also seems to act as a flow barrier. The
groundwater head drops to 15m below mean sea level SSE of Gara pit. Intersection of any
given fracture will only allow drainage of the rock matrix immediately adjacent to the fracture.
To achieve reasonable dewatering in the current decline, it was recommended that the
decline shear zone has to be depressurized. DWA06 and DWA08 were perfect candidate
boreholes for depressurization Gara decline shear zone. They might significantly reduce the
underground water ingress, and hence underground dewatering requirement in this area.
The rock matrix west of Gara pit has a very low drainable porosity. All water is
transmitted in fractures (the main fracture zone and interconnected microfractures). Main
fracture zone north and west of Gara pit, typically yield 1.3 to 2 L/s, sufficient to sustain flow
in boreholes. The multi-layered SQR beneath the limestone west of Gara pit is intersected
by a dolerite sill. Potential recharge to the aquifers beneath is partially blocked by the
dolerite sill.
Groundwater flow in shallow aquifers, controlled to some extent by relic structures, is
much more homogenous than flow beneath. There are no discrete highly transmissive
fracture zones. The fracture zones and matrix have both been completely weathered.
32
Compartmentalisation is much less important than in the deeper fractured aquifer. There is
also low lateral seepage from these aquifers to the pit walls. Isotope studies concluded that
the Falémé River has little or no contribution to Gara underground inflows. A steady state
groundwater flow model for the study area was constructed to simulate current
hydrogeological conditions at Gara study area. These conditions serve as starting heads for
the transient simulations of groundwater flow. Visual MODFLOW 2010.1, a MODFLOW
based modelling software package, was used for the simulations. MODFLOW and Visual
MODFLOW are internationally recognised modelling packages that have been proven to be
capable of simulating these types of groundwater flow and contaminant transport
assessments to a high level of accuracy. The simulation model is based on threedimensional groundwater flow equation.
A model, no matter how sophisticated, will never describe the investigated
groundwater system without deviation of model simulations from the actual physical process
(Spitz, 1996). The list of assumptions made can be found in the full report (DWA, 2012). The
model covers an area of 10.7 km x 9.7 km. The model is a finite difference model. The
individual cell sizes vary from 25 m x 25 m, within the vicinity of the Gara pit and
underground, to a maximum of 100 m x 100 m in the outer extremes of the model area
where less accuracy is required and dewatering impacts are expected not to be as
pronounced.
The numerical model design incorporates river/aquifer interaction features to enable
representation of both baseflow and recharge from the streams to the groundwater. The
Falémé River, Gara River and the creek in the southern boundary were represented using
the River (RIV) package, where river bed elevation, river stage and conductance of the
riverbed were specified. Groundwater does not flow across the Gara pit, thus the pit area
was represented internally with no flow boundaries. The area covered by the no flow
boundaries in the pit area was progressively reduced from layer 1 through to layer 4. Drain
boundaries set at the pit walls, in areas where groundwater seeps into the Gara pit. The
decline actuals were incorporated into the model as drain cells, to simulate steady state
groundwater flow condition at the end 2011. A regional recharge value calibrated from 2007
model 1.5E-6 m/d was used as recharge to the model.
33
Figure 19. Model setup (not all layers are shown) (Source: DWA, 2012)
The observed heads were used as calibration criteria. The inflow rate into the
modelled decline was compared to the average observed discharge rate from the Gara
underground. Steady state calibration of the model was accomplished once the flow and
head criteria were showed a reasonable resemblance to the observed values.
The predictive model produced time-variant dewatering requirements for the Gara pit,
and underground monthly developments from 40 Level to 285 Level, between January 2012
and December 2014. The mine plan from January 2015 onwards is not detailed.
Figure 20. Gara Underground mine plan (rotated 110̊ clockwise to show decline) (Source: Loulo Gold
mine, 2012)
Results
34
The predicted inflows increase with increasing depth below surface. This is due to
increases in the groundwater flow gradients and the flux area contributing to ingress into the
underground workings) as new levels are being developed. The model predicted a total
underground inflow of 3600m³/d for March 2012. A maximum inflow of 6100m³/d is expected
in November 2012. The cumulative effect of the unknown details of the mine plan after 2015
is depicted by an unusual rise in total influx from January 2015. However, the predicted
maximum inflow of 17000m³/d in January 2015 is not expected to occur. The findings of the
model are summarized below:

The groundwater model is more sensitive to the storage of the aquifer (than the
recharge and transmissivity). Increase in confined and unconfined storage may
increase the dewatering requirement by up to 55%. The actual storage coefficient will
be evaluated after aquifer tests.

Groundwater inflow into the actual decline is expected to drop from a maximum of
820m³/d in January 2012 to a minimum of 600m³/d in October 2012, after which it will
range below 620m³/d till the end of mining.

Groundwater inflow into developments in 40 Level is predicted to drop down to
34m³/d by October 2012. A maximum inflow, 300m³/d, into producing stopes of 40
Level is expected to occur in March 2012. Little or no seepage is expected into any
40 level developments from November 2012 onwards.

A maximum inflow of 1300 m³/d into 65 Level is expected to occur in March 2012. 65
Level inflows are expected to decrease to 12.6m³/d by the end of 2014.

Inflow into 85 Level is predicted to rise to a maximum of 1760.8m³/d in October 2012,
then decrease to 560m³/d at by the end of 2014.

Groundwater inflow into 110 Level is significantly less than the other levels. 110
Level inflows are not expected to rise above 500m³/d during mining. A maximum of
420m³/d is predicted to occur in October 2012.

Inflow into 135 Level is expected to be below 500m³/d in 2012. 135 Level inflows are
predicted to rise to 1020m³/d in June 2013, after which it decreases to 780m³/d at the
end of 2014.

An average inflow 560m³/d is predicted for the first three years of mining at 160
Level, during which a maximum inflow of 748.1m³/d is expected to occur in April
2012.

Groundwater inflow rates between 1000m³/d and 1200m³/d are predicted in 185
Level from November 2012 to January 2014. Higher inflows are expected from
January 2015 as more areas in 185 Level are mined.

A maximum inflow of 460m³/d into 210 Level is predicted between 2012 and 2014.
Higher inflows in should be expected from January 2014 onwards.
35

The predicted inflows from 235 Level to 285 Level range below 500m³/d, till the end
of 2014. Major developments in these levels are scheduled to occur after 2014.
These will result to elevated inflows, and dewatering requirements.

The model predicts decreasing groundwater inflow into Gara pit as the underground
mine develops. As more levels are mined, the groundwater gradient towards the
open pit walls are reduced, thus the reduction of inflow. The pit inflows are expected
to drop to 430m³/d by the end of 2014, and even less from 2015 onwards. It should
however be noted that during the rainy season, pit dewatering requirements are
expected to increase. Sump pumping will be suitable to handle rainfall events.

There are seven candidate dewatering boreholes at the Gara pit. They include:
DWA06, DWA08, DWA09, 3-850, 6-280N, 16-540N, and 16-940N. The maximum
borehole depth for these boreholes is 200mbgl. This correlates to fractured systems
extending to 110 Level.

Maximum pumping at a rate of 6100m³/d is required during the first three years of
mining to ensure workable conditions underground. During this time, Gara pit
dewatering requirement would decrease to 430m³/d.

The progress of underground mining will reduce the pressure in the pit walls,
reducing inflow into the pit as the underground mine advances. The two sumps in
Gara pit can effectively handle pit inflows. If however inflows into the pit are to be
reduced by any perimeter borehole, 16-540N should be used to reduce in-pit
dewatering requirement by 40%. 16-940N can also be used in conjunction with 16540N, the cumulative effect will reduce inflows through the north wall damaged zone.
Data calibration
Steady state calibration of the model was accomplished once the head criteria
showed a reasonable resemblance to the observed values. The calculated versus observed
groundwater levels are depicted in Figure 21. The steady state water balance is shown in
Figure 22. As depicted in Figure 21 the calculated groundwater levels plot close to the 45
degree line which represents the perfect fit between calculated and observed values. Eight
of the thirteen calculated groundwater levels fall within the 95% confidence interval, while the
five plot within the 95% interval. The 95% confidence interval show the range of calculated
values for each observed value with 95% confidence that the simulation results will be
acceptable for a given observed value. The 95% interval is the interval where 95% of the
total number of data points is expected to occur.
The maximum residual (difference between calculated and observed groundwater
level) is 11.63m as calculated for borehole DWA06. The minimum residual is -0.172m at
borehole BH12. The absolute mean residual is calculated to be 4 m, indicating that on
average the calculated groundwater levels are 2.7m above the levels measured in the field.
An inflow rate of 2,199.5m³/d, comparable to the average discharge rate of
2,528m³/d, was simulated as steady state groundwater inflow rate into the modelled decline.
36
The hydraulic conductivity zones used to achieve calibration are depicted in Table 4 and
Figure 22.
Figure 21. Calculated vs. observed groundwater levels
Table 4. Steady state calibrated hydraulic conductivity (K) values
Zone
Khx(m/d)
Kv(m/d)
Layer 1 matrix
3.92E-03
3.92E-05
Layer 2 matrix
9.00E-04
9.00E-05
Layer 3 matrix
2.00E-04
2.00E-05
Layer 4 matrix
5.00E-04
5.00E-05
Layer 5 matrix
3.30E-04
3.30E-05
25
4.80E-04
4.80E-05
NNE trending fault
2.50E+00
2.50E-01
NE trending fault
1.50E+00
1.50E-01
NW trending fault
1.00E-04
1.00E-05
EW trending fault
5.00E-02
5.00E-03
5.00E-01
5.00E-02
1.00E-04
1.00E-05
Layer
6
to
matrix
DTM
north of pit
Structure
Dolerite NW of pit
37
Zone
Khx(m/d)
Kv(m/d)
Dolerite SW of pit
1.00E-05
1.00E-06
Figure 22. Steady state water balance
6.1.3. Conclusion
Historical data averaged over time returned 1671m3/day and the predictive model
over the life of mine indicated an average flow of 2528 m3/day.
6.2.
Water quality
Discharge quality data has been generated monthly over 4 years (2010 - 2014) –
water sampling has followed standard sampling procedures and analyses were done at the
SGS Laboratory in Ghana (an ISO 17025 certified lab). The results are shown in Appendix
B. chemical elements of concerns include nitrate (NO3), total suspended solids (TSS), and to
a lesser extent heavy metals (Aluminium, Iron, and traces of copper, Arsenic, cadmium etc).
Below is a series of graphs showing the elements of concerns compared to the Malian water
discharge quality limit (see Table 2).
38
Figure 23. Variations of total suspended solids (TSS) in the final discharge
As shown above the maximum suspended solids (TSS) value recorded was 9870
mg/l in May 2013 and a minimum value of 18 mg/l in January 2012. Average value recorded
from 2010 to 2014 is 1435.7 mg/l (largely above the accepted limit as indicated in Appendix
B).
A grab sample of solids in the surface settling pond was analyzed in 2012 in South
Africa and in 2014 at the laboratory of the University of Sciences and Techniques (FAST) in
Bamako to determine its metal content (results are summarized in Table 5). Arsenic (As) and
Manganese (Mn) were determined to be elevated compared to limits established by
McDonald in 2000. The Gara ore being relatively Arsenic (As) depleted (Lawrence et al.,
2013), that being locked up in pyrite (<1 wt.% As), the 64-78 mg/kg (vs. 33 mg/Kg as
guideline) determined in the settling ponds maybe due to localised arseno pyrite zones
which hasn’t been picked up in previous petrography work. Another analysis done by
Patterson and Cooke on sediment in the effluent indicated no particular issue in the solids.
The high Manganese (Mn) levels (2050 mg/Kg vs. 1100 mg/Kg guideline) are likely sourced
from the Kofi limestones, which are common along the Senegal-Mali Shear Zone. The
effluent sediments are classified as toxic according to McDonald et al. (2000).
Table 5. Effluents Solids analyses results for heavy metals
Mn
Cr –
(ppm)
V
(ppm)
Ba
(ppm)
As
(ppm)
Mo
(ppm)
Nb
(ppm)
Zr
(ppm)
Sr
(ppm)
Rb
(ppm)
Cl
(ppm)
Au
(ppm)
Zn
(ppm)
-(ppm)
2012
Survey
Sample 1
(FAST,
2014)
Sample 2
2050
Cu
(ppm)
55
64
462
226
498
72
7
6
90
129
36
526
< 20
79
< 15
501
197
414
78
8
6
93
122
23
< 70
< 20
60
< 15
39
Mn
Cr –
(ppm)
V
(ppm)
Ba
(ppm)
As
(ppm)
Mo
(ppm)
Nb
(ppm)
Zr
(ppm)
Sr
(ppm)
Rb
(ppm)
Cl
(ppm)
Au
(ppm)
Zn
(ppm)
-(ppm)
(FAST,
2014)
Aquatic
ecosystem
integrity
survey,
2014
Guideline
(McDonald,
2000)
Cu
(ppm)
170
35
74
1100
111
33
38
22
149
As shown below, the maximum value recorded was 51.3 mg/l (May 2013) and the
last aluminum value was at 4.79 mg/l which is above the Malian discharge limit of 1mg/l.
Figure 24. Variations of Aluminium (Al) in the final discharge
Iron level is fluctuating around an average of 19.8 mg/l with a maximum of 110 mg/l
recorded on May 2013 - the limit is 2 mg/l – see graph below.
40
Figure 25. Variations of total Iron (Fe-Total) in the final discharge
Nitrate average value since 2010 is 159.2 mg/l – vs. a limit of 30 mg/l. the variation
is shown in the graph below
Figure 26. Variations of Nitrate (NO3) in the final discharge
Human health impacts is the norm used for nitrate standard determination in the
South African Drinking Water Standard and Target water quality guideline for domestic use,
which is set at 6 mg/L as N (equivalent to 26.6 mg/l as NO 3 ). The reason for this is that
Nitrate readily converts in the gastrointestinal tract to nitrite as a result of bacterial reduction.
Nitrite, upon absorption, combines with haemoglobin, the oxygen carrying red blood
pigment, to form methaemaglobin, rendering the blood incapable of carrying oxygen – a
disease known as methemoglobinanemia. Arterial blood with elevated methaemaglobin
levels has a characteristic chocolate-brown colour as compared to normal bright red oxygen
containing arterial blood. Values between 6 and 20mg/L as N could lead to
41
methaemaglobinanemia in infants, while values above 20mg/L will cause
methaemaglobinanemia in children, and mucus membrane irritations in adults. Signs and
symptoms of methemoglobinanemia (methaemaglobin >1%) include shortness of breath,
cyanosis, mental status changes, headache, fatigue, exercise intolerance, dizziness and
loss of consciousness. Severe methemoglobinanemia (methaemaglobin >50%) patients
have dysrhythmias, seizures, coma and death. Metabolically, nitrates may also react with
amines and amides, commonly found in food such as meat to form nitrosamines, which are
known carcinogens (cancer causing agents), and which can lead to especially stomach
cancer (SABS, 1996).
Sulfate level is below the guideline of 1000 mg/l but has increased since 2010. The
levels should continue to be monitored as it is possible that they may increase over time as
the mine working ages and continue to access new areas exposing new faces to oxidation.
A graph below shows the variation of sulfate levels.
Figure 27. Variations of Sulfate (SO4) in the final discharge
.
The Summary quality data of water being discharged from the underground before
(Table 6) and after (Table 7) current treatment is shown below:
Table 6. Quality data of the discharge before settling (2010-2014)
unit
Max
Min
Average
limits
Aluminium
mg/l
162
0.04
18.8
1
Nitrate
mg/l
428
2.02
134.7
30
Suspended
Solids
mg/l
17000
0.01
2130
30
Sulphate
mg/l
369
20
228.5
1000
Iron
mg/l
242
0.7
27.6
2
42
Table 7. Quality data of the discharge after existing settlers
unit
Max
Min
Average
limits
Aluminium
mg/l
51.3
0.41
13.4
1
Nitrate
mg/l
520
49.2
159.2
30
Suspended Solids
mg/l
9870
18
1435.7
30
Sulphate
mg/l
471
26
248.1
1000
Iron
mg/l
110
0.8
19.4
2
From what precedes, it is evident the quality of water being discharged into the
adjacent river is substandard (compared to the Malian discharge standards) and requires
improvement of the current treatment system prior to discharge as per the water code –
The next section evaluates the current health state of the Faleme river ecosystem to
determine any impact related to the past discharge of this substandard mine effluent.
7 Faleme River Ecosystem Assessment
7.1.
Information on the Faleme
The river originates in Guinea, approximately 250km upstream of the study area and
joins another large river, the Senegal river 270 km downstream of the mine, whereafter it
flows 590 km before entering the Atlantic Ocean as the Senegal River. The river is
managed by a Basin Management Authority (OMVS – Organisation for utilizing the Senegal
River) with the Governments of Guinea, Mali, Senegal and Mauritania having input. The river
is also a vital lifeline to communities living along it and is the only regional water supply in
the area during drought conditions. The Falémé River forms the Mali border with Senegal,
and has the westward flowing Garra River as a tributary; these two rivers converge
approximately 1.5 km north of the mine site. Other rivers in the Loulo region include the
Loulo Kaba River. The catchment areas for the Falémé River, Garra River and the Loulo
Kaba River are 17 100 km2, 561 km2 and 32 km2 respectively. (DWA, 2009).
Previous studies have analyzed the 41-year flow record in the Falémé River to
determine the Average, Minimum and Maximum monthly flow rates. There are two
catchments identified namely the Falémé and Gara catchments. By proportioning the flow
according to the catchment ratios, corresponding data was estimated for the Gara River as
summarised in Table 8.
Table 8. Summary of water quantity of the two catchments (Source: DWA, 2009)
Monthly Flow Rate (m3/s)
Month
Falémé Catchment
Sept
Annual
Gara Catchment
Sept
Annual
43
Min
69.1
20.7
2.2
0.7
Mean
515
110
16.2
3.5
Max
1 250
223
39.4
7.0
Note: September has the highest flow rates per annum.
The lowest annual flow for Falémé River was 20.7 m3/s compared to the Gara River
which is 0.7 m3/s. Gara River is a perennial river which is tributary to the Falémé River. The
surrounding villages of Djidian-Kenieba (DK), Loulo, Baboto and Sakola use the available
surface and groundwater (Falémé River and Gara Dam). Water collected by the villages is
used for personal drinking water, clothes washing, drinking water for cattle and crop
irrigation (Digby Wells, 2009).
7.2.
Methodologies
A study was carried out in July 2014 to evaluate the health of the Faleme river
ecosystem. The assessment consisted of chemical analysis of water (effluent and rievers
water), metal analysis of sediments and fish tissue and the assessment of the state of the
local fish and macroinvertebrates communities and habitat. Areas have been sampled also
in the Faleme at Gounkoto area (upstream of the study area) and at artisanal mining sites to
help in the interpretation of the results. Below are a table and a figure showing the different
sampling points.
Table 9. Sampling points and description (Modified from DWA, 2012)
Site
FL1
FL2
GR1
GR2
GK1
GK2
PC1
PC-UP
PC-DWN
Description
Falémé River downstream of Loulo
operation (weir)
Falémé River upstream of operation
In the lower reaches of the Garra Dam
adjacent to the Loulo camp
At the Garra River bridge downstream
of the local orpailleurs (artisanal
miners)
Falémé River downstream of Gounkoto
operation
Falémé River upstream of Gounkoto
operation
Surface settlers (referred to as Settling
ponds)
Immediately Upstream of the discharge
area into the Falémé River
Immediately Downstream of the
discharge area into the Falémé River
Water
Sediment
Fish Tissue
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
44
Figure 28. Arial map showing the sampling points (Source: DWA, 2012)
Figure 29. Photography of the sampling points at Faleme and Garra (Source: modified from DWA,
2012)
7.2.1. Habitat assessment
The habitat integrity of a river refers to the maintenance of a balanced composition of
physico-chemical and habitat characteristics on a temporal and spatial scale that are
comparable to the characteristics of natural habitats of the region (Kleynhans 1996).
Methodologies developed for the Rapid Bioassessment Protocols for use in Streams and
Wadeable Rivers (USEPA, 2006) for low gradient systems were primarily applied for the
assessment of the habitat. This was assessed and characterized according to section D of
the “Procedure for Rapid Determination of Resource Directed Measures for River
Ecosystems, 1999”. It should be noted that the Intermediate Habitat Assessment Integrity
(IHAI) was based on regions assessed in the current studies and therefore may only
constitute the assessment of conditions within a 50 km length of the potentially effected
water courses. The table below shows the scoring system.
45
Table 10. Description of IHAS scores with the respective percentage category (McMillan, 1998)
7.2.2. Water quality
Water quality analysis was completed for the project utilising a calibrated water
quality meter (EXTECH, DO700). In addition to this, laboratory analysis of water was
completed. The analysis involved the determination of metal content as well as the
concentrations of macronutrients such as nitrate, nitrite and phosphate. The analysis
involved the use of spectrophotometry applying techniques used in Inductively Coupled
Plasma Mass spectrophotometry (ICP-MS) as well as ICP- Optical Emission
Spectrophotometry (ICP-OES). The laboratory testing of water was be completed by SGS in
Bamako, Mali. The water quality data during the study was validated using the quality data
generated over 4 years for the Falémé river and then compared to the water quality
guidelines for aquatic ecosystems as described by the Department of Water Affairs for South
Africa (1996). Where guidelines were not prescribed for selected constituents, the guidelines
described for domestic use were considered since this is believed to be the top use of the
Faleme by the riverine communities.
7.2.3. Sediment
Sediment samples were collected using a plastic spoon where the top 2/3 cm of the
sediment surface was utilized. Sediment samples were then placed into plastic bottles and
analysed according to USEPA 2006 guidelines with ICP-MS. The following indices were
applied according to Pheiffer et al. (2014): the Enrichment Factor (Ef) which provides an
indication of enrichment of individual elements emanating from anthropogenic sources; the
Geo-accumulation index (Igeo) which provides an indication of pollution potential of specific
elements; the Pollution Load Index (PLI) which indicates the effect of a mixture of metals;
and the Sediment Quality Index. MacDonald et al. (2000a) determined the toxicity levels of
sediments based on a combination of a number of sediment quality guideline studies in the
United States (comprising a total of 17 data sets). Sediments can be placed into one of three
classes, based on the level of selected constituents in the sediments. These guidelines were
determined for Cadmium, Chromium, Copper, Lead, Mercury, Nickel and Zinc (MacDonald
et al., 2000a). In addition to categories described by (MacDonald et al., 2000a), sediment
contaminant concentrations obtained during the study were also evaluated against other
guidelines and authors, these references are shown in table below.
Table 11. Defined thresholds for metal concentrations in sediments (DWA, 2012)
46
Analyte
mg/kg dry weight (ppm)
EC
1
MEC
2
PEC
3
Reference
Antimony as Sb
Arsenic as As
2.0
.79
13.5
9.79-33
25.0
33
Long & Morgan, 1991
MacDonald et al., 2000a
Barium as Ba
Cadmium as Cd
100.0
.99
500.0
0.99-4.98
3000.0
4.98
I.P.C.S., 1990
MacDonald et al., 2000a
Chromium as Cr
3.4
43.4-111
111
MacDonald et al., 2000a
Cobalt as Co
Copper as Cu
9.73
1.6
13.83
31.6-149
37.1
49
Masoud et al. (2005)
MacDonald et al., 2000a
Lead as Pb
Manganese as Mn
5.8
460.0
35.8-128
780.0
128
1,100.0
MacDonald et al., 2000a
Persuad et al., 1993
Mercury as Hg
Nickel as Ni
0.18
22.7
0.18-1.06
22.7-48.6
1.06
48.6
MacDonald et al., 2000a
MacDonald et al., 2000a
Silver as Ag
1.6
1.9
2.2
MacDonald & MacFarlane, 1999
Zinc as Zn
121
121-459
459
MacDonald et al., 2000a
7.2.4. Aquatic benthic macroinvertebrates
Aquatic macroinvertebrate assemblages are good indicators of localised conditions
because many benthic macroinvertebrates have sedentary characteristics with relatively
long lives (±1 year) (Barbour et al., 1999). Macroinvertebrates are useful for their ability to
integrate pollution effects over time, their detectable response to environmental impacts as
well as the easy field sampling techniques involved in their collection. Benthic
macroinvertebrate assemblages are made up of species that constitute a broad range of
trophic levels and pollution tolerances, thus providing strong information for interpreting
cumulative effects (Barbour et al., 1999). The assessment and monitoring of benthic
macroinvertebrate communities forms an integral part of the monitoring of the health of an
aquatic ecosystem. The sampling protocols of the kick and sweep methodology of the South
African Scoring System (SASS, version 5) (Dickens and Graham, 2002) was used for the
current assessment. Identification of organisms was then made per family level and the
number of individuals was recorded (Thirion et al., 1995; Dickens and Graham, 2002; Gerber
and Gabriel, 2002). Both qualitative and quantitative assessments were undertaken. The
qualitative analysis followed the United States Environmental Protection Agency’s (USEPA)
Ephemeroptera (Mayflies), Plecoptera (Stoneflies) and Trichoptera (Caddisflies) (EPT) Taxa
Richness Metric. The EPT taxa are considered to be sensitive to pollution and therefore
provide information on the state and extent of pollution at a site (Barbour et al., 1996). The
results of this assessment provide baseline data for future rapid bio-assessment protocols.
The level of impairment at a site was then determined though the characterization of the
1
Threshold Effect Concentration (TEC)
2
Midpoint Effect Concentration (MEC)
3
Probable Effect Concentration (PEC)
47
EPT taxa, where a high EPT taxa richness would indicate no or low impairment levels and a
low EPT taxa richness indicating high impairment levels. The quantitative analysis included
the calculation of the following macroinvertebrate indices: the Margalef’s Measure of
Richness Index (1961), the Shannon-Wiener’s Diversity Index (1963), and the Pielou’s
Evenness Index (1986).
7.2.5. Fish
The use of fish to determine levels of ecological disturbance has many advantages
over other bio-assessment techniques (Zhou et al., 2008). This is because fish are long lived
and populations respond to environmental modification. They are continuously exposed to
aquatic conditions, are often migratory, and fulfil higher niches in the aquatic food web.
Therefore fish assemblages can provide an effective indication of the degree of modification
in the aquatic environment. A variety of techniques was used to sample the available fish
species within the project area. These sampling methods included cast nets, fyke nets, gill
nets and electro-fishing depending on site characteristics. Fish community structures and
diversity was determined at each site, this information is investigated as to determine
dominant species. The information and specific characteristics on dominant and present fish
species, in conjunction with macroinvertebrates data then allowed for the analysis of the
current state of the aquatic ecosystem. Muscle tissue samples were removed from the fish
species sampled from the Falémé River and frozen. In the laboratory, an inductive coupled
plasma mass spectrometry (ICP-MS) was used for metal screening for prepared whole body
tissues.
7.3.
Results
7.3.1. General Description of the Faleme
The Falémé River is located within the Senegal–Gambia freshwater ecoregion. The
river systems associated with this ecoregion experience “pronounced flooding during the wet
season” (Thieme, 2014). Due to the scale of the flooding within the ecoregion large
floodplains are created resulting in the creation of extensive feeding and breeding grounds
for ichthyofauna, with flooding conditions often being responsible for the triggering of
migrations. The terrestrial ecology associated with the drainage area of the Falémé River
consists of savannah/grasslands within the study focus area with wetter conditions in the
southern portion of the river catchment.
The Falémé River’s major habitat type, within the study focus area, is described best
as “a savannah river system within the middle reaches with a moderate gradient, resulting in
large colluvial deposits, moderate entrenchment, scoured and spaced pools with a gentle
meandering nature” (Rosgen, 1996). The sinuosity is approximately >1.2 producing stable
banks, however, within the study focus area, dolerite dykes are fairly common producing
bedrock riffles, large scour pools and at times diverting the path of the river. The general
sinuosity and common dykes are illustrated in the figure below (Figure 30).
48
Large cobble beds often associated with dolerite dykes are also found throughout the
study area (Figure 31). These cobble beds are usually dry during the low flow season and
inundated during periods of flooding. These cobble beds are likely fish spawning areas and
act as high aquatic biodiversity habitats, for this reason they are considered important in the
Falémé River system. Based on the importance of these cobble beds for aquatic biodiversity
and potential fish spawning grounds a critical habitat assessment would be included in this
study. Typical riparian habitat is illustrated in the images below (Figure 32 and Figure 33).
The longitudinal profile of the Falémé River is provided in the figure below (Figure 34). It
should be noted that this profile is a rough representation of the gradient associated with the
sites selected and is based on the height recorded at each sampling site considered in this
study. The profile will be used for basic reference purposes only.
Figure 30. Aerial Image of the Falémé River, depicting meandering nature and alluvial
deposition (Google Earth)
49
Figure 31. Cobble beds in the Falémé River. A: Natural cobble bed; B: Modified cobble bed;
C: Modified cobble bed.
Figure 32. Image of intact riparian habitat (Farandie).
50
Figure 33. Image of riparian habitat (Cobble Bed)
51
Figure 34. Longitudinal profile of the Falémé River (Google Earth™)
As seen in the above figure the gradient is relatively gentle and therefore geomorphological
characteristics should remain relatively constant between sites.
7.3.2. Habitat integrity
The Faleme river habitat was found to be moderately modified or class C with the
instream habitat being scoring less than the riparian habitat (62.36 and 76.84 respectively);
and the Gara river habiat was found to be largely modified or class D with the instream
habitat scoring also less than the riparian one (45.24 and 52.6 respectively). The results are
shown in the tables below.
Table 12. IHIA for instream Habitat within the Faleme River (Source: Tate, 2014)
Instream
Average score
Score
Water abstraction
8
4.48
Flow modification
10
5.2
Bed modification
15
7.8
Channel modification
15
7.8
Water quality
10
5.6
Inundation
5
2
Exotic macrophytes
3
1.08
Exotic fauna
4
1.28
Solid waste disposal
10
2.4
Total Instream
37.64
Category
62.36
Table 13. IHIA for Riparian Habitat within the Faleme River (Source: Tate, 2014)
Riparian
Average score
Score
Indigenous vegetation removal
10
5.2
Exotic vegetation encroachment
5
2.4
Bank erosion
3
1.68
Channel modification
2
0.96
Water abstraction
3
1.56
Inundation
5
2.2
Flow modification
5
2.4
Water quality
13
6.76
Total Riparian
23.16
Category
76.84
52
Table 14. IHIA for instream Habitat within the Gara River (Source: Tate, 2014)
Instream
Average score
Score
Water abstraction
10
5.6
Flow modification
18
9.36
Bed modification
18
9.36
Channel modification
18
9.36
Water quality
18
10.08
Inundation
10
4
Exotic macrophytes
5
1.8
Exotic fauna
5
1.6
Solid waste disposal
15
3.6
Total Instream
54.76
Category
45.24
Table 15. IHIA for Riparian Habitat within the Gara River (Source: Tate, 2014)
Riparian
Average score
Score
Indigenous vegetation removal
15
7.8
Exotic vegetation encroachment
5
2.4
Bank erosion
10
5.6
Channel modification
10
4.8
Water abstraction
10
5.2
Inundation
15
6.6
Flow modification
15
7.2
Water quality
15
7.8
Total Riparian
47.4
Category
52.6
7.3.3. Water quality
The results of the in situ water quality analysis are presented in the table below
(Table 16). Water quality guidelines utilized were the DWAF aquatic ecosystems guidelines.
The rationality for this is that these guidelines are designed specifically for aquatic biota.
Current available guidelines for East, Central and West Africa do not take into consideration
sensitive aquatic biota and are based on human needs.
Table 16. Water quality results obtained during the 2014 July survey
53
Constituent
pH
Temperature (ºC)
Conductivity
(µS/cm)
DO (mg/L)
DO saturation
(%)
Clarity (cm)
Guideline/Site
6.5-9
10-35
<700
>5
60-120
N/A
GK Upstream
7.3
30
84.0
4.9
70
41
GK Midstream
7.4
31
87.0
4.8
67
47
GK Downstream
7.3
31
93.0
5.2
71
13
Border Bridge
7.5
32
115
3.5
45
22
Downstream
Sasanba
7.4
29
73.0
5.4
71
6.0
Cobble Bed
7.4
31
93.0
6.2
85
17
Farandie
7.6
31
100
2.4
31
37
7.8
33
111.0
6.9
80
19
8
31
1332
6.5
87
2.0
7.4
29
140.0
5.4
71
18
Faleme Weir
7.5
32
103.0
7.2
105
28
Artisanal Tributary
7.5
29
37.0
4.4
71
6.0
DK Bridge
7.6
33
202.0
3.1
40
30
Gara Dam
7.8
33
182.0
7.3
87
25
Gara Settling Pond
7.8
33
1950
4.5
59
N/A
Upstream
Discharge
Gara
Gara
Discharge
Settling
Downstream
Discharge
Gara
*Red shading denotes values not within recommended guidelines.
Based on results observed in the above table (Table 16), the pH values of the
Faleme River range from 7.3 at upstream of gounkoto to 7.8 at the site upstream of the Gara
settling pond effluent. The range of pH in the Gara River was found to be relatively small,
from 7.5 at the Artisanal Tributary to 7.8 in the Gara Dam. The pH of the Gara settling ponds
was found to be basic and was at 8. Temperatures were found to be relatively constant
throughout the study and were predominantly measured at approximately to 30 ºC.
Conductivity was found to be relatively low in the Faleme River and ranged from 73 µS/cm at
the site downstream of the artisanal mining village Sasanba to 115 µS/cm at the Border
Bridge site. Conductivity in the Gara River was also found to be relatively low and ranged
from 37 µS/cm at the Artisanal Tributary to 202 µS/cm at the DK bridge site. Concentrations
of Dissolved Oxygen (DO) in the Faleme River ranged from 2.4 mg/l to 7.2 at the Faleme
Weir site. DO levels in the Gara River were found to range from 3.1 at DK Bridge to 7.3 in
the Gara Dam.
The results of the chemical analysis of samples from the sites considered in this
study are presented in Table 17. Total phosphorus concentrations were found to be
exceeding guideline values at several sites including Faleme downstream of Gounkoto,
Faleme downstream of artisanal mining areas, Faleme downstream of gara settling pond,
and the Gara settling pond. Nitrite values were found to be exceeding recommended
guideline values in the Gara effluent. The concentrations of Nitrate were found to be
54
elevated beyond guideline concentrations at Gara settling discharge. Ammonia levels at
sites Faleme upstream of Gounkoto, Gara settling discharge, Faleme Weir, Gara Bridge,
and the Gara settling pond were found to be elevated. Dissolved arsenic (As) concentrations
were found to be exceeding guideline values at Faleme downstream of artisanal mining
areas and slightly at the gara effluent settling ponds. Concentrations of Copper (Cu) were
found to be below recommended guideline values. Chromium concentrations were found to
be above guideline values at Faleme downstream of artisanal mining areas and the Gara
settling pond. Concentrations of Cobalt (Co), Cadmium (Cd) and Mercury (Hg) were found to
be below guideline values at all sites. However the mercury values were noteworthy at
Faleme close to artisanal mining areas. Levels of dissolved Lead (Pb) were found to be
above guideline concentrations at Faleme close to artisanal mining areas. The following
sites had elevated concentrations of Manganese (Mn) above recommended guideline
values; Faleme at Gounkoto, Faleme downstream of artisanal mining areas and the Gara
settling pond. Concentrations of Zinc (Zn) were found to be above recommended guideline
values at sites upstream of Gounkoto, Faleme at artisanal mining areas as well as the Gara
settling ponds. The concentrations of dissolved Aluminium (Al) were found to be exceeding
recommended guideline values at all sites with the exception of the Faleme Weir.
-1
Table 17. Chemical analysis of water from the sites considered in the 2014 survey (mg.l )
Guideline
69
<
200
<6*
<6*
<0.
2
<0.0
1
<0.1
<0.01
2
N/A
<0.15
<0.006
<0.01
<0.1
8
<0.00
2
Description
pH*
SO4
NO2
NO3
NH3
As
Cu
Cr
Co
Cd
Hg
Pb
Mn
Zn
Upstream GK
7.
3
1
<0.0
5
<0.0
6
0.2
2
0.00
1
0.00
1
<0.00
1
<0.00
1
<0.000
1
<0.000
1
<0.000
5
0.03
1
0.021
Midstream GK
7.
4
2
<0.0
5
0.52
0.1
8
0.00
1
0.00
1
<0.00
1
<0.00
1
<0.000
1
<0.000
1
<0.000
5
0.02
0.009
Downstream GK
7.
3
5
<0.0
5
0.22
0.1
9
0.00
2
0.00
6
0.004
<0.00
1
<0.000
1
<0.000
1
0.0008
0.04
5
0.009
Sasanba DS
7
9
0.11
3.07
0.1
5
0.24
0.05
4
0.089
0.058
0.0004
0.003
0.1
1.42
0.083
Farandie
7.
6
3
<0.0
5
0.6
0.1
2
0.00
4
0.00
1
<0.00
1
<0.00
1
<0.000
1
<0.000
1
<0.000
5
0.01
6
0.006
Us settling
7.
3
4
0.06
0.92
0.2
0.00
7
0.00
4
0.001
0.003
<0.000
1
<0.000
1
0.0005
0.05
6
0.006
7.
7
94
10.2
96.5
3.5
9
0.01
7
0.00
4
0.008
0.001
<0.000
1
<0.000
1
<0.000
5
0.01
1
0.007
Ds settling
7.
2
4
0.08
0.89
0.1
4
0.01
2
0.00
6
0.004
0.002
<0.000
1
<0.000
1
0.0012
0.08
5
0.008
Faleme Weir
7.
7
4
<0.0
5
0.65
0.4
9
0.00
3
0.00
3
<0.00
1
<0.00
1
<0.000
1
<0.000
1
<0.000
5
0.01
0.005
Artisanal Area
6.
7
27.
0.26
0.7
0.1
0.02
0.01
2
0.013
0.007
0.0007
0.004
0.0043
0.18
0.021
Artisanal tributary
6.
4
8.
0.07
2.96
0.1
0.01
1
0.00
9
0.014
0.004
<0.000
1
<0.000
1
0.0042
0.07
5
0.028
DK bridge
7.
3
4.
<0.0
5
<0.0
6
0.4
1
0.00
6
0.00
4
0.005
0.002
<0.000
1
<0.000
1
0.0014
0.07
8
0.011
Gara dam
7.
6
4.
<0.0
5
<0.0
6
0.2
2
0.00
4
0.00
1
0.002
<0.00
1
<0.000
1
<0.000
1
0.0005
0.09
2
0.005
Gara settling pond
7.
6
198.
8.9
180.
22.
7
0.04
3
0.01
7
0.025
0.007
<0.000
1
<0.000
1
0.0019
0.2
0.038
Gara
discharge
settling
55
69
Guideline
<
200
<6*
<6*
<0.
2
<0.0
1
<0.1
<0.01
2
N/A
<0.15
<0.006
<0.1
8
<0.01
<0.00
2
* Red Shading denotes constituent exceeding recommend guideline, Yellow shading denotes elevated constituent
7.3.4. Sediment quality
Sediment samples analyses results are shown below in Table 18.
The majority of elements analysed were found to be below the Threshold Effect
Concentrations (TEC). Elements found to be exceeding the Probable Effect Concentrations
(PEC) were arsenic (As), chromium (Cr) and mercury (Hg). Levels of As were found to be
exceeding the Midpoint Effect Concentrations (MEC) at sites Faleme Downstream of
Gounkoto, Faleme & gara at artisanal mining areas, and Gara Bridge. Concentrations of As
were found to be exceeding PEC concentrations at Faleme Midstream of Gounkoto, Faleme
at artisanal mining areas, Faleme at Gara Settling Discharge, and Gara underground
Settling pond. Levels of As were found to be excessively high (above 100 mg.kg -1) at
Faleme Midstream of Gounkoto, and Faleme at artisanal mining areas.
-1
Table 18. Chemical analysis of sediment samples collected during the July 2014 survey (mg.kg )
TEC
9.79
31.6
43.4
9.73
0.99
35.8
460
121
N/A
N/A
0.18
MEC
9.7933
31.6149
43.4111
11.8
3
0.994.98
35.8128
780
121459
N/A
N/A
0.181.06
PEC
>33
>149
>111
>37.
1
>4.98
>128
>11
00
>459
N/A
N/A
>1.06
Description
As
Cu
Cr
Co
Cd
Pb
Mn
Zn
Al
Fe
Hg
UPSTREAM GK
4
8.4
37
5
<0.3
4
120
11
430
0
1400
0
<0.06
MIDSTREAM GK
410
62
97
8.5
0.4
8
210
24
290
0
3500
0
0.1
DOWNSTREAM GK
16
8.8
33
3.8
<0.3
3
110
10
310
0
1300
0
<0.06
SASAMBA
STREAM
260
22
38
11
<0.3
10
270
25
430
0
2400
0
1
DS SASAMBA
320
30
89
19
0.4
15
820
19
460
0
5300
0
0.5
FARANDIE
6
11
55
4.2
<0.3
8
54
15
450
0
1400
0
<0.06
US SETTLING
67
18
42
7.4
<0.3
6
250
25
460
0
2000
0
0.2
GARA
SETTLING
DISCHARGE
74
22
35
6.9
<0.3
3
170
38
570
0
1700
0
<0.06
DS SETTLING
54
21
54
7.6
<0.3
5
210
20
470
0
2100
0
0.2
WEIR FALEME
6
14
240
9.2
<0.3
9
140
16
410
0
5400
0
<0.06
ARTISINAL AREA
28
10
23
3.7
<0.3
3
77
7.1
280
0
1300
0
7.5
ARTISINAL Tributary
8
6.6
14
2.4
<0.3
3
49
10
300
8600
0.1
ARTISINAL
56
0
DK BRIDGE
7
6.2
19
3.4
<0.3
7
180
9.7
270
0
1100
0
<0.06
GARA DAM
12.
9.7
21.
4.2
<0.3
4.
130.
13.
390
0.
1200
0.
<0.06
GARA UNDERGROUND
42.
20.
16.
5.2
<0.3
2.
160.
31.
390
0.
1000
0.
<0.06
7.3.5. Aquatic Benthic microinvertebrates
The results of the IHAS completed at the various sites are presented in the table
below (Table 19).
Table 19. Integrated Habitat Assessment System (IHAS) results for the 2014 survey (Source: Tate,
2014).
Site
Midstream GK
Border Bridge
Farandie
Faleme Weir
DK Bridge
Flow
Moderate
Fast
Slow
Fast
Slow
Clarity (cm)
47
22
37
28
30
Score
64
67
63
69
50
Suitability
Fair
Good
Fair
Good
Poor
A variety of macroinvertebrates were sampled throughout the 2014 survey. The
results of the macroinvertebrate survey are presented in Table 20. It should be noted that
not all survey sites were included in the invertebrate survey. Only sites with suitable and
available “sampleable” invertebrate habitat were selected. The sensitivity ratings are based
on the South African Scoring System (Dickens and Graham, 2002). The Average Score Per
Taxon (ASPT) is also calculated based on the above biotic index.
Table 20. Macroinvertebrates sampled during the July 2014 survey (Source: Tate, 2014).
Family
Sensitivity
Aeshnidae
8
Baetidae 1sp.
4
Baetidae 2 sp
6
Baetidae >2 sp
8
Caenidae
6
Ceratopogonidae
5
Midstream GK
Border
Bridge
Farandie
DK
bridge
Faleme
Weir
1
3
2
3
17
4
1
1
4
3
4
57
Family
Sensitivity
Midstream GK
Border
Bridge
Farandie
DK
bridge
Faleme
Weir
Chironomidae
2
12
14
22
9
19
Coenagrionidae
4
7
9
23
Corbiculidae
5
1
1
Dytiscidae
5
Elmidae
8
Gerridae
5
4
Gomphidae
6
3
Gyrinidae
5
2
Heptageniidae
13
3
Hydracarina
8
21
Hydropsychidae 2 sp
8
Hydropsychidae >2 sp
12
Hydroptilidae
6
Leptoceridae
6
Leptophlebiidae
9
Libellulidae
4
Oligochaeta
3
Oligoneuridae
15
Polymitarcyidae
10
Potamonautidae
3
Protoneuridae
8
Psychomyiidae
8
Simuliidae
5
4
Tabanidae
5
1
Teloganodidae
12
Thiaridae
3
Veliidae
5
Total Taxa
1
2
1
8
1
1
1
1
1
5
15
10
23
16
7
2
22
2
4
4
6
4
4
4
3
2
4
10
9
8
10
2
6
1
1
8
1
4
23
4
1
7
3
12
14
17
10
17
58
Midstream GK
Border
Bridge
Farandie
DK
bridge
Faleme
Weir
Total individuals
65
80
81
54
167
Total Sensitivity Score
70.0
92.0
112
48.0
105
ASPT
5.8
6.5
6.5
4.8
6.1
EPT
2
4
8
5
1
Family
Sensitivity
Total macroinvertebrate taxa recorded at the various sites ranged from 10 at the DK
Bridge to 17 at Farandie and Faleme Weir. The overall SASS5 scores for the sites ranged
from 54 at DK Bridge to 112 at Farandie. The ASPT values ranged from 4.8 at DK Bridge to
6.5 at Border Bridge and Farandie.
The percentage contribution of Ephemeroptera, Plecoptera and Trichoptera (% EPT)
at the sites is expressed in Figure 35. Results for the Margalef’s diversity index are
presented in Figure 36. Shannon’s Diversity index results are presented in Figure 37 with
Pielou’s Evenness index represented in Figure 38.
Figure 35. Percentage contribution of EPT at the sites assessed in the 2014 survey (Tate, 2014).
59
Figure 36. Margalef’s Diversity Index results for the July 2014 survey (Tate, 2014).
Figure 37. Results of Shannon’s Diversity Index during the July 2014 surveys (Tate, 2014).
60
Figure 38. Results of Pielou’s Evenness Index during the July 2014 surveys. (Tate, 2014).
7.3.6. Ichthyofauna
The results are summarized in Table 21 below where the various family groups of
fish sampled during the 2014 is presented.
33 species of fish were collected representing 12 families, of which 24 fish were
identified to species level and the remaining 9 identified to genus level. The different species
were collected using various techniques, and as would be expected, were found in their
associated meso-habitats. Figure 39 and Figure 40 show examples of fish sampled during
the study.
Table 21. Families recorded during the 2014 (Source: Tate, 2014).
Family
No. of species
No. of individuals
Alestidae
7
68
Cichlidae
5
166
Cyprinidae
6
145
Distichodontidae
1
8
Claroteidae
4
44
Clariidae
1
1
Gobiidae
1
4
Malapteruridae
1
1
Mochokidae
3
4
61
Family
No. of species
No. of individuals
Mormyridae
2
11
Poeciliidae
1
4
Schilbeidae
1
17
Total
33
473
Figure 39. Photographs of fishes from the Alestid family (Source: Tate, 2014)
62
Figure 40. Photographs of fishes from the Cichlid family (Source: Tate, 2014)
7.3.7. Bioaccumulation
The results of the bioaccumulation of metals within T. zillii samples are presented in
the figures below (Figure 41, Figure 42, and Figure 43).
63
64
Figure 41. Bioaccumulation of metals in the Faleme River (Source: Tate, 2014)
Figure 42. Bioaccumulation of metals in the Gara River (Source: Tate, 2014)
65
Figure 43. Comparison of metal bioaccumulation between the Gara and Faleme River (Source: Tate,
2014)
66
8 Gara Underground Effluent treatment
Two types of pollutants occurred in the Gara discharge as described in the risk
characterization – underground effluent section: Total suspended solids (TSS) and High
nitrate (100 – 400 mg/l) with Sulfate level gradually increasing. The pollutants together
present a complex treatment system. It was therefore necessary to isolate each pollutant
and determine its source, transfer pathway and devise solutions accordingly with associated
risks.
8.1.
Proposed Solutions
8.1.2. Source of pollutants
Specialist input was seeked to describe the contributing factors to the effluent quality
issue. Below is a summary of the findings from Digby wells in 2013 assessment.
Total Suspended Solids (TSS)

Poor settling of water in the shallow underground sumps and settling dams (see
Figure 44) prior to pumping to surface resulting in the water containing high
content of solids;

Poor housekeeping in terms of solids and water separation. The dams and
sumps do not have a bund wall thus solids easily enter the dams;

Due to the high volume, there is a high flow and this adds to mixing of water
particularly in the last settling dam before pumping water to surface;

The lack of a thickener exacerbates the issue. There is definitely more solids
content and the settling dams are silted. There is also poor housekeeping at Gara
where the solids are washed on the surface and have formed a layer at least 0.2
m thick on the surface. The discharge washes away this silt straight into the river;

The use of the settling dams with inlet and outlet serves no purpose in trying to
achieve a good sedimentation.
67
Figure 44. Surface settling dams showing large amount of suspended solids
Heavy Metals

Geology (released with solids and pyrite oxidation)

Hydrocarbon spillage
Nitrate Source

No major agricultural activity where large amount of fertilizers/ Herbicides and
pesticides are used;

No major sewage source exists;

Natural groundwater is low in nitrates;

Nitrate based explosives seem to be the only real source.
8.1.3. Recommended Solutions
A number of expert advice has been seeked in the past as to the most effective way
of treating the effluent prior to discharge. The last two consultations were in January 2011
when Digby Wells Environmental undertook a site visit to evaluate the issues and to
recommend measures. Proposed measures were to settle the solids in a thickener and
construct a wetland/reedbeds to deal with the nutrient rich – effluent. Then sulphate was not
identified to be a problem like it appears to be today. The International Network for Acid
Prevention (INAP) in 2003 published a report on the treatment of sulphate in mine effluents
(“Lorax report”) where it was shown that out of the existing methods, biological treatment
seems to have the greatest potential for sulphate removal at a low cost – however it was
pointed out in the same report that more research and development is required to design a
wetland that can achieve expected results as it seems the least effective among the
biological methods- energy and carbon are key drivers of the cost. Another visit undertaken
in October 2013 recommended a more elaborated action plan as stated below.

Use of reeds/ plants, lignocellulose, sewage sludge, rock and soil in a wetland to
remove. use gabions4 (as indicated in Figure 45 and Figure 46)
–
Residual sulphate;
–
Residual nitrate;
–
Heavy Metal;
4
gabion (from Italian gabbione meaning "big cage"; from Italian gabbia and Latin cavea
meaning "cage") is a cage, cylinder, or box filled with rocks, concrete, or sometimes sand and soil for
use in civil engineering, road building, and military applications (Source: Wikipedia)
68
Figure 45. Gabion system conceptual design (DWA, 2013)
Figure 46. Gabions (left) and gabions meshed cages (right) (DWA, 2013)
The following sections below cover the design of the constrcuted wetalnd together
with that of the pre-treatment system to remove the suspended solids.
8.2.
Constructed Wetland – conceptual design
Digby Wells Environmental has been commissioned to propose a design based on
the available information (effluent data, site climate and topography and discharge quality
requirement). Part of Information from the report has been used to propose design for the
purpose of this study.
8.2.1. Definition of wetland
Hammer (1994) defines a constructed wetland (CW) as man-made systems
designed, built and operated to emulate functions of natural wetlands for the removal of
pollutants from wastewater in a more controlled environment. Understanding and designing
a constructed wetland require the involvement of expertise in a variety of fields, including
chemistry, hydrology, soil science, plant biology, natural resources, environmental
management, ecology, environmental engineering, surveying and project management.
8.2.2. Classification of wetlands
Constructed wetlands are classified into two groups (Kayombo et al, 2004; Wetland
International, 2003): horizontal flow system (HFS) and the vertical flow system (VFS). The
horizontal flow systems are classified into two types: the surface flow (SF) and the sub69
surface flow (SSF) systems. In the horizontal flow systems, water is fed through an inlet and
flows horizontally through the bed to the outlet and in the vertical flow systems, water is fed
intermittently and drains vertically through the bed via a network of drainage pipes. Below
are figures showing examples of a surface flow and a sub-surface flow systems.
Figure 47. Free Surface flow (FSW) constructed wetland – section view (Source: Wetland
International, 2003)
Figure 48. Sub-Surface flow (SSF) constructed wetland (Source: Wetland International, 2003)
8.2.3. History of wetlands
The editorial of the ecological engineering journal (issue no. 25) in 2005 (p. 475-477)
describes the history of wetlands development around the world.
It states that the first attempts to use the wetland vegetation to remove various
pollutants from water were conducted by K. Seidel in Germany in early 1950s. The first fullscale free water surface (FWS, surface flow) CW was built in The Netherlands to treat
wastewaters from a camping site during the period 1967–1969. Within several years, there
were about 20 FWS CWs built in The Netherlands. However, FWS CWs did not spread
throughout the Europe but constructed wetlands with horizontal sub-surface flow (HF CWs)
became the dominant type of CWs in Europe. The first full-scale HF CW was built in 1974 in
Othfresen in Germany. The early HF CWs in Germany and Denmark used predominantly
heavy soils, often with high content of clay. These systems had a very high treatment effect
70
but because of low hydraulic permeability, clogging occurred shortly and the systems
resembled more or less FWS systems. In late 1980s in the United Kingdom, soil was
replaced with coarse materials (washed gravel) and this set-up has been successfully used
since then. In the 1980s, treatment technology of constructed wetlands rapidly spread
around the world.
In 1990s, increased demand of nitrogen removal from wastewaters led to more
frequent use of vertical flow (VF) CWs which provide higher degree of filtration bed
oxygenation and consequent removal of ammonia via nitrification. In late 1990s, the inability
to produce simultaneously nitrification and denitrification in a single HF or VF CWs and thus
remove total nitrogen lead to the use of hybrid systems which combine various types of
CWs. The concept of combination of various types of filtration beds was actually suggested
by Seidel in Germany in the 1960s but only a few fullscale systems were built (e.g. Saint
Bohaire in France or Oaklands Park in UK) in 1980s and early 1990s. At present, hybrid
CWs are commonly used throughout Europe as well as other parts of the world. VF–HF
combination is the dominant set-up but HF–VF combination is also used and FWS CWs are
commonly used in hybrid systems. In 1970s and 1980s, constructed wetlands were nearly
exclusively built to treat domestic or municipal sewage. Since 1990s, the constructed
wetlands have been used for all kinds of wastewater including landfill leachate, runoff (e.g.
urban, highway, airport and agricultural), food processing (e.g. winery, cheese and milk
production), industrial (e.g. chemicals, paper mill and oil refineries), agriculture farms, mine
drainage or sludge dewatering. they offer a cheaper, low raw material, sustainable and
energy-efficient alternative technology to wastewater treatment. They cater for secondary
and tertiary treatment of wastewaters. Wetlands have many functions: creation of natural
habitat, water quality improvement, flood control and the production of food and fiber
(production of biomass). They are now well-established methods for wastewater treatment in
tropical climate (Kayombo et al, 2004; Wetland International, 2003).
8.2.4. Conceptualization of wetland
Morgan (2014) described the steps for the design of a wetland.
Conceptualisation begins with the theoretical understanding of the entire wetland
system, followed by data collection and the refinement of that understanding.
Characterization of natural (ambient background or baseline) conditions involves the
following tasks:

Locating, collecting, and organizing basic types of data from available published
and unpublished sources; and

Conducting specifically designed field, laboratory, and modelling studies for the
sites selected.
71
It is imperative to develop a broad overview of the site and identifying those issues
that may assist or hamper the overall delivery of the wetland objective. This would
specifically respond to the site conditions. Careful assessment and interpretation of the site
conditions is a fundamental part of designing and development of the constructed wetland
that can be effective. There are several key characteristics of a site that need to be
understood as these can influence the level of confidence in providing an effective design.
For simplistic purposes these should include the following factors; i.e. climate, soils, average
slope, depth to groundwater. An overall water management plan should provide;

Site plan showing location, size and dimensions of the wetland and associated
measures, and

Conceptual design calculations to establish the quantitative estimates.
Effective environmental planning often demands qualitative and quantitative predictions of
the effect of future management activities. The conceptual design and appropriate models
can be applied to solve a wide range of wetland related problems under very different
situations. This study will also identify the most important variables in the development of the
constructed wetland, since the combination of flow, transport and processes in models is
often only possible when simple solutions for each problem are applied to keep the
mathematical complexity of the model low.
Development of an appropriate conceptual design (specifically the macrophyte zone)
of water flow and pollution transport within the constructed wetland is critical for developing
adequate predictive modelling methods and designing cost effective remediation techniques.
These processes must still take into account the changes under episodic natural climate.
The complexity resulting from the effects of episodic infiltration and preferential flow on a
field scale must be taken into account when predicting flow and transport and developing the
constructed wetland. The change of water flow and contaminant transport, which is difficult
to detect, poses unique and difficult problems for characterization, monitoring, modelling,
engineering of the constructed wetland, and remediation of contaminants. Lack of
understanding in this area has led to severely erroneous predictions of contaminant
transport and incorrect remediation actions. Therefore, it is imperative to develop a strategy
to investigate the macrophyte zone including a comprehensive plan to assess these
conditions specifically.
Kayombo et al, 2004, published a design manual for constructed wetlands and waste
stabilization ponds. Most of the following paragraphs refer to their work together with the
Wetland International’s 2003 publication on the use of constructed wetlands in wastewater
treatment. Focus is placed on the sub-surface flow systems (Figure 48) as they appear to
outperform the free surface ones (Hammer, 1989).
8.2.5. Pollutants Removal Mechanism
72
Wetlands have been found to be effective in treating BOD, TSS, N and P as well as
for reducing metals, organic pollutants and pathogens. The principal pollutant removal
mechanisms in constructed wetlands include biological processes such as microbial
(bacteria, fungi, algae and protozoa) metabolic activity and plant uptake as well as physicochemical processes such as sedimentation, adsorption and precipitation at the watersediment, root-sediment and plant-water interfaces (Reddy and DeBusk, 1987). A pollutant
may be removed as a result of more than one process at work. Since Nitrate and heavy
metals appear to be our main pollutants- focus has been placed on their removal
mechanism.
Nitrate removal mechanism
Before describing the processes at work for Nitrogen removal, it is important to
remind the nitrogen cycle. Below extract was taken from Bosman, 2009.
The Nitrogen Cycle refers to the inter-conversion between nitrogen (N), nitrite (NO2),
nitrate (NO3), ammonia (NH3), and ammonium (NH4+) in the environment. A simplified
illustration of the natural (not altered by anthropogenic activity) nitrogen cycle is outlined in
Figure 1 below:
Figure 49. Nitrogen cycle (Source:
http://users.rcn.com/jkimball.ma.ultranet/BiologyPages/N/NitrogenCycle.html )
Nitrate (NO3-) is the end product of the oxidation of elemental nitrogen (N),
ammonium (NH4+) and/or nitrite (NO2-) and is measured either as the salt, NO3-, or as the
amount of Nitrate-Nitrogen (N). Natural soil is generally Nitrate-rich. Nitrates and Nitrites
occur together in soil from natural erosion of geological Nitrogen. In normal geological
processes such as erosion, desertification and soil formation, nitrate release occurs slowly
over long periods of time, allowing the release of low levels of nitrogen that is essential for
73
the formation of fertile soil. Metasedimentary and metavolcanic lithologies, such as
greenstone and slate, typically contain high levels of elemental Nitrogen which could be
released as nitrate in this manner. The Nitrogen Cycle thus entails the uptake of atmospheric
or soil Nitrogen by plants or animals, and the conversion and use there-of by plants and
animals as essential building blocks for amino-acids and genetic material. As a result of
animal excretion and plant decay, ammonia (NH3) or ammonium (NH4+) is formed, which
readily oxidises to nitrite (NO2-) and nitrate (NO3-) under aerobic (presence of oxygen)
conditions (process of nitrification). Denitrifying bacteria can convert nitrate back to
ammonium, ammonia, or atmospheric Nitrogen under reducing circumstances (process of
de-nitrification).
The Nitrogen cycle is modified by anthropogenic activities such as the introduction of
oxidizing or reducing circumstances or chemicals, or by the large scale disturbance of
Nitrogen-rich geological formations. Ammonium (NH4+) will convert to nitrite (NO2), and
nitrate (NO3-) under oxidising conditions, such as aeration or excavation, while nitrate (NO3)
will convert to ammonium under reducing circumstances, or in the presence of a reducing
agent, such as acids. Large scale anthropogenic disturbances to the natural Nitrogen Cycle
are a serious cause of concern, since this can cause nitrate to be released in large quantities
into water resources, which has significant detrimental ecological and human health effects.
There are sufficient studies to indicate some roles being played by wetland in
Nitrogen removal but the significance of plant uptake vis-à-vis nitrification/denitrification is
still being questioned (Wetland International, 2003). Nitrogen (N), as per the cycle described
above, can exist in various forms, namely Ammoniacal Nitrogen (NH3 and NH4+), organic
Nitrogen and oxidised Nitrogen (NO2- and NO3-). The removal of Nitrogen is achieved
through nitrification/denitrification, volatilisation of Ammonia (NH3) storage in detritus and
sediment, and uptake by wetland plants and storage in plant biomass (Brix, 1993). A
majority of Nitrogen removal occurs through either plant uptake or denitrification. Nitrogen
uptake is significant if plants are harvested and biomass is removed from the system.
At the root-soil interface, atmospheric oxygen diffuses into the rhizosphere through
the leaves, stems, rhizomes and roots of the wetland plants thus creating an aerobic layer
similar to those that exists in the media-water or media-air interface. Nitrogen transformation
takes place in the oxidised and reduced layers of media, the root-media interface and the
below ground portion of the emergent plants. Ammonification takes place where Organic N
is mineralised to NH4+-N in both oxidised and reduced layers. The oxidised layer and the
submerged portions of plants are important sites for nitrification in which Ammoniacal
Nitrogen (AN) is converted to nitrite N (NO2-N) by the Nitrosomonas bacteria and eventually
to nitrate N (NO3-N) by the Nitrobacter bacteria which is either taken up by the plants or
diffuses into the reduced zone where it is converted to N2 and N2O by the denitrification
process . The first reaction in the nitrification process produces hydroxonium ions (acid pH),
which react with natural carbonate to decrease the alkalinity (Mitchell, 1996a). In order to
74
perform nitrification, the nitrosomonas must compete with heterotrophic bacteria for oxygen.
The BOD of the water must be less than 20 mg/l before significant nitrification can occur
(Reed et al., 1995). Temperatures and water retention times also may affect the rate of
nitrification in the wetland. Denitrification is the process in which nitrate is reduced in
anaerobic conditions by the benthos to a gaseous form. The reaction is catalyzed by the
denitrifying bacteria Pseudomonas spp. and other bacteria.
Denitrification requires nitrate, anoxic conditions and carbon sources (readily
biodegradable) (Kayombo et al, 2004). Nitrification must precede denitrification, since nitrate
is one of the prerequisites. The process of denitrification is slower under acidic condition
(Kayombo et al, 2004). At a pH between 5-6, N20 is produced. For a pH below 5, N2 is the
main nitrogenous product (Nuttall et al., 1995). NH+4 is the dominant form of ammonianitrogen at a pH of 7, while NH3 (present as a dissolved gas) predominates at a pH of 12.
Nitrogen cycling within, and removal from, the wetlands generally involves both the
translocation and transformation of nitrogen in the wetlands, including sedimentation
(resuspension), diffusion of the dissolved form, litter fall, adsorption/desorption of soluble
nitrogen to soil particles, organism migration, assimilation by wetland biota, seed release,
ammonification (mineralisation) (Orga-N – NH+4), ammonia volatilization (NH+4 – NH3
(gas)), bacterially-mediated nitrification/denitrification reactions, nitrogen fixation (N2, N2O
(gases – organic-N)), and nitrogen assimilation by wetland biota (NH+4, Nox organic – N,
with NOx usually as NO-3). Precipitation is not a significant process due to the high solubility
of nitrogen, even in inorganic form. Organic nitrogen comprises a significant fraction of
wetland biota, detritus, soils, sediments and dissolved solids (Kadlec and Knight, 1996).
Denitrification is the permanent removal of Nitrogen from the system, however the
process is limited by a number of factors, such as temperature, pH, redox potential, carbon
availability and nitrate availability (Johnston, 1991). The annual denitrification rate of a
wetland could be determined using a Nitrogen mass-balance approach, accounting for
measured influx and efflux of Nitrogen, measured uptake of Nitrogen by plants, and
sediment, and estimated NH3 volatilisation (Frankenbach and Meyer, 1999).
The extent of Nitrogen removal depends on the design of the system and the form
and amount of Nitrogen present in the wastewater. If influent Nitrogen content is low,
wetland plants will compete directly with nitrifying and denitrifying bacteria for NH4+ and
NO3-, while in high Nitrogen content, particularly Ammonia, this will stimulate nitrifying and
denitrifying activity (Good and Patrick, 1987).
75
Figure 50. Nitrogen transformations in a constructed wetland treatment system (Source: Cooper et
al., 1996)
Nitrogen will also be taken up by macrophytes in a mineralised state and
incorporated it into plant biomass. Accumulated Nitrogen is released into the system during
a die-back period. Plant uptake is not a measure of net removal. This is because dead plant
biomass will decompose to detritus and litter in the life cycle, and some of this Nitrogen will
leach and be released into the sediment. Johnston (1991) shows only 26-55% of annual N
and P uptake is retained in above-ground tissue, the balance is lost to leaching and litter fall.
Heavy Metals removal mechanism
Heavy metals is a collective name given to all metals above calcium in the Periodic
3
Table of Elements, which can be highly toxic, and which have densities greater that 5g/cm
(Skidmore and Firth, 1983). The main heavy metals of concern in freshwater include lead,
copper, zinc, chromium, mercury, cadmium and arsenic. There are three main wetland
processes that remove heavy metals (Kayombo et al, 2004); namely, binding to soils and
organic material, sedimentation and particulate matter, precipitation as insoluble salts, and
uptake by bacteria, algae and plants (Kadlec & Knight, 1996). These processes are very
effective, with removal rates reported up to 99% (Reed et al., 1995). A range of heavy
metals, pathogens, inorganic and organic compounds present in wetlands can be toxic to
biota. The response of biota depends on the toxin concentration and the tolerance of
organisms to a particular toxin. Wetlands have a buffering capacity for toxins, and various
processes dilute and break down the toxins to some degree.
Evapotranspiration as a pollutant removal mechanism
76
Evapotranspiration is one of the mechanisms for pollutant removal. Atmospheric
water losses from a wetland that occurs from the water and soil is termed as evaporation
and from emergent portions of plants is termed as transpiration. The combination of both
processes is termed as evapotranspiration. Precipitation and evapotranspiration influence
the water flow through a wetland system. Evapotranspiration slows water flow and increases
contact times, whereas rainfall, which has the opposite effect, will cause dilution and
increased flow.
8.2.6. Design Requirements
The principal design criteria for a constructed wetland system includes substrates
types, pollutant loading rate and retention time, choice of wetland plant species, and area of
reed bed (Kayombo et al, 2004).
Substrates
Substrates may remove pollutants by ion-exchange, specific adsorption/precipitation
and complexation. Hydraulic permeability is one of the substrate selection criteria. The long
term efficiency of an emergent bed system is improved if the effluent is pre-treated prior to
discharge to the active bed.
The media depth should be about 0.6 m and the bottom is a clay layer to prevent
seepage (Kayombo et al, 2004). Media size for most gravel substrate range from 5 to 230
mm with 13 to 76 mm being typical. Wastewater flows by gravity horizontally through the root
zone of the vegetation about 100-150mm below the gravel surface. Outlet is typically 0.3 to
0.6 mm below bed surface. The environment within the SSF bed is either anoxic or
anaerobic. Oxygen is supplied by the roots of the emergent plants and is used up in the
Biofilm growing directly on the roots and rhizomes, being unlikely to penetrate very far into
the water column itself. SSF systems are good for nitrate removal (denitrification), but not for
ammonia oxidation (nitrification), since oxygen availability is the limiting step in nitrification
(Kayombo et al, 2004). The most common problem with SSF is, according to Kayombo et al,
2004, blockage, particularly around the inlet zone, leading either to short circuiting, surface
flow or both. This occurs because of poor hydraulic design, insufficient flow distribution at the
inlet, and inappropriate choice of porous media for the inlet zone. Properly-designed SSF
systems are very reliable.
Wetland plant species
While there is a recognition that the improvement of water quality in treatment
wetland applications is primarily due to microbial activity (Faulwetter et al., 2009; Kadlec and
Wallace, 2009), experience has shown that wetland systems with vegetation or macrophytes
has a higher efficiency of water quality improvement than those without plants (Coleman et
al., 2001; Tanner, 2001; Brisson and Chazarenc, 2009). The emphasis of constructed
wetland technology to date has been on soft tissue emergent plants including Cyperus
77
papyrus, Phragmites, Typha and Schoenoplectus (Okurut, 2000; Kadlec and Wallace, 2009).
A higher reduction efficiency for mass balances of N and P could be achieved by Phragmites
if water retention time is more than 5 days.
Wetland International, 2003, also extends on the role of vegetation in wetlands. The
most significant functions of the wetland plants in relation to water purification are their
physical effects. They provide a huge surface area for attachment and growth of microbes.
Therefore they play a vital role in the retention and removal of nutrients and help in
preventing the eutrophication of wetlands. A range of plants have shown their ability to
assist in the break down of wastewater. As good examples, there are the common reed
(Phragmites karka) and cattail (Typha Angustifolia) – these plants have large biomass both
above (leaves) and below (underground stem and roots) the substrate surface. Below a
figure showing the extensive root system of the plant:
Figure 51. Extensive root system of plants (source: modified from Cooper et al., 1996)
The other five roles of the wetland plants, besides the physical one, are:

Soil hydraulic conductivity,

Organic compound release through the root system at rates up to 25% of the
total photosynthetically fixed carbon which may act as a source of food for the
microorganisms (Brix, 1997),

Microbial growth,
78

Creation of aerobic soil – nitrification requires a minimum of 2 mg O2/l to
proceed at a maximum rate. It is evident that the rate of nitrification is most
likely the rate limiting for overall nitrogen removal from a constructed wetland
system (Sikora et al., 1995),

Aesthetic values.
The vegetation tends to increase the rates of water loss through evapotranspiration
when compared to rates of evaporation from bodies of open water (Jones and Humphries,
2002). Therefore, the aim of establishing the macrophyte zone will increase hydraulic
retention time; consequently increased biomass will facilitate evapotranspiration.
The ability for soft tissue macrophytes to grow and perform well has been
documented, especially for the high latitudes, temperature climate regions. Empirical
exploitation of plants is a common practice. Availability, expected water quality, normal and
extreme water depths, climate and latitude, maintenance requirements and project goals are
among the variables that determine the selection of plant species for constructed wetlands
(Stottmeister et al., 2003).
Climate
Precipitation and temperature must be monitored on site and must be used in the
climatic water balance. In terms of the climate monitoring, rainfall and actual evaporation
rate are the driving forces for determining the water balance (Morgan, 2014). Understanding
the wetland water balance will play an important role in establishing the success or failure of
the system, which is again determined by monitoring and understanding the meteorological
variables. These will be further used in the understanding of the macrophyte conditions,
specifically the water retention characteristics. Thus rainfall is measured directly with site
rain gauges and the potential evaporation can be determined by an empirical equation or
literature if the latter becomes impossible.
Potential evaporation (ETo) is a measure of the ability to remove water from the
surface through the processes of evapotranspiration and assuming non-limiting water
conditions (Clark et al. 1989). Actual evaporation (ET) is the quantity of water that is actually
removed from the surface through evapotranspiration. No single method of ETo
determination is likely to be ideal for all circumstances. Depending on the availability of data,
common methods may infer potential ETo for the modelling calculated from the potential
evaporation by application of an appropriate crop coefficient through a direct measure of
evaporation from a Class A-pan or an indirect estimate using an empirical equation, for
example, the Penman - Monteith equation.
The Food and Agriculture Organisation (FAO) methodology is considered the
international standard for predicting crop water requirements (Allen, 1998). Reference crop
evapotranspiration refers to ET from a uniform green crop surface, actively growing, of
uniform height, completely shading the ground, under well-watered conditions.
79
Sufficient data must be made available for calculating ETo using the Penman –
Monteith method, thus this method would be used in this study as presented in the equation,
below.
𝐸𝑇0 =
= 0.408∆(𝑅𝑛 − 𝐺) + γ
900
𝑢 (𝑒 − 𝑒𝑎 )
𝑇 + 273 2 𝑠
∆ + 𝛾(1 + 0.34𝑢2 )
Where
𝐸𝑇0
is the potential evaporation [mm/day],
𝑅𝑛
is the net radiation at the crop surface [MJ/m2/day],
𝐺
is the soil heat flux density [MJ/m2/day],
T
is the mean daily temperature at 2m height [℃],
𝑢
is the wind speed at 2 m height [m/s],
𝑒𝑠
is the saturated vapour pressure [kPa],
𝑒𝑎
is the saturated vapour pressure [kPa]
𝑒𝑠 − 𝑒𝑎
is the saturated vapour pressure deficit [kPa]
∆
is the slope of vapour pressure curve [kPa/℃], and
𝛾
is the phychometric constant [kPa/℃]
With ∆ the slope of the saturation vapour pressure curve in kPa/℃
∆=
4098𝑒𝑠
(𝑇𝑎 + 273.3)2
And 𝐺 the soil heat flux (MJ/m2/day) calculated form the current day’s (DOY) and
pervious days (DOY-1) average air temperature (𝑇𝑎𝑣𝑔 ).
𝐺 = 0.38[𝑇𝑎𝑣𝑔 (𝐷𝑂𝑌) − 𝑇𝑎𝑣𝑔 (𝐷𝑂𝑌 − 1)]
Where
𝑇𝑎𝑣𝑔 =
𝛾
(𝑇𝑚𝑎𝑥 + 𝑇𝑚𝑖𝑛 )
2
is the psychrometer constant (kPa/℃) and is calculated as:
𝛾=
0.00163𝑃𝑎
𝜆
With 𝜆 = 2.501 – 2.361X 10-3𝑇𝑎𝑣𝑔
Therefore, the climatic water balance components of the wetland consist of:
80
S = I + P + ET – D
where:
S
is the change in volume of water held in storage per unit area (mm)
P
is the volume of precipitation per unit area (mm)
ET
is the water lost through evaporation (climate and porous media
dependent) and transpiration (species dependent) per unit are (mm),
and
D
is the volume of water draining out the bottom of the macrophyte zone
per unit area (mm)
The constructed wetland should have the following key design features;

An inlet zone pond/basin that acts as a sedimentation pond and buffer to disperse
the water flows into the constructed wetland system. This feature reduces the
velocity of inflows, traps remaining coarse sediments and generally protects the
macrophytes zone. Wherever possible, sedimentation ponds should be separate
from the macrophytes zone so they can be isolated for maintenance.

Connection of the inlet to the macrophyte zone can be either by pipe or porous
rock weir. Where pipe connections are used it is important to have an initial open
water section in the macrophyte zone to help disperse flows. A high flow bypass
channel (to protect the macrophyte zone from scour and vegetation damage).

An extensively vegetated macrophyte zone. The vegetation is predominantly
emergent aquatic plants that support a complex of algal and bacterial microscopic
organisms, known as biofilms, which grow on the surface of the plants. This zone
also traps finer sediments and potential soluble pollutants.

Outlet zone which will allow for aeration of water before discharge into the river.
The disadvantages of a constructed wetland are (1) the land requirements (cost and
availability of suitable land), (2) current imprecise design and operation criteria, (3) biological
and hydrological complexity and our lack of understanding of important process dynamics,
(4) the costs of gravel or other fills, and site grading during the construction period, and (5)
possible problems with pests. Mosquitoes and other pests could be a problem for an
improperly designed and managed SSF (Kayombo et al, 2004).
8.2.7. Design of the constructed wetland
Design of a Settling system
tests
81
Samples of the Gara underground effluents were sent to Paterson & Cooke
(Johannesburg) to carry out some test on the effluent and advice on the thickener design.
Samples of Gara decant water and bench-top thickener underflow samples were sent for
comprehensive water quality analysis and solids elemental study at SGS respectively; and
samples of the settled solids removed from the effluents as received from the mine were
sent for standard fire assay at Mintek to determine the gold content in the material.
The following suite of laboratory tests was conducted in order to characterize the
static sedimentation behavior of the solids in the effluent and to determine the optimum
process conditions for flocculation and settling within a thickener:

Water quality and specific gravity tests

Particle size distribution (PSD)

Flocculant type screening (performed in 100 ml measuring cylinders)

Optimisation of feed effluent solids concentration

Optimisation of flocculant dose

Settling rate envelope under static consolidation conditions (un-raked)

Static mud bed consolidation (24h raking)
The results of the static sedimentation tests also provided the operating parameters
for set-up of the dynamic thickening test. The bench top dynamic thickening test is designed
to determine the critical solids flux rate and/or rise rate values for calculating the optimum
thickening area required for a specific process load (t/h). The data generated is required to
adequately size the diameter and number of thickeners required by the process. This test
was conducted in a 100 mm diameter bench-top dynamic thickening rig ( see Figure 52
below). The Gara effluent characteristics was also determined in the laboratory to
understand the nature of the initial state of the effluent and the preliminary thickener feed
effluent solids concentration was determined through a series of 250 ml cylinder settling
tests at a range of effluent feed solids concentrations.
82
Figure 52. Bench-top dynamic thickening rig (P&C, 2011)
Results
Table 22 below shows the characteristics of the Gara effluent - The conductivity of
the effluents, which is higher than that of the process water due to the solids content, falls
within the moderately high range (1.5 to 3 mS/cm). Settlings effluents could be expected in
the presence of clay ores as a result of the double layer compression effect at effluent
conductivity values of 2 mS/cm and above.
It was observed that Gara effluents were coagulated in the natural state (prior to
flocculation). A naturally coagulated material is defined in layman’s terms as a material that
settles overnight (without any flocculation) and producing a layer of clear supernatant water
above a settled bed of solids. As a result, no further effluent conditioning measures is
required and therefore no flocculation and/or settling problems are expected based on the
natural coagulated state of the thickener feed effluents.
Table 22. Effluent characteristics (SGS, 2011)
Gara
Parameter
effluent
pH
8.0
Conductivity (mS/cm)
Specific gravity (g/cm3)
1.6
5
Solids
2.72
Liquid
Particle size distribution
1.00
6
5 Water pycnometer method
6 Laser particle size analysis method
83
Parameter
Gara
effluent
d10
1.80
d50
9.16
d80
% clay size material
19.90
7
Natural colloidal state
~ 27
settling
The particle size distribution also is shown below – more than 30% of solids of the
effluent is less than 5 micron.
Figure 53. Particle size distribution of the solids in effluent (P&C, 2011)
Flocculant DP8629 performs better in the gara effluent than thenother tested
flocculant (see figure below)
7 Clay size refers to the minus 5 micron fraction and has no reference to the clay mineral content.
84
Flocculant Selection
Settling rate (m/h) - 100ml cylinder
45.0
40.0
35.0
30.0
25.0
20.0
Yalea
15.0
Gara
10.0
5.0
0.0
10
919
336
333
156
DP8629 6260
5250
Flocculant
Figure 54. Flocculant screening results (SGS, 2011)
The preliminary thickener feed slurry solids concentration was determined
through a series of 250 ml cylinder settling tests at a range of slurry feed solids
concentrations.
Figure 55 and Figure 56present the results of the thickener feed slurry solids
concentration tests
85
Preliminary f eed slurry solids concentration
Settling rate (m/h) - 250ml cylinder
45.0
40.0
35.0
30.0
25.0
20.0
Yalea
15.0
Gara
10.0
5.0
0.0
0.0
2.5
5.0
7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0
Slurry f eed solids conc. (%m)
Figure 55 : Thickener Feed Effluent Solids Conc. (settling rate)
Preliminary f eed slurry solids concentration
50
45
Clarity (wedge no.)
40
35
30
25
Yalea
20
Gara
15
10
5
0
0.0
5.0
10.0
15.0
20.0
25.0
30.0
Slurry f eed solids conc. (%m)
Figure 56 : Thickener Feed Effluent Solids Conc. (clarity)
8
8 The clarity value was measured in a turbidity wedge and represents a relative clarity measurement where 0 represents complete ly unclear supernatant
and 50 presents the clarity of tap water.
86
Bearing in mind that this optimization is conducted prior to the optimization of
flocculant dose rate, this range of settling tests have been conducted at an arbitrary
flocculant dose.
These conservative design parameters should assist in ensuring that the equipment
is able to deal with fluctuating water qualities and quantities. The actual suspended solids
readings are much less than 5%.
Preliminary f locculant dosage
Settling rate (m/h) - 250ml cylinder
90.0
80.0
70.0
60.0
50.0
40.0
Yalea
30.0
Gara
20.0
10.0
0.0
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
Flocculant dose (g/t)
Figure 57 : Flocculant Dose (settling rate)
87
Preliminary f locculant dosage
50
45
Clarity (wedge no.)
40
35
30
25
Yalea
20
Gara
15
10
5
0
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
Flocculant dose (g/t)
Figure 58 : Flocculant Dose (clarity)
Error! Bookmark not defined.
The preliminary flocculant dose was verified during the subsequent settling rate
envelope and dynamic thickening tests to be 25 g/t for gara effluent.
Figure 59 and Figure 60 present the static settling rate as a function of flocculant
dose and effluent solids concentration as determined through 1 litre cylinder tests.
88
Settling rate as a f unction of f locculant dose
Settling rate (m/h) - 1 litre cylinder
50
40
30
Yalea
20
Gara
10
0
0.0
5.0
10.0
15.0
20.0
25.0
Flocculant dose (g/t)
Figure 59 : Static Settling Rate
Settling rate as a f unction of thickener f eed slurry solids concentration
Settling rate (m/h) - 1 litre cylinder
50
40
30
Yalea
20
Gara
10
0
0.0
2.5
5.0
7.5
10.0
Slurry solids conc. (%m)
Figure 60 : Static Settling Rate (P&C,2011)
Figure 61 presents the underflow solids concentrations achieved in the 1 litre cylinder
settling tests under static un-raked conditions vs. the underflow solids concentrations
achieved after a static 24 hour raking period.
89
Underf low solids concentration as a f unction of f locculant dose
65
Underf low solids conc. (m%)
60
55
Yalea
50
Gara
45
24h raked U/F
40
24h raked U/F
35
30
25
20
0.00
5.00
10.00
15.00
20.00
25.00
Flocculant dose (g/t)
Figure 61 : Static Mud Bed Compaction
The results of the different laboratory dynamic tests are presented below tables and figures Figure 62 and Figure 63 present the solids flux and hydraulic flux (rise rate) curves
that were generated for each of the two underground discharge effluent samples in a 100
mm diameter bench-top dynamic thickener at the following test conditions:
Table 23. Dynamic Thickening Test Settings and Data
Solids flux
(t/m2.h)
YALEA
0.2
0.3
0.4
GARA
0.2
0.3
0.4
Rise rate
(m/h)
Test Settings
Mud bed
Rake
height
speed
(mm)
(rpm)
Results
Feed slurry Avg floc Overflow Underflow
solids conc.
dose
clarity solids conc
(%m)
(g/t)
(wedge)
(%m)
3.5
5.3
7.1
100
100
100
2.5
2.5
2.5
5.5
5.5
5.5
25
25
25
25
25
25
59.3
56.4
55.1
4.1
6.1
8.2
100
100
100
2.5
2.5
2.5
4.8
4.8
4.8
25
25
25
49
49
49
51.9
51.4
47.7
90
70
14
60
12
50
10
40
8
30
6
20
Rise Rate (m/h)
Underflow Solids Conc. (%m)
Solids Flux Curve
4
Yalea U/F
Gara U/F
10
2
Yalea Rise Rate
Gara Rise Rate
0
0
0.0
0.1
0.2
0.3
0.4
0.5
Solids Flux Rate (t/(m^2.h))
Figure 62 : Solids Flux Curves
Hydraulic Flux Curve
50
45
Overflow Clarity
(turbidity wedge no.)
40
35
30
25
20
15
10
Yalea O/F
5
Gara O/F
0
0
1
2
3
4
5
6
7
8
9
Rise Rate (m/h)
Figure 63 : Hydraulic Flux Curves
91
Design parameters
Table 24 presents the recommended thickener operating parameters for the Gara
samples based on the combined results of the static sedimentation and dynamic thickening
tests.
Table 24. Thickener Operating Parameters
Parameter
Optimum flocculant type
Flocculant dosing concentration (%m)
Thickener feed solids concentration (%m)
Optimum flocculant dosage (g/t)
Gara
effluent
DP8629
0.025
5
25
Table 25 presents the recommended thickener sizing parameters for the Gara
samples based on the combined results of the static sedimentation and dynamic thickening
tests.
Underground discharge thickeners are typically sized on rise rate as the critical sizing
parameters due to the low solids concentration expected in the thickener feed. A maximum
rise rate of 4 m/h, corresponding to a solids flux rate of 0.2 t/(m 2h), is therefore
recommended for thickener sizing.
Table 25. Thickener Sizing Parameters
Parameter
Optimum solids flux rate (t/m2.h)
Hydraulic rise rate (m/h)
Overflow clarity
(turbidity wedge no.) - estimate
Underflow solids concentration (%m) - estimate
Gara
effluent
0.2
4
49
50
92
Table 26 presents the minimum thickening area and estimated thickener diameter to
treat the above process flows based on a critical rise rate of 4 m/h.
Table 26. Minimum Thickening Area (P&C, 2011)
Process stream
Critical Rise
Rate (m/h)
Max Flow
Rate (m3/h)
Thickening
Area (m2)
Thickener
Diameter (m)
Gara
4
160
40
7
underground discharge
Table 27 presents a more conservative approach were the thickener sizes are
rounded up to the next available standard size, which thereby incorporates a large safety
factor. This approach is preferred due to the following reasons:



Thickener sizes are small and therefore cost implications of going larger
should not be prohibitive
The larger thickening area for the Yalea material would in all probability sort
out the somewhat poor overflow clarity
Operational problems around surges of high solids concentration effluents or
larger volumetric flow rates should be reduced due to the larger available
thickening area.
Table 27. Recommended Thickener Diameter (P&C, 2011)
Process stream
Critical Rise
Rate (m/h)
Max Flow
Rate (m3/h)
Thickening
Area (m2)
Thickener
Diameter (m)
Gara
underground discharge
2.0
160
79
10
The underflow solids concentrations quoted in table above are based on a mud bed
height of 100 mm and essentially exclude the consolidation effect of bed height. Typically, an
increase of 10 to 15% in underflow solids concentration is expected at bed heights
comparable to full scale (i.e. +1m).
93
Figure 64. A 3D representation of the proposed thickener design (TWP, 2011)
Inlet zone
The inlet zone of a constructed wetland is designed as a ponding basin and has two
key functional roles.

The primary role is to remove coarse to medium sized sediment (i.e. 125 μm
or larger) prior to flows entering the macrophyte zone. This ensures the
vegetation in the macrophyte zone is not smothered by coarse sediment and
allows this zone to target finer particulates, nutrients and other pollutants.

The second role of the inlet zone is the control and regulation of flows
entering the macrophyte zone and bypass of flows during ‘above design flow’
conditions. The outlet structures from the inlet zone (i.e. sedimentation basin)
should be designed such that flows up to the ‘design flow’ (typically the 80
percentile flow rate) enter the macrophyte zone whereas ‘above design flows’
are bypassed around the macrophyte zone. In providing this function, the inlet
zone protects the vegetation in the macrophyte zone against scour during
high flows.
Macrophyte Zone Design
The layout of the macrophyte zone needs to be configured such that system
hydraulic efficiency is optimised and healthy vegetation sustained. Design considerations
include:
94

The range of suitable extended retention depths is 0.25-0.6 metres (providing
suitable plant species are selected for deeper extended detention depths),
depending on the underground pumping operations of the wetland and target
pollutant.

Flow velocities must be kept low and maintained through the zone.

The ground contours under the water of the macrophyte zone should be
designed to promote a sequence of ephemeral, shallow marsh, marsh and
deep marsh zones in addition to small open water zones. The relative
proportion of each zone will be dependent on the target pollutant and the
wetland hydrologic effectiveness.

The macrophyte zone is required to retain water permanently and therefore
the base must be of suitable material to allow water flow at the effective rate.
If in-situ soils are unsuitable for water retention, a clay liner (e.g. compacted
300 millimetres thick) should be used to ensure there will be permanent water
for vegetation and habitat.

The optimum treatment configuration is a wetland densely vegetated with
species that provide a high density of stems in the submerged zone (thereby
maximising the contact between the water and the surfaces on which
microorganisms grow), while providing uniform flow conditions.

The main potential advantage to an overall densely vegetated system (for this
study) would be the reduction of dissolved oxygen in the near bottom water
and the surface sediment layer. The presence of anaerobic sediment is
desirable for denitrification.

The macrophyte zone outlet structure needs to be designed to provide a
detention time (usually 48 to 72 hours) for a wide range of flow depths.
(1)
Retention Time and Hydrologic Effectiveness
Retention time is the time taken for each ‘particle’ of water entering the wetland to
travel through the macrophyte zone assuming ‘saturated’ flow conditions. It should be noted
that retention time is rarely a constant and the term ‘retention time’ is used to provide a point
of reference in modelling and determining the design criteria.
Hydrologic effectiveness is a measure of the volume of water captured and treated
within the wetland and is expressed as a percentage of the water pumped form the
underground. For example the constructed wetland presented in this study is designed for
80% percentile of the measured pumping rates. Analysis of the flow rates indicates this to be
approximately 2800m3/day, which is equivalent to 32 L/s.
The range of retention times achieved in a constructed wetland is influenced by the
type of structure used. The volume of the permanent pool also has a significant effect on the
range of retention times achieved. Water level control is desirable in wetland design to
95
enable maintenance and to assist with vegetation establishment. The HEC RAS 9 Model was
set up to simulate the flow conditions based on the input site characteristics. The main aim
was to establish the flow inundation (water level height) and flow velocities within the
wetland. The inputs into the model were based on the requirements for the wetland to be
effective at the proposed location. Figure 65, is an illustration of the proposed wetland
location, where Figure 66 provides the cross sectional profile of the current land topography
against the proposed wetland topography. The proposed wetland characteristics where set
up in the HECRAS Model. The output from the model is presented in the Appendix D, which
aims to establish the height of water and flow boundaries/velocities for the proposed
wetland.
The different sections of the constructed wetland are described below.
9
HEC-RAS is designed to perform one-dimensional hydraulic calculations for a full
network of natural and constructed channels. Designed by the US Army Corps of Engineers
96
Figure 65. Illustration of the constructed wetland layout (Source: DWA, 2014)
97
Figure 66. Illustration of the cross sectional through the constructed wetland layout. Inlet zone (0m –
250m), Macrophyte zone (250m – 650m) and Outlet zone (650m - 800m)
(2)
Saturated Hydraulic Conductivity (Darcy’s Law)
To determine the hydrologic properties of the macrophyte zone material, a
comparison is made based on the material tests (by example only). A one cubic metre (1 m3)
volume of empty space would hold to 1000 L of water. If this same volume is filled with dry
macrophyte material it will be able to hold, for example 240 L of water as total storage, which
represents a porosity of 24%. However, if this volume is drained under gravity, only 80L (8%)
of the total volume will drain. This is an important physical property of the material known as
specific yield. The specific yield represents 33% of the porosity and the rest of the water
(160 L) is retained in the material particle grains. It is therefore not available as water as it is
trapped in the material, only plants can use it via root uptake as moisture.
Specific yield is one of the most important parameter as it is the component of water
that can drain from the material under gravity, basically allowing for flow in the wetland.
Below the saturation level of specific yield, the material will retain the water. The retained
component of the water in the material will be only accessible by plants such as reeds.
Therefore, if water into the macrophyte zone is greater it will force water out the macrophyte
material and flow above surface.
A very basic relationship for the one dimensional flow in porous macrophyte
zone/media is established through Darcy’s Law. The formula derives the volumetric flow rate
through the function of flow area, and elevation for saturated flow conditions. Based on the
Darcy’s equations it was calculated that the required hydraulic conductivity (K = 0.32m/s), to
allow for residence time within the macrophyte zone of 48 hours. The K value in this very
basic example will be in-line with a gravel material for the constructed wetland dimensions
which have been setup in this study. A very basic illustration of the macrophyte zone of this
study is illustrated in the Figure 67.
98
Porous Media
K=
f=
dh = (h 2 - h 1) =
J = dh /dx =
q = -KJ =
Q = qA =
v = q /f =
Hydraulics
Units
0.32 m/s
3
1.00 m
-0.004 m/m
1.28E-03 m/s
6.40E-02 m3/s
4.27E-04 m/s
5529.6 m3/d
110.5920 m/d
h2 =
121
/”
h
Area
Flow through macrophyte zone
122
h1 =
1
Flow
Macrophyte
zone
L=
L
50
250
Figure 67. Darcy’s flow equation for saturated flow through the macrophyte zone
Modelling horizontal flow in the constructed wetlands with intermittent loadings
requires transient variably saturated flow models as these systems are highly dynamic,
which adds to the complexity of the overall system. Based on professional experience of
applicable models the two dimensions such as HYDRUS-2D model can be used to provide
an improved estimate of the saturated flow conditions. It is thus suggested, for practical
application:

Undertake experiments to measure the porosity and the saturated hydraulic
conductivity parameters of the macrophyte zone material;

Instrumentation and monitoring methods for determining the macrophyte zone
material saturation should be assessed as these collective parameters are
used to determine the flow rates. These parameters should be assessed as
they are used in the HYDRUS-2D model for simulations of the water flow and
ultimately assess the achievable targets of the wetland.

It may be possible to further establish the influence of the particulates and
biomass growth on the hydraulic properties of the macrophyte zone, as these
estimates could further increase its prediction water quality through the
wetland.
The HYDRUS-2D model is a finite-element model, which numerically solves the
Richards’ equation for saturated-unsaturated water flow. Richards equation for one
dimensional flow requires knowledge of the soil hydraulic functions, i.e., the soil water
retention curve, (h) , describing the relationship between water content () and the
99
pressure head ( h ), and the unsaturated hydraulic conductivity function, K(h) defining the
hydraulic conductivity (K) as a function of (h).
The above equation presents a very broad understanding of the Darcy’s Law through
the macrophyte material, but it is essential that experiments be undertaken for composite
samples. In order to adequately quantify the flow in the macrophyte zone, it is imperative
that sufficient information of the porous media hydraulic characteristics and the dynamics of
the interstitial water are assessed. There are numerous techniques for measuring and
monitoring these variables. For example, the porosity is defined as the ratio of volume of
voids to the total volume of the macrophyte material. The effective porosity defines the
volume of water that a given volume of bulk porous medium can contain. Water content
reflects the volume of the void space that is filled with water, relative to the bulk porous
medium.
Soil hydraulic properties will required in order to model transport of water in the
macrophyte zone. The ability of a porous media to retain and transmit water is characterized
by the relationships between water content, (), matric pressure head, (h), within the profile
and the hydraulic conductivity, (K). The most substantial challenges for experimental and
theoretical investigations of fluid flow in unsaturated porous media result from the extremely
nonlinear behaviour of the hydraulic properties as a function of saturation and the highly
irregular nature of pore geometry. The aim of these macrophyte zone studies must provide a
perspective of saturated and unsaturated hydraulic characterization and measurement used
in the suitable material.
(3)
Soil Water Measurements and Field Monitoring Experiments
It is important also to provide understanding into direct observations into key
hydrological processes and techniques in measuring soil hydraulic properties, monitoring soil
water dynamics. The understanding of these measurements and data adds value as a
reference for the evaluations of the soil and macrophyte media hydraulic behaviour, which
ultimately is used in a model. Measurements of the macrophyte media may include;

soil physical characteristics,

water retention properties, and

hydraulic conductivity
characteristics.
characteristics
as
well
and
in-situ
hydraulic
Soil hydraulic characterisation may involve both in-situ and laboratory
measurements. A consistent procedure must be adopted for characterising macrophyte
media profile. An undisturbed core is taken and the water retention characteristic, bulk
density and particle size distribution are determined in the laboratory. The particle size
distribution and bulk density may be used to develop pedotransfer relationships, within the
macrophyte zone material. However, since the macrophypte zone will be developed
100
ensuring heterogeneity, only a few samples may be analysed. Contractors should therefore
be scrutinized and consistently monitored during construction. One could look at relating
STP tests to the hydraulic properties, since it would be assumed heterogeneous and the
spatial distribution of the hydraulic characteristics will therefore be constant.
Monitoring the pore water for contaminants is a difficult task and requires the use of
indirect methods. There are, however, a number of sensing techniques for detecting
moisture or the electrical changes due to moisture change that will be considered for longterm monitoring. These techniques include tensiometers and TDR probes.
(4)
Water retention characteristic
Conventional methods of soil water retention characteristic determination include
both laboratory and field techniques (Klute, 1986; Bruce and Luxmore, 1986). Although
many of the standard laboratory techniques can be used to determine the water retention
characteristics of the porous media, the modified controlled outflow method is reported in
detail. This method holds some promise for the accurate characterisation of the soil pore
structure over the range of moisture contents close to saturation.
Accurate measurements over this range are desirable to characterise the larger pore
characteristics which conduct water rapidly during intense rainfall. The retention
characteristics samples can be used used together with field measurements of both
saturated and unsaturated conductivity to define the possible macrophyte material
composite.
Recent emphasis on characterising unsaturated hydraulic characteristics of soils in
hydrology, soil science and chemical transport studies has fostered renewed interest in liquid
retention measurements. It is recommended that the methodology for defining the liquid
retention characteristic of macrophyte material established by monitoring equilibration of the
matric pressure rather than equilibration of the liquid volumetric content, as in conventional
methods.
(5)
Macrophyte zone monitoring
Wells and piezometers should be installed to measure the saturated water level
dynamics through this zone.
Soil moisture alone does not provide information on the driving force for unsaturated
fluid flow in the macrophyte zone or allow the direction of water movement to be determined.
The driving force for water movement in macrophyte zone soils is the matrix (or suction)
potential, which is usually expressed in terms of a vacuum (using negative pressure units).
The matrix potential of a given soil is determined by the texture and moisture content of the
soil. If two soils of two different textures (one fine, the other coarse) are placed in contact
with each other and allowed to reach equilibrium, water will move between them until they
come to the same matrix potential. The moisture content of the fine-textured soil, having
more small pores, will be appreciably higher than that of the coarse-textured soil. With
101
knowledge of the soil texture, the matrix potential can be estimated from the measured soil
moisture content; however, it is much better to measure the matrix potential directly with a
tensiometer.
Tensiometers should be considered for measurement of the macrophyte zone
material and ultimately establishing water pressure, which would translate to the water at the
surface. Tensiometers can also be used to measure positive pressure in soils that are
saturated and thus can be used for monitoring perched water tables, which is an objective of
the constructed wetland. Tensiometers have the advantage in that they provide a direct
measure of the water potential. They require a continuous water column that extends from
the measurement point to the pressure transducer located at the surface. A tensiometer
typically consists of a porous ceramic cup or plate attached to a nonporous tube, with some
means to measure a vacuum inside the tube. When the porous cup is placed in hydraulic
contact with soil, and the cup and tube filled with water and sealed, a vacuum will develop in
the tube as water flows from the ceramic into the soil. At equilibrium, the measured vacuum
inside the tensiometer is equal to the matrix potential of the soil. An example of how the
position of the tensiometers would be placed in the macrophyte zone in order to study the
water within this zone is illustrated in Figure 68.
102
Figure 68. Schematic of the root zone and components of the tensiometers
Surface flow can be monitored within the macrophyte zone consisting mainly of
successive runoff plots, (20 x 3 m) may be installed to further develop understanding of the
surface flow.
(6)
Quantification of the hydrological processes
The macrophyte zone will be constructed as an isolated catchment; therefore the
boundary conditions will be the sides and the bottom of the wetland. In order to quantify the
hydrological processes detailed observations of the surface and subsurface dynamics will
need to be made at various experimental sites. Developing understanding of these
mechanisms at these respective sites allow for some generalisation or insight into the
classification of critical wetland flow criteria for estimating the outflow responses.
Consequently, the wetland flow generation processes and nitrate breakdown can then be
deduced from the hydrometric observations of the dynamics of soil water, and flow
responses to the inflow and evaporation. It is recommended that these mechanisms will
initially be quantified using, physically based techniques. These techniques may be applied
103
to the wetland in order to predict the outflow estimate and to define algorithms and
methodologies for application to the macrophyte zone deterministic hydrological models.
(7)
Material and stratigraphy
The landscape of the wetland must be characterised by a broad flat drainage system,
with no dendritic drainage systems especially in the macrophyte zone. This will aim to
ensure the ability for water to be spread evenly throughout the zone. If the material
conditions are constant, theoretically the floor gradients would ensure flow velocity in the
macrophyte zone. The flow gradients may also be established by developing stratigraphic
material units of varying hydraulic properties. For example, horizontal sandy units and be
developed allowing water to build up (like sponge), spread the water throughout the unit and
only flow out (at the bottom) when saturated.
Problems associate with clogging of macrophyte zone is the clogging of the gravel
media, which in turn creates surface flow. This results in reduced treatment efficiency since
the water does not come into contact with the macrophyte material.
Wetland plants
In the subtropical and tropical climate, C. papyrus is one of the most interesting
macrophytes because it is among the most productive plants in wetlands (Kansiime et al.,
2005; Heers, 2006; Perbangkhem and Polprasert, 2010).
Figure 69. Photos of the above ground Cyperus papyrus in the water body
In natural and constructed wetlands, macrophyte root structures provide microbial
attachment sites. In an experimental microcosm set-up, Gagnon et al. (2007) found that
microbes were present on substrates and roots as an attached biofilm and abundance was
correlated to root surface throughout depth. Indeed planted wastewater treatment systems
outperform unplanted ones, mainly because plants stimulate below ground microbial
populations (Gagnon et al., 2007). Plant species root morphology and development seems
to be a key factor influencing microbial– plant interactions. Kyambadde et al. (2004)
measured a higher root surface and microbial density in a constructed wetland planted with
104
C. papyrus (average root surface area 208.6 cm2), where the root recruitment rate per
constructed wetland unit was 77 roots per week for C. papyrus, and C. papyrus had more
adventitious roots and larger root surface area. Furthermore, C. papyrus seems to promote
greater nitrogen removal efficiencies, through nitrification and denitrification rates of bacteria
associated with it roots (Morgan et al., 2008).
The average daily water vapour flux from the papyrus vegetation through canopy
evapotranspiration in a wetland located near Jinja (Uganda) on the Northern shore of Lake
Victoria was approximated by Saunders et al. (2007) as 4.75 kg H2O m2 d-1 (= 4.75 mm/d),
which was approximately 25% higher than water loss through evaporation from open water
(approximated as 3.6 kg H2O m2 d-1). Jones and Muthuri (1985) reported an
evapotranspiration rate of 12.5 mm/day at the fringing papyrus swamp on Lake Naivasha,
while Kyambadde et al. (2005) reported 24.5±0.6 mm/d for a subsurface horizontal flow
wetland in Kampala (Uganda). Evapotranspiration rates vary sharply since they depend on
numerous factors influencing the ecosystem's prevailing micro-climate, as listed by Kadlec
and Wallace (2009). For example, common reed transpiration rates oscillate between 4.712.4 mm/day depending on meteorological conditions (Holcová et al., 2009).
Evapotranspiration (ET) by plants can significantly affect the hydrological balance of
treatment wetlands. The water lost through ET concentrates pollutants within the wetland,
while the volume reduction results in longer hydraulic retention times (Kadlec and Wallace,
2009). For low loaded systems or systems with longer retention times, such
evapotranspiration rates can exceed the influent wastewater flow, leading to a zero
discharge.
The Outlet Zone
A outlet will have the primary purpose of the outlet zone provides a cascade of water
at the outlet before the discharge point. This will facilitate the aeration of water by allowing
water to drop over an elevated area. The drop zone will have gabions over the surface to
prevent erosion and facilitate the mixing of the water.
9 Conclusions & Discussions
9.1.
Impact of the Underground Effluent on the Faleme river
9.1.1. Characteristics of the effluent discharged
Discharge into the Faleme river started from 2010 to date indicating there has been
an effective 4 years of effluent being discharged into the Faleme after passing through the
existing settlers. Flow monitoring records indicated an average 1671m3/day has been
105
discharged. Quality issues (refer to the discharge limits in Table 2) with the effluent as per
the analyses record include elevated total suspended solids (TSS), relatively high arsenic
(As) levels in the solids, elevated aluminium (Al) and iron (Fe) levels, high nitrate (NO3) and
increasing sulphate (SO4) concentrations.
Average TSS value recorded over the reporting period is 1435.7 mg/l with As content
varying from 64 to 78 mg/kg (vs. 33 mg/Kg as guideline – McDonald, 2000a) and
Manganese (Mn) level reaching 2050 mg/Kg vs. a guideline 1100 mg/Kg (McDonald,
2000a). No dissolved Cr & As (mobile form) was picked up in the past analyses, only in July
2014 when traces (0.017 mg/l versus a limit of 0.01 mg/l) of dissolved As were identified.
The Gara ore being relatively Arsenic (As) depleted (Lawrence et al., 2013), that being
locked up in pyrite (<1 wt.% As), As identified in the settling ponds sediments maybe due to
localised arseno pyrite zones which hasn’t been picked up in previous petrography work.
The high Manganese (Mn) levels (2050 mg/Kg vs. 1100 mg/Kg guideline) are likely sourced
from the Kofi limestones, which are common along the Senegal-Mali Shear Zone. It is to be
however noted that the recent TSS analyses showed Mn level at 150 mg/kg (much below
the 1100mg/kg guideline). With regard to its As & Cr content, the TSS of the effluent is
classified as toxic according to McDonald et al. (2000a).
2014 survey indicated the nitrite (NO2) value of 10.2mg/l, the nitrate (NO3) value of
96.5 mg/l and Ammonia (NH3) of 3.59 mg/l in the effluent are within the water discharge
limits (15mg/l for Ammonia and no limit is specified for Nitritre) except Nitrate which
exceeded the Malian discharge limit of 30 mg/l. Average nitrate level over the reporting
period is 159 mg/l. the most probable source of the nitrate is the explosive used in mining.
Fe and Al, although not analyzed for in the 2014 survey, record over the 4-years
reporting period indicated high levels of these two elements in solution. Iron level is
fluctuating around an average of 19.8 mg/l with a maximum of 110 mg/l recorded on May
2013 (vs. a limit is 2 mg/l) and the maximum value recorded was 51.3 mg/l (May 2013) and
the last aluminum value was at 4.79 mg/l which is above limit of 1mg/l. the source of both Fe
and Al is believed to be the local and regional geology.
Sulfate level is below the guideline of 1000 mg/l but has increased since 2010. The
levels should continue to be monitored as it is possible that it may increase over time as the
mine working ages and continue to access new areas exposing new faces to oxidation.
From above, it is clear that Gara effluent could only impact on the Faleme through
sedimentation (with associated relatively small amount of immobile As in the sediments) and
Eutrophication through the elevated NO3 levels. Although below the discharge limit, SO4
which is traditionally known to be associated to mining, should be seen as a potential
pollutant.
9.1.2. Characteristics of the faleme
106
Faleme water quality
When water quality results are compared to guidelines derived from appropriate
sources (DWAF, 1996, Bain and Stevenson, 1990) the water quality in the Faleme River can
be considered fair. The pH values obtained from the in situ analysis at the sites considered
in the 2014 survey were found to be mostly neutral and would not be seen as limiting factor
for local aquatic biota.
The concentrations of dissolved ions (conductivity) in the Faleme River were found to
be relatively consistent, with few fluctuations as sites move downstream. Although
anthropogenic activities are occurring adjacent to the rivers, the activities were not
dramatically influencing dissolved ion concentrations. The consistent conductivity results
also indicate that the dilution capacity of the Faleme River is sufficient to dilute dissolved
salts to an extent whereby negligible effects would be observed in local aquatic biota. The
conductivity was seen to slightly increase between the upstream and downstream sites of
the Gara settling pond discharge. This increase may be attributed to the discharge itself
which had a conductivity of 1332 µS/cm resulting in a slight increase between the sites. The
conductivity then returned to normal levels at the Faleme Weir site.
Sulphate concentrations are on average low (2 -5 mg/l) in the Faleme River, sulphate
levels increased downstream of the artisanal mining areas and then decrease again at the
nearest downstream site, possibly as a result of dilution. The Gara settling pond was found
to contain elevated concentrations of sulphate, at 198 mg/l, which decreased to 94 mg/l at
the confluence point between the Gara settling pond discharge and the Faleme River.
Dilution is occurring to such an extent in the Faleme River that no negative effects on
aquatic biota would be observed at downstream regions.
Nitrate molecules were found to be exceeding guideline concentrations in the Gara
settling discharge point at the confluence with the Faleme River. Although high
concentrations were observed at this site, the concentrations measured immediately
downstream of the site, at DS Gara settling, was found to be below guideline values
indicating a negligible effect on local aquatic biota. This is evident by the absence of any
algae bloom on the water surface.
No dissolved As & Cr could be observed in the sites located within the upper reaches
of the study focus area indicating that limited sources exist between these sites. Dissolved
As concentrations were observed to increase and were in excess (24 times), above
guideline values downstream of the artisanal mining village Sasanba. These concentrations
were such, that negative effects would be observed in local aquatic biotic structures.
Furthermore, the elevated As concentrations provide an indication that artisanal mining
activities can produce dissolved As metals in associated water sources. Dissolved
concentrations of Cr were observed to be elevated at sites associated with artisanal mining
activities and these included sites downstream of Sasanba as well as artisanal activities in
107
the Gara River. These results indicate that artisanal activities are elevating Cr concentrations
in the local river systems.
Concentrations of Cadnium (Cd) and Mercury (Hg) were found to be below detection
limits at most sites with the exception of sites associated with artisanal activities which
served to increase these dissolved metals to concentrations which were detectable.
Dissolved Lead (Pb) concentrations were also observed to be above threshold effect levels
at the site downstream of Sasanba indicating that lead is possibly produced as a result of
artisanal mining activities as Pb is found in batteries and fuels/lubricants extensively used in
artisanal mining activities.
Above shows that the quality of the faleme is largely impacted by anthropogenic
activities such as artisanal mining activities and very negligible impact from the gara effluent.
Faleme sediment quality
Arsenic (As) and Chromium (Cr) are above acceptable levels at Faleme downstream
of the operation. These sediments according to McDonald et al. (2000a) are classified as
toxic and constitute a threat to the aquatic biota. These same results are observed at the
Faleme upstream of the Gounkoto operation with even higher concentrations in the past
surveys – indicating that the mine effluent is not the only source for this high level of metals.
As and Cr seem to be occurring naturally even though sediments sample from the settling
ponds (PC1) returned also a high value of As. The Chromium (Cr) enrichment maybe
coming from the mafic-intermediate intrusive/extrusive Faleme rocks or younger dolerites, or
sericite alteration associated with the As-rich ore bodies.
The analysis of Hg in sediments revealed elevated and enriched concentrations in
sites located downstream of artisanal mining activities. The highest concentrations of Hg
were obtained at the artisanal area at Djidian and can be linked directly to gold washing
practices in place there. As seen in the figure below (Figure 70) Hg is used and left adjacent
the river and thereby can be seen as a point source of the pollutant.
108
Figure 70. Artisanal gold processing chemicals. A: Mercury and other metals used in gold processing.
B: Processing chemicals left adjacent the river
The Faleme sediments are classified as toxic according to McDonald et al. (2000a)
and constitute a threat to the aquatic biota. The primary source of the pollutants is from the
surrounding geology followed by artisanal mining activities. The mine through discharge has
a negligible contribution to the Faleme polluted sediment.
Aquatic Macroinvertebrates
The macroinvertebrate assemblages at the sites generally represented medium term
conditions in the local aquatic habitats. Due to the time of the July survey, water quality
constituents may have been diluted, due to rainfall and increased flow, allowing for the
presence of improved water quality based on water quality analyses. However,
macroinvertebrate assemblages at the sites were composed of largely pollution tolerant
taxa. This is confirmed by low ASPT and EPT % contribution values obtained at the sites.
This is further confirmed by the absence of families such as Perlidae, Prosopistomatidae and
in general the Ephemeroptera and Trichoptera groups.
Based on these results the long term conditions at the sites in the Gara River can be
considered fair/poor whilst conditions in the Faleme River can be considered poor at sites in
the upper reaches (artisanal mining areas) and fair in the lower reaches (improved due to
the presence of the Weir). Furthermore, it can be stated that there is a definite overall
modified status result for the macroinvertebrate communities present and identified at sites
due to numerous anthropogenic activities as seen in the figures below.
Figure 71. Possible negative impacts at Farandie. A: Dredging, B: Hydrocarbon pollution.
109
Figure 72. Heavily impacted faleme cobble bed due to sand mining at 13.001471°; -11.405753°
Figure 73. Photographs illustrating urban pollutants. A: washing of vehicles and clothing; B: urban
runoff/solid waste.
110
Figure 74. Extensive artisanal activities in the Gara River
Fish and Bioaccumulation
The different fish species sampled at shown in Table 21 and the metal contents in
Figure 41, Figure 42, and Figure 43.
The use of the bioaccumulation analysis of metals in fish tissue provides a measure
of conditions that were occurring approximately 12 months prior (long term) to the moment of
sampling (Zhou et al., 2007). Based on this, it can be stated that bioaccumulation analyses
provide tertiary evidence of what the true aquatic conditions are, and can establish which
factors are influencing aquatic conditions prior to and during the moment of sampling.
The bioaccumulation of metals in aquatic biota often allows for the magnification of
elements to detectable limits and therefore allows for the accurate determination of
contaminants, despite potential low concentrations in the ambient environment (Tate and
Husted, 2014). This previous statement was observed when considering the concentrations
of Cd in fish tissue from the Faleme River, which were determined to be highest at the
Faleme Weir followed by the GK Midstream site. Based on this information it can be stated
that Cd concentrations in the ambient environment are such that bioaccumulation occurs to
a greater extent at these sites. Based on sediment analyses presented earlier in this study,
Cd was found to be predominantly below detection limits in the sediment, with exception of
the GK Midstream site. Although Cd was below detection limits in the sediment of the
downstream site (Faleme Weir) it has been determined to be present in elevated
111
concentrations in previous assessments (during different seasons). Therefore, based on
these results it can be stated that Cd concentrations are elevated in the Faleme Weir. This
increased concentration may be due to the effects created by the impoundment and
activities at the Weir, which includes artisanal mining that makes use of batteries and
therefore presents a potential source resulting in increased concentrations of Cd in habitat
utilized by T. zillii and subsequent increased bioaccumulation.
When considering the bioaccumulation trends of As in the Faleme River, the largest
concentrations were observed at Farandie followed by the Faleme Weir and then the
Upstream GK site. When considering the sediment and water quality results of the study
these trends do not directly confirm this at the site, however, when considering previous
studies, as well as the location of the sites, it can be stated that As was confirmed in
elevated concentrations at these sites, being at 252mg/kg at the Upstream GK site and
44mg/kg at the downstream Faleme Weir site during the previous low flow 2012 survey.
Furthermore, the location of the Farandie site immediately downstream of artisanal village
Sasanba, which was shown to emit high concentrations of As dissolved in the water column,
could be responsible for these increased As concentrations as dissolved As would be more
bioavailable than As which is present in the sediment.
Based on this information the following hypothesis can be stated. The concentrations
of As at the site GK Upstream is elevated as a result of upstream artisanal mining activities
during the low flow period, as established in the sediment analysis of the 2012 survey. After
rainfall and subsequent increased flow, sediments are deposited in the pool at the GK
Midstream site where they were determined to be at highly elevated concentrations in this
study. The bioaccumulation of As at the Upstream GK site is therefore higher as fish would
be representing conditions prior to the survey (in the low flow period) with the Midstream GK
fish showing lowered concentrations as sediments had not yet settled into the pools
associated with sites. It is therefore further hypothesized that after the high flow period the
fish in the GK Midstream site would have higher concentrations of As when compared to the
upstream site (Upstream GK) as a result of this increased concentration during the high flow
period.
To conclude, the concentrations of As in fish from Farandie are a result of the
presence of extensive artisanal mining upstream of the site producing increased dissolved
As concentrations (as determined in this study) and therefore a greater component of
bioavailable As resulting in the site having the highest bio-concentration of As when
compared to the other sites. Concentrations of As in fish from the Faleme Weir were
elevated due to the cumulative effects of the impoundment (weir) and the upstream artisanal
activities resulting in the second highest concentrations. Although sediment analyses in this
study determined lowered concentrations of As in the Faleme Weir, it is hypothesized that
recent high flows diluted As concentrations in the sediment and therefore the more
accurate/representative bioaccumulation results would be favored.
112
The bioaccumulation trends of Hg in the Faleme River were found to closely emulate
trends seen in the Cd concentrations, in that the concentrations of Hg in the sites Upstream
GK, Farandie and the Faleme Weir were below detection limits in the sediment.
Bioaccumulation concentrations of Hg were found to be highest at the Faleme Weir, which is
expected due to the large extent of artisanal miners at the site, as observed during the 2014
survey. The second highest concentration was observed at the GK Midstream site with the
metal also being picked up in the Midstream GK sediment possibly indicating rapid uptake
by local T. zillii populations and further confirming aquatic contamination from upstream
mining activities and deposition at the site.
The bioaccumulation trends of Pb showed similar trends as As bioaccumulation
levels and were determined to be highest in the Faleme Weir followed by Farandie and then
the Upstream GK site. Possible explanations for this will be similar as the explanation of As
and Cd concentrations as dissolved Pb was also observed at the site downstream of
Sasanba but is possibly accumulating in fish at the Faleme Weir to a greater extent because
of impoundment effects.
The bioaccumulation concentrations of the metal Cr were found to be highest in the
Faleme Weir and this can be explained through the continued high concentrations of the
metal observed in the Faleme Weir sediments (2010, 2012 and 2014).
9.1.3. Conclusion
Although the mine effluent had a few quality issues (suspended solids with high
arsenic content due to the nature of the rocks, and elevated nitrate concentrations), the
aquatic assessment indicated that it has not had any significant impact on the Faleme River
over the past 4 years of discharge. The current deterioration of the Faleme ecosystem is
associated primarily with the artisanal mining activities, aggregate mining and other
uncontrolled activities by the riverine communities. Going forward it is of paramount
importance that the mine develop an adequate treatment system to isolate the solids and
nitrate from the effluent before discharged as it is believed that continued discharge of the
effluent may have a long term impact on the Faleme locally (i.e. immediately at the
discharge point).
9.2.
Design of the Gara Effluent Treatment System
For an effective design of a treatment system it was necessary to predict the future
discharge volume from the gara underground. A predictive geohydrological model was
constructed for the gara area by Digby Wells Environmental and it was found that an inflow
rate of 2,199.5m³/d, comparable to the average discharge rate of 2,528m³/d, was simulated
as steady state groundwater inflow rate into the modelled decline. An average flow of 2600
m3/day and a maximum peak flow of 3840m3/day is used for the purpose of this
assessment.
113
Digby Wells Environmental recommended a thickener as a measure to remove the
solids from the effluent before discharge and a wetland to remove the nitrate. The rationales
behind these choices are not covered; instead the approaches used to design the two
systems (thickener and wetland) form the basis of the next section.
9.2.1. Removal of solids
Patterson and Cooke (P&C) in Johannesburg was contracted to help with the design of the
thickener. The final water quality objective (Malian discharge limit) is <30 mg/l before
discharge.
The conductivity of the slurry falls within the moderately high range (1.5 to 3 mS/cm). A high
clay size (-5 micron) percentage of more than 30% could indicate high flocculant dose rates.
Settlings slurries could be expected in the presence of clay ores as a result of the double
layer compression effect at slurry conductivity values of 2 mS/cm and above.
It was observed that solids in gara samples coagulated in the natural state. A naturally
coagulated material is defined in layman’s terms as a material that settles overnight (without
any flocculation) and producing a layer of clear supernatant water above a settled bed of
solids. As a result, no further slurry conditioning measures is required and therefore no
flocculation and/or settling problems are expected based on the natural coagulated state of
the thickener feed effluent. But because of the relatively high volume of effluent, it was
necessary to induce a quicker settling rate via flocc addition. The current plant flocculant
(DP6829) performed well for the Gara materials – this is easier as it will save both cost and
time. For this exercise, a reasonable maximum solids concentration of 5%m was selected as
the basis for all further test work.
Due to the irregular nature of the solids in the effluent (see Figure 23), inclusion of an
internal thickener feed-well dilution system in the thickener design is important in order to
deal with possible surges of high solids concentration material that might come through.
Because of the solids quality concerned raised in the characterization of the effluent, the
overflow clarity was the overriding factor that led to the selection of a preliminary flocculant
dose rate of 25 g/t for Gara samples to achieve a minimum settling rate of 20 to 30 m/h and
also ensure acceptable overflow clarity (+40 on clarity wedge).
Underground discharge thickeners are typically sized on rise rate as the critical sizing
parameters due to the low solids concentration expected in the thickener feed. A maximum
rise rate of 4 m/h, corresponding to a solids flux rate of 0.2 t/(m2h), is therefore
recommended for thickener sizing. Based on all the test results and assuming a discharge
rate of 160m3/h below is a table with the minimum parameters.
Process
stream
Critical Rise
Max Flow
Thickening
Thickener
114
Gara
Rate (m/h)
Rate (m3/h)
Area (m2)
Diameter (m)
4
160
40
7
underground
discharge
The thickener is expected to provide a physically good quality water to feed the
constructed wetland for the nutrients and heavy metals removal. A carefully designed and
constructed wetland is able to remove the pollutants in the discharge water. However, from
the discussion, it is evident that wetlands represent complex pollutants removal systems and
their constructions require some knowledge in many disciplines. The hydraulic properties of
the macrophyte zones and climate are the parameters controlling retention time, macrophyte
growth, therefore the rate of pollutant removal – it is therefore of paramount importance to
get them right.
9.2.2. Removal of nitrate and heavy metals
The wetland has been designed to treat high nitrate levels and high dissolved heavy
metals and to protect the wetland should there be a lot of suspended solids entering the
system.
The new mechanical settler will be used for primary removal of solids but there needs
to be a system which will stop solids from entering the wetland if the settlers are not
operational. This will prevent blinding of the wetland.
The clear water will be fed into the constructed wetland which be divided into two
different compartments/zones, the first one will include aerobic and anaerobic zones, with
aerobic zones near the surface and anaerobic zones below. The last one will be an aerobic
zone and being the maturation pond/cascading pond.
The anaerobic part will employ the use of bacteria and macrophytes to provide a
biological treatment activity, where the metals will be removed by the process of
rhizofiltration (removal of pollutants from the contaminated waters by accumulation into plant
biomass). The heavy metals could also form compounds with the ferricrete rocks where
these are placed in the wetland or precipitate as metal sulphides with the sulphides
generated by sulphate reduction in the anaerobic zone.
Plants or microorganisms can assimilate nitrate, or anaerobic bacteria may reduce
nitrate (denitrification) to gaseous nitrogen (N2) when nitrate diffuses into anoxic (oxygen
depleted) water. The gaseous nitrogen volatilizes and the nitrogen is eliminated as a water
pollutant. Thus, the alternating reducing and oxidizing conditions of wetlands completes the
needs of the nitrogen cycle and maximize denitrification rates (Johnston 1991).
115
The wetlands would have to be deep (at least 500mm depth) in order to create an
anoxic/anaerobic environment that is required for the reduction of sulphate to sulphide and
to achieve the required retention time for a biological activity to take place. The produced
sulphide would precipitate the metals and the precipitate would settle out and sediment over
time. The macrophytes planted in the wetland would serve a purpose of metal uptake
through rhizofiltration. Ferns, Pteris vitta commonly known as Braken fern has been
identified as an As hyperacumulator. It can accumulate up to 7500 mg As/kg on a
contaminated site (Ma et al., 2001) without showing toxicity symptoms. The ecological status
of this species in the area would have to be established to ensure it is not a potential
invasive.
The effluent would be aerated through a cascading pond prior to discharge to the
environment. The wetland would require regular but infrequent maintenance over the longterm for clearing the site of weeds, harvesting the reeds, ensuring that the pipes aren’t
clogged and collecting water quality samples. The declining slope would also be useful in
terms of gravitationally feeding the wetland and ensuring that a required retention time is
achieved. Gravity flows ensure that the system is operated cheaply.
De-silting of the settling areas and wetland will be needed dependant on the
effectiveness of the settlers.
Removal of Arsenic will also be influenced by two main mechanisms which are; coprecipitation with Fe-oxides caused by the oxidation of Fe II to Fe III and Mn-oxides, and
adsorption on organic matter. Thus, the As behaviour is governed by redox chemistry of
Fe(III) and Mn(IV) oxides under oxidising conditions. Increases in sediment redox potential
equally increased the affinity of both Fe (III) and Mn (IV) oxides for As, while a decreases in
sediment redox potential results in high affinity between particulate organic matter and As.
Water will then flow to the cascaded pond designed to remove as much residual Mn
and Fe as possible (by aeration at higher pH and allowing the precipitates to settle) prior to
discharging the water into the surface water environment through oxidation cascade to help
with the oxygenation of water.
The oxidation cascade will serve as a post-treatment pond and the following are
proposed requirements of the post-treatment pond:
■
200 – 500mm shallow clay/concrete/plastic lined earth pond;
■
constructed at a slight decline (5 to 15o angle) to facilitate mixing and aeration of the
water;
■
None-reactive stone (+50mm) to facilitate mixing of the flow and to enhance aerobic
conditions; and
■
Monitoring of the water quality must be implemented and the frequency of monitoring
should be adhered to in order to monitor the success of the implemented technology.
116
This could be achieved by monitoring the quality of the underground water after the
settlers and at the wetland outlet.
The use of coarse material and large pipes is imperative to maintain continuous flow. This
recommendation is from experience gained by constructing trail sized wetlands in the
past. Finer sediment and/or thinner pipes will clog or choke and endless frustration will be
experienced.
9.2.3. Conclusion
The design of the solid removal system was straight forward as the thickening
system is a well understood field and involved relatively simple calculation. A 10m diameter
thickener will provide the necessary surface area to remove the solids from the gara effluent
before entering the wetland. Designing a horizontal flow subsurface constructed wetland,
however, revealed to be very complex as many physical, biological and chemical processes
are involved in the system. The research has allowed an understanding of the different
processes and because of the complexity of their interaction a broad design has been
proposed with a set of designs to be trial tested (real life situation) in order to select the most
effective pollutants system. Figure 75 below shows the design of the proposed treatment
integrated system and the different wetland designs to be trial tested:
Figure 75. Proposed surface treatment system (Source: DWA, 2014)
A summary of requirements for each design trial is shown in table below
117
Design requirements
Trial 1
The secondary compartment is filled with a later of waste rock of >200mm as
a base layer. Large rock is used to facilitate subsurface flow. Previous trials
at Western Areas Gold Mine have shown that consistent flow is a challenge
if uniform finer material is used.
The layer above the rock is gravel (<50 mm), the middle layer will be
ferricrete layer with fragmented ferricrete blocks >100mm. Again the size is
chosen to facilitate subsurface flow and maintain anaerobic conditions. The
ferricrete will increase surface area for the bacteria to establish and as a
secondary benefit facilitate arsenic removal.
The surface layer will be waste rock with a minimum diameter 200 mm or
similar material which is rounded and allows maximum interstitial space to be
maintained.
Hydrophytes are maintained as a floating mass on wooden/reed/bamboo
frames. Roots may enter the scat size layer.
Trial 2
will vary with the first trial by excluding the gravel and only use ferricrete
layer with fragmented ferricrete blocks >100mm as well as the surface layer
which will be waste rock with a minimum diameter 200 mm or similar
material which is rounded and allows maximum interstitial space to be
maintained.
Trial 3
this will only include gravel (<50 mm) and the surface layer will be waste
rock with a minimum diameter 200 mm or similar material which is rounded
and allows maximum interstitial space to be maintained. The hydrophytes
are retained.
Trial 4
This will consist of the Gravel layer above (<50 mm), the middle layer will be
ferricrete layer with fragmented ferricrete blocks >100mm. and the
hydrophyte layer is retained. This should give a limited retention time and
good hydrological performance.
Trial 5
This trial will have similar porous material as trial 1 but will not consist of any
plants or hydrophytes. The compartment is filled with a later of waste rock of
>200mm as a base layer.
The layer above the rock is gravel (<50 mm), the middle layer will be
ferricrete layer with fragmented ferricrete blocks >100mm. Again the size is
chosen to facilitate subsurface flow and maintain anaerobic conditions. The
ferricrete will increase surface area for the bacteria to establish and as a
secondary benefit facilitate arsenic removal
118
Size of the A surface size of 10 m wide by 30 m long is suggested. This is because the
pilot
final expected size is 100m by 300m to deal with the full flow. The base
needs to be sealed so that it is not permeable. This could be via concrete,
clay or plastic lining. Concrete will enable for a more robust floor if rock
needs to be changed or items need to be replaced.
Monitoring
A redox probe needs to be inserted into each wetland to determine the redox
potential at different heights to determine regimes in this regard as
biochemical reactions are dependant and driven by this variable.
The pond exit should be such that the height of the pond could be managed
to ensure full coverage of the surface area.
Additionally carbon could be added in the form of leaves, sewage plant
water, hay, manure, compost etc to increase ammonifying bacteria to assist
in the nitrogen cycling. Various plants like papyrus and or other hydrophytes
readily available can be used to determine effectiveness of nitrogen removal
compared to the papyrus.
119
120
Figure 76. Examples of the trials that may be undertaken as iterative process for optimisation of the
constructed wetland (Source: DWA, 2014)
Additionally trials will be used to see if crops can be produced without accumulating
unwanted elements or compounds. Rice and/or Cassava for example can be included in the
trails. Additionally a final polishing pond can be added and fish farming can be tested,
specifically if algae populations can be maintained as a fish fodder.
121
Experience has shown that simulation results match the measured data when the
hydraulic behaviour of the system can be described well. A good match of experimental data
to reactive transport simulations can then be obtained using literature values for the model
parameters (Langergraber, 2003; Langergraber and Šimůnek, 2005). For practical
applications, it is advisable to measure at least the porosity and saturated hydraulic
conductivity of the filter material to obtain reasonable simulation results for water flow.
10 Recommendations
Fish metal bioaccumulation assessment
Due to the presence of potentially toxic metal elements within the fish associated with
the project area a requirement of a human health risk assessment arises. Due to the low
confidence in standalone threshold comparisons metal concentrations will be used in human
health risk indices such as the (du Preez et al., 2003). This work is currently not complete
owing to the fact that particular threshold effect metal concentrations need to be confirmed,
however some preliminary findings suggest the following.
 The Faleme River has As, Co, Cu, Pb and Hg which show Hazard Quotients over 1.0
as standalone values (not a mixture) with As concentrations showing a level that is
13x its recommended safety level. Further, As concentrations present a high risk to
consumers and show a 3 in 1000 chance of generating cancer in consumers.
 Similarly in the Gara River, As, Hg and Pb all exhibit Hazard Quotients higher than 1.0
with As 15 times above the recommended safety level presenting a 3 in 1000 chance
of acquiring cancer and therefore presenting a high risk to consumers.
 As stated above it has been recommended that a full Human Health Risk Assessment
(HHRA) be undertaken. Through the completion of this assessment safe
consumption guidelines could be established to limit the potential harmful effects of
metal toxicity. The model used to generate the above information is largely adapted
from the following reference du Preez et al. (2003; 2004).
This has not been completed as yet and the results will be presented to Randgold at
a later stage.
Eflluent Solids removal system
Table below presents a more conservative approach were the thickener sizes are
rounded up to the next available standard size, which thereby incorporates a large safety
factor. This approach is preferred due to the following reasons:

Thickener sizes are small and therefore cost implications of going larger
should not be prohibitive
122


The larger thickening area for the Yalea material would in all probability sort
out the somewhat poor overflow clarity
Operational problems around surges of high solids concentration slurries or
larger volumetric flow rates should be reduced due to the larger available
thickening area.
Table 28. Recommended thickener size (P&C, 2011)
Process
stream
Gara
Critical Rise
Max Flow
Thickening
Thickener
Rate (m/h)
Rate (m3/h)
Area (m2)
Diameter (m)
2.0
160
79
10
underground
discharge
Nitrate removal system – constructed wetland
It is recommended that the constructed wetland systems will consist of five trial pilot
scale cells of HSSF-CW receiving primary discharge from the underground mine. Examples
of the trials are illustrated Figure 76, where an iterative process should be used to optimise
the constructed wetland. The overall objective of the trials will be to establish under which
materials the process of denitrification will perform best. In pilots holding basins or lagoons,
wastewater discharge is stabilized in a confined environment. Area requirements will be
based on volumetric and surface loading rates. In the of the HSSF-CW cells, the macrophyte
Cyperus papyrus will planted and it may be further compared against test cell, which may be
planted with other appropriate vegetation.
The area determination is initially based on a rule of thumb or first order design
equations. This assumption will be checked during the operation of the trial wetlands to
ensure that they operate sufficiently. In this case a primary assumption is that the rate of
nitrogen removal is 2 g/m2/day. We need to be able to control the rate at which the water
enters our trial wetlands and to be able to measure this so that we can calculate rates of
removal. We have assumed that the settlers will operate sufficiently for there not to be too
many solids entering the system to clog it up but we have designed it so that some solids will
be removed upfront to protect the wetland in case where the settlers are not functioning
optimally. The trials further aim to prevent the clogging of the macrophyte material.
The biological conditions in these pilot systems will perform under similar conditions
of the wetland; water near the bottom is in an anoxic/anaerobic state, while a shallow zone
near the water surface tends to be aerobic. The source of oxygen is planned to be
atmospheric reaeration, photosynthesis (in facultative pond) and root oxygen leaching. The
123
treatment for the nitrate will be dealt with by the development of microorganisms on the plant
roots and the solid medium. There will be no need to input chemicals or external energy, as
treatment mechanisms are natural and wastewater flow within the systems is driven by
gravity.
In the of the HSSF-CW cells, a pair of the cells will use the macrophyte Typha which
already exist on site and vary the material in the macrophyte bed to ensure that results are
comparable and establish the most effective material in achieving the set objectives.
Figure 77. Examples of pilot scale constructed wetlands (Source: Vymazal, 2009)
All trial are set to use the same plant species (typha/pyparus) which is available on
site; the variation will only be on the macrophyte material, this will ensure the establishment
of the most effective materials when comparing the results within the trials.
A sub surface flow inlet into a pre-impoundment will allow the remaining suspended
solids or sediment to settle out and it will ensures anaerobic conditions as water is forced
through the substrate.
.
124
Appendix A – Boreholes Logs
125
Appendix B – Gara Discharge water quality
Site
Nam
e
DateTi
meMea
s
p
H
IFC
69
Mali
an
6.
59.
5
7.
7
0
7.
2
0
7.
7
0
7.
3
0
7.
0
0
7.
3
0
7.
3
0
7.
0
0
7.
2
0
7.
1
0
6.
6
0
7.
0
0
6.
7
0
7.
1
0
7.
1
0
6.
8
0
6.
2
0
2010/0
7/01
00:00
2010/0
8/01
00:00
2010/0
9/01
00:00
2010/0
9/02
00:00
2010/0
9/03
00:00
2010/0
9/04
00:00
2010/0
9/05
00:00
2010/0
9/06
00:00
2010/0
9/07
00:00
2010/0
9/08
00:00
2010/0
9/09
00:00
2010/0
9/10
00:00
2010/0
9/11
00:00
2011/0
1/03
00:00
2011/1
0/01
00:00
2011/1
2/01
00:00
2012/0
1/01
00:00
EC
mS
/m
TD
S
mg/
l
Ca
mg
/l
M
g
m
g/l
Na
m
g/l
K
m
g/l
MA
LK
mg/
l
Cl
m
g/l
SO
4
mg
/l
Nit
rat
eN
mg
/l
F
m
g/
l
Al
m
g/l
Fe
m
g/l
M
n
m
g/l
Am
oni
a
mg/
l
N
O
2N
m
g/l
2
25
0
13
5.0
0
69.
80
850
.00
50.
00
440
.00
69.
30
98.
30
528
.00
66.
40
400
.00
10
4.0
0
87.
20
12
4.0
0
67.
20
958
.00
468
.00
68.
30
83.
00
590
.00
90.
00
97.
80
490
.00
69.
00
48.
30
294
.00
57.
30
10
4.0
0
86.
50
582
.00
95.
90
578
.00
91.
60
614
.00
11
7.0
0
11
7.0
0
17
2.0
0
14
8.0
0
732
.00
634
.00
20
0.0
0
16
8.0
0
15
5.0
0
11
8.0
0
12
0.0
0
12
0.0
0
78.
00
108
0.0
0
10
8.0
0
564
.00
732
.00
24
.7
0
31
.8
0
34
.2
0
33
.5
0
28
.3
0
26
.2
0
22
.1
0
29
.2
0
58
.9
0
51
.1
0
54
.8
0
32
.9
0
41
.7
0
41
.7
0
27
.4
0
57
.5
0
12
00
10
00
30
40
6.0
0
51.
30
26
.0
0
26
.0
0
40
.0
0
38
.0
0
57
.0
0
48
.0
0
15
.0
0
40
.0
0
6.
00
8.
70
216
.00
3.
00
26.
00
3.
40
177
.00
2.
30
5.
00
236
.00
4.
30
3.
50
201
.00
0.
40
5.
50
210
.00
0.
90
2.
80
286
.00
1.
70
11
3.0
0
19
6.0
0
11
3.0
0
16
9.0
0
89.
00
4.
30
202
.00
2.
30
4.
90
213
.00
3.
10
0.
90
232
.00
0.
10
59
.1
0
59
.7
0
9.
80
15
.9
0
10
.3
0
8.
90
228
.00
6.
80
169
.00
6.
70
257
.00
11
.9
0
34
.0
0
52
.0
0
52
.0
0
53
.0
0
55
.0
0
6.
80
185
.00
9.
10
9.
10
190
.00
8.
00
213
.00
17
.3
0
8.0
0
12
.3
0
12
.3
0
6.
90
11
.3
0
17
9.0
0
18
1.0
0
Ni
m
g/l
H
g
m
g/l
As
m
g/l
50
0.0
5
0.
30
0.
50
0.
10
30
0.0
2
0.
10
2.
00
0.
00
2
0.
00
5
227.00
0.0
0
0.
02
0.
01
0.
00
0.
01
304.00
0.0
0
0.
01
0.
01
402.00
0.0
0
0.
03
0.
02
0.0
0
0.
01
0.
01
409.00
0.0
0
0.
01
287.00
0.0
0
332.00
2
1.
80
1.
10
0.
03
6.
08
90.00
1.
12
1.
00
0.
04
55.00
64.
00
8.
50
8.
70
0.
12
14.
00
25
.6
0
3.
33
6.7
5
5.
34
4.
90
0.
08
0.0
1
0.
22
140.00
5.
38
6.
60
0.
20
0.
87
565.00
0.
02
0.
05
0.
01
0.
03
44.
70
7.
26
7.
90
0.
11
0.
03
350.00
86.
90
7.
80
9.
80
0.
26
0.
03
495.00
0.
01
0.
00
0.
03
0.
21
0.
03
605.00
0.
31
18
.2
0
2.
38
550.00
735.00
0.
5
1
0.2
5
19
0.0
0
19
4.0
0
23
0.0
0
22
6.0
0
26
3.0
0
26
3.0
0
22
1.0
0
28
7.0
0
C
u
m
g/l
2
0.
5
4
18
0.0
0
83.
50
11
.5
0
10
.7
0
15
Cd
mg/
l
1
20
4.0
0
0.1
0
6
TotalSuspe
ndedSolids
230.00
80.
00
3.
04
14
.1
0
13
.1
0
21
.0
0
3.
50
18
0.0
0
18
0.0
0
63.
20
7.
77
8.
50
0.
18
17
.8
0
0.
03
7.
77
8.
50
0.
18
0.
03
343.00
12
.3
0
0.
41
14
.9
0
0.
80
0.
28
0.
03
759.00
66
.1
0
18.00
99.
00
52
0.0
0
0.
34
0.
12
0.
04
10.
70
965.00
343.00
P
O
4
m
g/
l
0.
0
1
0.
0
1
0.
0
1
0.
0
1
0.
0
1
0.
0
1
0.
3
3
0.
0
1
0.
0
3
0.
0
1
0.
0
1
0.
0
1
0.
0
1
0.
0
1
0.
0
1
0.
0
1
0.
3
2
TotalHa
rdnessCaCO3
mg/l
C
r6
m
g/
l
Zn
m
g/l
Pb
mg
/l
D
B
O
CO
D
Oil
&
Gre
ase
0.
1
0
0.
2
0
0.
50
0.2
0
0.2
0
15
0.0
0
15
0.0
0
10.
00
0.
50
50
.0
0
50
.0
0
0.
04
0.
07
0.0
1
2.
50
0.0
5
0.
01
0.
02
0.
08
0.0
1
7.
00
18
0.0
0
12.
50
0.
02
0.
03
0.
03
0.0
1
2.
50
51.
00
0.0
5
0.
01
0.
02
0.
06
0.0
1
2.
50
31.
00
0.0
5
0.
01
0.
03
0.
06
0.0
1
11
5.0
0
0.0
5
0.
01
0.
01
0.
01
0.
03
0.0
1
13
.0
0
2.
50
0.0
5
0.
02
0.
01
0.
01
0.
01
0.
03
0.0
1
16
.0
0
89.
00
264.00
0.0
0
0.
03
0.
01
0.
01
0.
04
0.
10
0.0
1
2.5
0
263.00
0.0
0
0.
00
0.
00
0.
00
0.
02
0.0
0
2.5
0
0.0
0
0.
01
0.
03
0.
02
0.
04
0.
12
0.0
0
0.0
0
0.
01
0.
02
0.
02
0.
02
0.
14
0.0
1
2.50
0.0
0
0.
02
0.
04
0.
02
0.
05
0.
09
0.0
1
2.
50
431.00
0.0
0
0.
01
0.
01
0.
01
0.
01
0.
03
0.0
1
2.
50
0.0
0
0.
01
0.
01
0.
02
0.
02
0.
18
0.0
1
472.00
0.0
0
0.
01
0.
01
0.
02
0.
02
0.
18
308.00
0.0
0
0.
03
0.
03
0.
02
0.
03
505.00
0.0
0
0.
02
0.
01
0.
11
0.
01
0.
00
0.
00
Cr
m
g/l
0.
50
20.
00
2.5
0
C
NFr
e
e
m
g/
l
0.
1
0
0.
5
0
C
NTo
tal
m
g/l
W
c
n
m
g/
l
1.
00
0.
5
0
0.
0
3
0.
0
4
0.
09
0.
0
1
0.
0
1
0.
0
1
0.
01
0.
0
1
0.
01
0.
0
1
1.
00
0.
07
0.
0
3
0.
0
4
2.5
0
44.
00
2.5
0
2.
50
12.
50
2.5
0
0.0
1
2.
50
12.
50
2.5
0
0.
05
0.0
1
2.
50
11
0.0
0
2.5
0
0.
09
0.0
1
2.
50
2.5
0
126
Site
Nam
e
DateTi
meMea
s
p
H
IFC
69
Mali
an
6.
59.
5
6.
9
0
6.
7
0
6.
8
0
6.
7
2012/0
2/01
00:00
2012/0
3/01
00:00
2012/0
4/01
00:00
2012/0
5/01
00:00
2012/0
6/01
00:00
2012/0
8/01
00:00
2012/0
7/01
00:00
2012/0
9/01
00:00
2012/1
0/01
00:00
2012/1
1/01
00:00
2013/0
1/01
00:00
2013/0
2/01
00:00
2013/0
3/01
00:00
2013/0
4/01
00:00
2013/0
5/01
00:00
2013/0
6/01
00:00
2013/1
1/01
00:00
2013/1
2/01
00:00
EC
mS
/m
TD
S
mg/
l
Ca
mg
/l
M
g
m
g/l
Na
m
g/l
K
m
g/l
MA
LK
mg/
l
Cl
m
g/l
SO
4
mg
/l
Nit
rat
eN
mg
/l
F
m
g/
l
Al
m
g/l
Fe
m
g/l
M
n
m
g/l
Am
oni
a
mg/
l
N
O
2N
m
g/l
2
25
0
10
0.0
0
88.
00
670
.00
90.
60
672
.00
11
1.
6.
7
12
00
10
00
30
58.
40
6
15
TotalSuspe
ndedSolids
Cd
mg/
l
C
u
m
g/l
Ni
m
g/l
H
g
m
g/l
As
m
g/l
50
0.0
5
0.
30
0.
50
0.
10
30
0.0
2
0.
10
2.
00
0.
00
2
0.
00
5
655.00
0.0
0
0.
05
0.
05
612.00
0.0
0
0.
03
0.
03
0.
12
1140.00
0.0
0
0.
07
0.
08
0.
09
738.
0.0
005
891.
0.0
01
0.
01
5
0.
04
0.
03
3
0.
01
537.
0.0
01
0.
04
0.
03
371.
0.0
5
0.
04
0.
03
805.
0.0
32
0.
12
0.
1
852.
0.0
01
0.
07
437.
0.0
01
1040.
1
2
2
22
.2
0
15
.8
0
47
.5
0
12
0.
50
7.
88
1240.00
0.
45
3.
63
91.00
1.
30
0.
96
9290.00
14
2.
26
.6
0
12
.1
0
36
.2
0
7.
3
0.
68
13.
9
6.
9
1660
19.
1
23
.4
3220
3.
51
892
60
.1
0
49
.5
0
77
.7
0
60
.3
50
.0
0
50
.0
0
54
.3
0
59
.1
10
.6
0
8.
00
228
.00
9.
40
182
.00
6.
20
12
.4
0
11
.1
229
0.0
0
296
.
8.
30
558
.
16
3.0
0
16
4.0
0
33
0.0
0
19
6
10
.3
23
8.0
0
23
7.0
0
21
0.0
0
23
8.
13
0.
760
.
25
0.
64
.8
14
3.
14
.4
340
.
17
.4
30
8.
20
6.
4.
77
7.
5
0.
81
6.
8
11
8.
680
.
14
0.
45
.9
56
.4
9.
5
181
.
14
.
29
1.
16
4.
7.
71
9.
1
0.
49
6.
5
98.
1
582
.
10
2.
28
.1
45
.5
6.
2
162
.
8.
5
23
1.
10
2.
9.
1
16
.
0.
38
1.8
6
26
.8
1110
6.
5
12
1.
740
.
18
9.
80
.8
56
.9
11
.3
151
.
10
.4
25
4.
14
4.
31
.5
34
.5
0.
91
3.7
4
23
.
2700
6.
8
10
5.
764
.
22
2.
72
.4
51
.5
10
.7
413
.
8.
9
25
1.
13
4.
22
.7
33
.3
0.
74
0.0
1
2240
6.
8
10
7.
744
.
11
2.
38
.1
51
.9
8.
3
179
.
10
.2
25
5.
14
0.
2.
25
2.
8
0.
08
0.
02
5
33
.2
7.
2
14
9.
892
.
23
6.
11
0.
71
.3
17
.2
238
.
18
.4
31
3.
25
6.
34
.
45
.5
1.
01
30.
1
11
.4
1360
7.
2
13
8.
900
.
19
3.
74
.2
67
.3
14
.6
202
.
17
.4
31
9.
21
6.
15
.5
24
.7
0.
61
18.
16
.7
1060
7.
14
2.
848
.
20
1.
83
.2
71
.7
19
.1
213
.
19
.2
30
5.
41
6.
23
.7
32
.4
0.
5
18
.9
1860
7.
1
11
3.
752
.
11
9.
9.
2
55
.9
9.
2
149
.
12
.9
27
3.
14
2.
7.
44
13
.5
0.
41
11
.2
611
6.
8
14
3.0
102
0.0
54
3.0
30
.9
69
.5
22
.9
457
.0
19
.5
34
6.0
22
6.0
51
.3
11
0
3.
07
24
9870
7.
2
11
2.0
722
.0
13
3.0
36
.2
57
.7
13
.9
309
.0
13
.0
25
1.0
13
0.0
14
.3
30
.2
1.
14
21
.4
2310
7.
2
10
2.0
710
.0
15
5.0
45
.9
59
.0
10
.4
167
.0
12
.8
31
8.0
15
2.0
12
.2
33
.4
0.
69
2230
7.
0
17
6.0
119
0.0
15
0.0
39
.9
81
.7
15
.2
210
.0
22
.5
47
1.0
37
2.0
38
.7
87
.6
2.
38
0.
02
5
25
.4
606
.00
49.
20
92.
50
6.6
3
355
2750
P
O
4
m
g/
l
0.
1
0
0.
0
1
0.
0
2
0.
0
1
0.
0
4
0.
5
5
0.
0
3
0.
0
1
0.
0
1
0.
1
3
0.
0
4
0.
0
1
0.
0
7
0.
0
3
0.
0
8
0.
0
4
0.
0
3
0.
0
1
TotalHa
rdnessCaCO3
mg/l
Cr
m
g/l
0.
50
0.
03
C
r6
m
g/
l
Zn
m
g/l
Pb
mg
/l
D
B
O
CO
D
Oil
&
Gre
ase
0.
1
0
0.
2
0
0.
50
0.2
0
0.2
0
15
0.0
0
15
0.0
0
10.
00
0.
50
50
.0
0
50
.0
0
0.
21
0.0
1
0.
06
2.
50
61.
00
0.
20
0.0
1
0.
03
8
0.
01
0.
13
0.0
06
10
.0
0
14
.
60
5.0
0
11
5.
0.
08
0.0
05
10
.
11
0.
0.
04
0.
07
0.0
05
0.
04
0.
06
0.0
05
2.
5
0.
11
0.
23
0.0
2
2.
5
0.
05
0.
07
0.
11
0.0
05
9.
21
0.
0.
02
0.
01
5.
43.
0.
15
0.
1
0.
02
5
0.
15
0.0
05
0.0
01
0.
00
5
0.
1
787.
0.0
01
0.
06
0.
04
0.
04
0.
16
0.0
1
843.
0.0
01
0.
06
0.
04
0.
05
0.
09
0.0
05
1050.
0.0
01
0.
01
0.
03
0.
03
0.
09
0.0
05
12
.
12
0.
1480
0.0
04
0.
23
0.
17
0.
15
0.
33
0.0
1
930
0.0
01
0.
08
0.
05
0.
04
0.
19
0.0
1
9
20
5
575
0.0
000
5
0.0
01
0.
04
4
0.
22
0.
05
0.
12
8
22
5
0.
26
0.
07
8
0.
39
0.0
07
0.
13
0.
04
8
0.
12
538
0.
04
4
0.
04
2
C
NTo
tal
m
g/l
W
c
n
m
g/
l
1.
00
0.
5
0
1.
00
2.5
0
0.0
1
0.
02
5
0.
12
0.
0
3
20.
00
C
NFr
e
e
m
g/
l
0.
1
0
0.
5
0
2.5
0
12.
5
33.
0.0
05
0.0
05
127
Site
Nam
e
DateTi
meMea
s
p
H
EC
mS
/m
TD
S
mg/
l
Ca
mg
/l
M
g
m
g/l
Na
m
g/l
K
m
g/l
IFC
69
Mali
an
6.
59.
5
6.
7
25
0
22
2
155
0
21
0
55
.8
12
2
24
.1
7
86.
5
606
81
25
45
.9
6.
6
11
6.
770
.
92.
37
.5
6.
9
17
1.
123
0.
15
5.
7.
2
11
4.
810
.
7.
2
19
5.
7.
5
13
3.
2014/0
1/01
00:00
2014/0
2/01
00:00
2014/0
3/01
00:00
2014/0
4/01
00:00
2014/0
5/01
00:00
2014/0
6/01
00:00
2014/0
7/01
00:00
MA
LK
mg/
l
Cl
m
g/l
SO
4
mg
/l
Nit
rat
eN
mg
/l
F
m
g/
l
Al
m
g/l
Fe
m
g/l
M
n
m
g/l
Am
oni
a
mg/
l
N
O
2N
m
g/l
2
12
00
10
00
30
83
55
.7
59
8
50
8
5
177
9.
9
24
2
41.
5
57
.1
7.
149
.
17
.5
29
0.
10
8.
52
.6
10
6.
13
.9
155
.
37
.5
46
8.
25
2.
11
8.
42
.5
68
.2
10
.1
154
.
19
.9
25
9.
10
0.
134
0.
14
3.
42
.4
12
2.
34
.4
93.
64
.7
27
5.
28
8.
970
.
11
2.
39
.4
10
0.
18
.3
191
.
37
.2
29
9.
10
6.
6
1
2
2
15
4.
79
8.
1
0.
16
62.
8
2
2.
7
4.
2
3.
16
5.
3
TotalSuspe
ndedSolids
Cd
mg/
l
C
u
m
g/l
Ni
m
g/l
H
g
m
g/l
As
m
g/l
50
0.0
5
0.
30
0.
50
0.
10
30
0.0
2
0.
10
2.
00
0.
00
2
0.
00
5
0.0
000
5
0.0
01
0.
01
1
0.
01
0.
02
385.
0.0
001
603.
0.0
003
0.
01
1
0.
00
4
0.
01
4
0.
00
9
0.
01
2
0.
02
4
0.
03
32
.9
421
0.
04
3.
61
98
18
.
75
5.
75
112
7.
9
0.
07
9
0.
03
6
0.
13
8.
7
243
3.
51
5.
8
0.
09
1.3
7
14
.4
202
8.
11
11
.8
0.
16
5.6
5
4.
56
655
P
O
4
m
g/
l
0.
0
1
0.
0
3
0.
0
4
0.
0
1
0.
0
1
0.
0
5
0.
0
1
TotalHa
rdnessCaCO3
mg/l
755
305
470.
532.
0.0
001
442.
0.0
001
0.
01
0.
02
7
0.
02
3
Cr
m
g/l
C
r6
m
g/
l
Zn
m
g/l
Pb
mg
/l
D
B
O
CO
D
Oil
&
Gre
ase
0.
1
0
0.
2
0
0.
50
0.2
0
0.2
0
15
0.0
0
15
0.0
0
10.
00
0.
50
50
.0
0
50
.0
0
0.
02
8
0.
02
5
0.
04
8
0.
01
9
0.
07
0.0
01
3
0.0
05
2.
5
37.
0.
02
1
0.
01
4
0.
00
5
0.
01
1
0.
00
5
0.
02
1
0.
06
9
0.0
01
2
0.0
01
4
22
.
74.
0.
03
3
0.
01
4
0.
01
5
35
.
70.
0.
50
0.
02
4
0.
02
2
0.0
02
20.
00
C
NFr
e
e
m
g/
l
0.
1
0
0.
5
0
C
NTo
tal
m
g/l
W
c
n
m
g/
l
1.
00
0.
5
0
1.
00
0.0
00
3
128
Appendix C - STATIC SEDIMENTATION TESTS
Appendix C.1. Physical properties of the effluents
PCCS Number
Yalea
Gara
Conductivity
(mS/cm)
2.040
1.600
pH
8.27
8.00
Solids SG
(g/cm 3)
2.749
2.716
Liquid SG
(g/cm 3)
1.000
1.000
Appendix C.2. settling test results
PCCS Number
Yalea
Gara
PCCS Number
Yalea
Gara
PCCS Number
Yalea
Gara
Sample Number
Yalea
Gara
Clarity
4.0
2.0
M10
Settling distance
Settling time
(mm)
(s)
50
8.13
50
28.10
Settling rate
(m/h)
22.14
6.41
Clarity
4.0
3.0
M336
Settling distance
Settling time
(mm)
(s)
50
6.20
50
23.61
Settling rate
(m/h)
29.03
7.62
Clarity
4.0
3.0
M156
Settling distance
Settling time
(mm)
(s)
50
6.89
50
23.21
Settling rate
(m/h)
26.12
7.76
Optimum
flocculant type
M6260
DP8629
Clarity
4.0
3.0
M919
Settling distance Settling time
(mm)
(s)
50
6.99
50
23.54
Settling rate
(m/h)
25.75
7.65
Clarity
5.0
4.0
M6260
Settling distance Settling time
(mm)
(s)
50
4.65
50
20.00
Settling rate
(m/h)
38.71
9.00
Clarity
4.0
4.0
DP8629
Settling distance Settling time
(mm)
(s)
50
5.43
50
15.00
Settling rate
(m/h)
33.15
12.00
Secondary
flocculant type
DP8629
M6260
129
Process water added (ml)
232
Target slurry volume (ml)
250
Target slurry mass (g)
254.0
Target slurry density (g/cm 3)
1.016
Floc volume added (ml)
0.4
Settled distance (mm)
232
Settling time (s)
20.00
Settling rate (m/h)
41.76
Appendix
C.3. Thickener
Clarity
49.0feed
Sample Number
Gara
Stock slurry solids conc.
24.97%
Parameters
2.5
Floc dose (g/t)
Stock slurry volume (ml)
Stock slurry mass (g)
Stock slurry density (g/cm 3)
Process water added (ml)
Target slurry volume (ml)
Target slurry mass (g)
Target slurry density (g/cm 3)
Floc volume added (ml)
Settled distance (mm)
Settling time (s)
Settling rate (m/h)
Clarity
21
25.4
1.187
229
250
254.0
1.016
0.5
218
20.00
39.24
49.0
Sample Number
Yalea
Gara
Prelim slurry
solids conc
5.0
5.0
214
250
258.2
1.033
0.8
192
20.00
34.56
concentration
49.0
195
175
250
250
262.5
267.0
1.050
1.068
1.2
1.6
160
98
20.00
20.00
28.80
17.64
determination
49.0
30.0
134
250
276.4
1.106
2.5
30
20.00
5.40
20.0
m% solids
Slurry % solids
7.5
10.0
20
44
66
90
51.7
78.8
106.9
1.187
1.187
1.187
206
184
160
250
250
250
258.2
262.4
266.9
1.033
1.050
1.067
1.0
1.6
2.1
184
120
90
20.00
20.00
30.00
33.12
21.60
10.80
49.0
49.0
49.0
Prelim floc
Sample Number
dose (g/t)
Yalea
15.0
Gara
15.0
5.0
15.0
140
165.9
1.187
110
250
276.2
1.105
3.3
25
30.00
3.00
49.0
130
Sample Number
Gara
Stock slurry solids conc.
24.97%
Parameters
Optimum feed solids m%
Stock slurry volume (ml)
Stock slurry mass (g)
Stock slurry density (g/cm 3)
Process water added (ml)
Target slurry volume (ml)
Target slurry mass (g)
Target slurry density (g/cm 3)
Floc volume added (ml)
Settled distance (mm)
Settling time (s)
Settling rate (m/h)
Clarity
m% solids
10.0
15.0
44
51.7
1.187
206
250
258.2
1.033
0.5
100
30.00
12.00
30.0
44
51.7
1.187
206
250
258.2
1.033
0.8
170
20.00
30.60
49.0
Floc dose (g/t)
20.0
5
44
51.7
1.187
206
250
258.2
1.033
1.0
190
20.00
34.20
49.0
25.0
30.0
44
51.7
1.187
206
250
258.2
1.033
1.3
198
20.00
35.64
49.0
44
51.7
1.187
206
250
258.2
1.033
1.5
198
20.00
35.64
49.0
131
Sample Number
Yalea
Slurry % solids
5.00
Slurry make-up
Stock slurry solids conc. (%m)
Stock slurry volume (ml)
Stock slurry mass (g)
Stock slurry density (g/cm 3)
Process water added (ml)
Target slurry volume (ml)
Target slurry mass (g)
Target slurry density (g/cm 3)
29.07%
145
177.6
1.227
855
1000
1032.9
1.033
Base case
+50% floc
-50% floc
+50% solids
-50% solids
Floc dose (g/t) Floc needed (ml) Floc dose (g/t) Floc needed (ml) Floc dose (g/t) Floc needed (ml) Slurry % solids Floc needed (ml) Slurry % solids Floc needed (ml)
15.00
3.1
22.50
4.6
7.50
1.5
7.50
4.7
2.50
1.5
Distance (mm)
Distance (m)
Distance (mm)
Distance (m)
Distance (mm)
Distance (m)
Distance (mm)
Distance (m)
Distance (mm)
Distance (m)
0
0.000
0
0.000
0
0.000
0
0.000
0
0.000
72
-0.072
106
-0.106
30
-0.030
63
-0.063
105
-0.105
175
-0.175
193
-0.193
54
-0.054
115
-0.115
197
-0.197
278
-0.278
262
-0.262
81
-0.081
173
-0.173
294
-0.294
290
-0.290
273
-0.273
106
-0.106
225
-0.225
316
-0.316
296
-0.296
280
-0.280
123
-0.123
245
-0.245
319
-0.319
300
-0.300
287
-0.287
140
-0.140
257
-0.257
322
-0.322
310
-0.310
298
-0.298
240
-0.240
277
-0.277
326
-0.326
317
-0.317
308
-0.308
300
-0.300
295
-0.295
328
-0.328
322
-0.322
313
-0.313
310
-0.310
300
-0.300
331
-0.331
325
-0.325
315
-0.315
314
-0.314
304
-0.304
331
-0.331
327
-0.327
318
-0.318
320
-0.320
310
-0.310
331
-0.331
327
-0.327
321
-0.321
323
-0.323
310
-0.310
331
-0.331
Settling rate (m/h)
37.08
Settling rate (m/h)
28.08
Settling rate (m/h)
9.18
Settling rate (m/h)
19.80
Settling rate (m/h)
34.02
Clarity
49
Clarity
49
Clarity
15
Clarity
49
Clarity
49
Time (s)
0
10
20
30
40
50
60
120
300
600
900
1800
3600
+50% floc dose
-50% floc dose
+50% solids
-50% solids
0.0
-0.1
-0.1
-0.1
-0.1
-0.1
-0.2
-0.3
-0.2
-0.3
-0.4
-0.3
-0.4
0
450
900
1350
Time (s)
1800
2250
2700
-0.2
450
900
1350 1800 2250 2700
Time (s)
-0.2
-0.3
-0.4
0
Distance (m)
0.0
Distance (m)
0.0
Distance (m)
0.0
Distance (m)
Distance (m)
Base case
0.0
-0.3
-0.4
0
450
900
1350 1800 2250 2700
Time (s)
-0.2
-0.4
0
450
900
1350 1800 2250 2700
Time (s)
0
450
900
1350 1800 2250 2700
Time (s)
132
Sample Number
Gara
Slurry % solids
5.00
Slurry make-up
Stock slurry solids conc. (%m)
Stock slurry volume (ml)
Stock slurry mass (g)
Stock slurry density (g/cm 3)
Process water added (ml)
Target slurry volume (ml)
Target slurry mass (g)
Target slurry density (g/cm 3)
24.97%
174
206.8
1.187
826
1000
1032.6
1.033
Base case
+50% floc
-50% floc
+50% solids
-50% solids
Floc dose (g/t) Floc needed (ml) Floc dose (g/t) Floc needed (ml) Floc dose (g/t) Floc needed (ml) Slurry % solids Floc needed (ml) Slurry % solids Floc needed (ml)
15.00
3.1
22.50
4.6
7.50
1.5
7.50
4.7
2.50
1.5
Distance (mm)
Distance (m)
Distance (mm)
Distance (m)
Distance (mm)
Distance (m)
Distance (mm)
Distance (m)
Distance (mm)
Distance (m)
0
0.000
0
0.000
0
0.000
0
0.000
0
0.000
71
-0.071
72
-0.072
23
-0.023
37
-0.037
70
-0.070
140
-0.140
158
-0.158
47
-0.047
78
-0.078
122
-0.122
208
-0.208
235
-0.235
70
-0.070
123
-0.123
192
-0.192
260
-0.260
261
-0.261
104
-0.104
148
-0.148
245
-0.245
268
-0.268
269
-0.269
139
-0.139
180
-0.180
270
-0.270
274
-0.274
276
-0.276
173
-0.173
200
-0.200
278
-0.278
291
-0.291
290
-0.290
258
-0.258
245
-0.245
292
-0.292
303
-0.303
303
-0.303
283
-0.283
277
-0.277
307
-0.307
310
-0.310
310
-0.310
293
-0.293
283
-0.283
314
-0.314
310
-0.310
315
-0.315
300
-0.300
290
-0.290
317
-0.317
317
-0.317
317
-0.317
310
-0.310
302
-0.302
323
-0.323
320
-0.320
320
-0.320
315
-0.315
305
-0.305
325
-0.325
Settling rate (m/h)
24.66
Settling rate (m/h)
29.34
Settling rate (m/h)
8.46
Settling rate (m/h)
15.48
Settling rate (m/h)
21.96
Clarity
49
Clarity
49
Clarity
25
Clarity
49
Clarity
49
Time (s)
0
10
20
30
40
50
60
120
300
600
900
1800
3600
+50% floc dose
-50% floc dose
+50% solids
-50% solids
0.0
-0.1
-0.1
-0.1
-0.1
-0.1
-0.2
-0.3
-0.2
-0.3
-0.4
-0.3
-0.4
0
450
900
1350
Time (s)
1800
2250
2700
-0.2
450
900
1350 1800 2250 2700
Time (s)
-0.2
-0.3
-0.4
0
Distance (m)
0.0
Distance (m)
0.0
Distance (m)
0.0
Distance (m)
Distance (m)
Base case
0.0
-0.3
-0.4
0
450
900
1350 1800 2250 2700
Time (s)
-0.2
-0.4
0
450
900
1350 1800 2250 2700
Time (s)
0
450
900
1350 1800 2250 2700
Time (s)
133
Appendix C.4. Flocculant dosage test
Sample Number
Parameters
Yalea
Base case
Flocculant dose
+50%
-50%
Slurry % solids
+50%
-50%
24h rake test
(2l cylinder)
Settling rate (m/h)
37.08
28.08
9.18
19.80
34.02
-
Floc dose (g/t)
15.00
22.50
7.50
15.00
15.00
15.00
5.00
5.00
5.00
7.50
2.50
5.00
Underflow % solids
51.80
49.32
46.79
51.49
38.08
61.97
Supernatant clarity
49
49
15
49
49
49
Slurry solids conc. (%)
Sample Number
Parameters
Gara
Base case
Flocculant dose
+50%
-50%
Slurry % solids
+50%
-50%
24h rake test
(2l cylinder)
Settling rate (m/h)
24.66
29.34
8.46
15.48
21.96
-
Floc dose (g/t)
15.00
22.50
7.50
15.00
15.00
15.00
5.00
5.00
5.00
7.50
2.50
5.00
Underflow % solids
50.05
49.39
43.06
50.39
47.43
59.81
Supernatant clarity
49
49
25
49
49
49
Slurry solids conc. (%)
134
Appendix D – HECRAS Output
Appendix D.1. cross section across the wetland previously proposed area
135
Appendix D.2. Cross section information for the wetland
Reach
River Sta
Q (m3/s)
Min Chl El (m)
W.S Elev
(m)
Crit. W. S
(m)
E.G. Elev
(m)
E.G.Slope
(m/m)
Vel Chl
Flow Area
(m2)
Top Width
(m)
Froude
No.
Outlet
203.9204
0.04
121.1
121.31
121.11
121.31
0
0
11.42
60.35
0
Outlet
183.9204
0.04
121.1
121.31
121.31
0
0
11.42
60.35
0
Outlet
163.9204
0.04
121.2
121.31
121.31
0.000001
0.01
5.67
55.95
0.01
Outlet
143.9203
0.04
121.3
121.31
121.31
121.31
0.008329
0.12
0.31
50.39
0.5
Outlet
123.9204
0.04
121.1
121.11
121.11
121.11
0.005626
0.11
0.35
50.35
0.42
Outlet
110.2809
0.04
120
120.01
120.01
120.01
0.007668
0.12
0.32
50.15
0.48
Outlet
92.68047
0.04
119.6
119.61
119.61
119.61
0.005609
0.11
0.35
50.14
0.42
Outlet
63.92042
0.04
119.37
119.38
119.38
119.38
0.002844
0.09
0.43
50.16
0.31
Outlet
43.92041
0.04
119.3
119.31
119.31
119.31
0.004521
0.1
0.37
50.13
0.38
Outlet
23.92037
0.04
119.1
119.11
119.11
119.11
0.003934
0.1
0.39
50.13
0.36
Outlet
3.920386
0.04
119
119.01
119.01
119.01
0.007032
0.12
0.33
50.1
0.46
Macrophyte
301.3511
0.04
121
121.04
121.01
121.04
0.000035
0.02
2.15
51.92
0.03
Macrophyte
281.3511
0.04
121
121.04
121.04
0.000037
0.02
2.12
51.89
0.03
Macrophyte
261.3511
0.04
121
121.04
121.04
0.000039
0.02
2.08
51.86
0.03
Macrophyte
241.3511
0.04
121
121.04
121.04
0.000042
0.02
2.04
51.82
0.03
Macrophyte
221.3511
0.04
121
121.04
121.04
0.000045
0.02
1.99
51.78
0.03
Macrophyte
201.3511
0.04
121
121.04
121.04
0.000048
0.02
1.95
51.74
0.03
Macrophyte
181.3511
0.04
121
121.04
121.04
0.000053
0.02
1.89
51.69
0.03
Macrophyte
161.3511
0.04
121
121.04
121.04
0.000059
0.02
1.84
51.64
0.04
Macrophyte
141.3511
0.04
121
121.04
121.04
0.000066
0.02
1.77
51.59
0.04
136
Macrophyte
121.3512
0.04
121
121.03
121.03
0.000076
0.02
1.7
51.52
0.04
Macrophyte
101.3511
0.04
121
121.03
121.03
0.000089
0.02
1.62
51.45
0.04
Macrophyte
81.35113
0.04
121
121.03
121.03
0.00011
0.03
1.52
51.36
0.05
Macrophyte
61.35114
0.04
121
121.03
121.03
0.000144
0.03
1.4
51.26
0.05
Macrophyte
41.3511
0.04
121
121.02
121.02
0.000234
0.03
1.21
51.08
0.07
Macrophyte
24.7913
0.04
121
121.01
121.01
121.01
0.012003
0.1
0.37
50.33
0.39
Inlet
200
0.04
121.67
121.68
121.68
121.68
0.00571
0.11
0.35
50.8
0.42
Inlet
180
0.04
121.47
121.47
121.48
0.022176
0.16
0.23
50.37
0.78
Inlet
160
0.04
121.23
121.24
121.24
121.24
0.007224
0.12
0.32
50.37
0.47
Inlet
140
0.04
121.08
121.09
121.09
121.09
0.007736
0.12
0.32
50.31
0.48
Inlet
120
0.04
120.74
121.01
120.75
121.01
0
0
14.6
59.78
0
Inlet
99.99999
0.04
120.48
121.01
121.01
0
0
30.57
66.23
0
Inlet
80
0.04
120.28
121.01
121.01
0
0
43.55
69.95
0
Inlet
60.00002
0.04
120.17
121.01
121.01
0
0
50.85
71.66
0
Inlet
39.99997
0.04
120.3
121.01
121.01
0
0
42.23
69.61
0
Inlet
20.00003
0.04
120.82
121.01
120.83
121.01
0
0
9.98
57.27
0
Inlet
1.420576
Macrophyte Zone Weir
Inlet
0.5
0.04
120.84
120.84
120.84
0.003505
0.09
0.4
50.32
0.34
120.83
137
Appendix D.3. Hydraulic Output from the respective wetland sections.
Wetland
Plan: Plan 03
2014/07/31
Wetland Inlet
Le gend
EG f low
121.6
WS flow
Crit flow
121.4
Ground
Elevation (m)
121.2
121.0
120.8
120.6
120.4
120.2
120.0
0
50
100
150
200
Main Channel Distance (m)
Wetland
Plan: Plan 03
2014/07/31
Wetland Macrophyte
Le gend
EG f low
WS flow
Crit flow
121.04
Elevation (m)
Ground
121.03
121.02
121.01
121.00
0
50
100
150
200
250
300
Main Channel Distance (m)
Wetland
Plan: Plan 03
2014/07/31
Wetland Outlet
121.5
Le gend
Crit flow
EG f low
WS flow
121.0
Elevation (m)
Ground
120.5
120.0
119.5
119.0
0
50
100
150
200
Main Channel Distance (m)
138
11 References
Allen, R., G, Pereira, L. S., Raes, D., Smith, M., 1998. Guidelines for Predicting Crop Water
Requirements. Food and Agriculture Organisation of the United Nations, FAO Irrigation and
Drainage Paper 56, Rome.
Arrêté Interministériel No. 09-0767/MEA-MEIC-MEME-SG du 06 Avril 2009 rendant obligatoire
l’application des normes maliennes de rejet des eaux uses ;
Barbour MT, Gerristen J, Snyder DB, Stribling JB (1999) Rapid bioaassessment protocols
for Use in Stream and Wadeable Rivers: Periphton, Benthic Macroinvertebrates and
Fish, Second Edition. EPA 841-B-99-002. US Environmental Protection Agency,
Office of Water, Washington DC.
Bosman, C., 2009. The Hidden Dragon: Nitrate Pollution from Open-pit Mines – A case study from the
Limpopo Province, South Africa. Carin Bosman Sustainable Solutions. PO Box 26442.
Gezina, 0031, Pretoria, Gauteng, Republic of South Africa. Email: [email protected]
Brix, H. 1993. Wastewater treatment in constructed wetlands: system design and treatment
performance. In: Constructed wetlands for water quality improvement (ed Moshiri, G.A.). CRC
Press Inc., Boca Raton. pp. 9-22.
Brix, H. 1997. The macrophytes play a role in constructed treatment wetlands? Water Science and
Technology 35(5): 11-17.
Chapman, P.M., Wang, F., Janssen, C.R., Goulet, R.R., & Kamunde, C.N. (2003). Conducting
ecological risk assessments of inorganic metals and metalloids: Current status. Human Ecol.
Risk Assess. 9: 641–697.
Charte des Eaux du Fleuve Sénégal
Clark, G.A., Smajstrla, A.G. and Zazueta, F.S. 1989. Atmospheric Parameters Which Affect
Evapotranspiration. University of Florida, Florida.
Convention d’Etablissement entre l’Etat du Mali et SOMLO SA du 02 Avril 1993
Cooper, P.F., Job, G.D., Green, M.B. & Shutes, R.B.E. 1996. Reed beds and constructed wetlands
for wastewater treatment. WRc Swindon. 184 pp.
Décret no. 01-395/P-RM du 06 Septembre 2001 fixant les modalités de gestion des eaux usées et
des gadoues ;
DWA.2007. Loulo Baseline Report. Loulo gold mine. Digby Wells Environmental. Fern Isle, Section
10, 359 Pretoria Ave Randburg Private Bag X10046, Randburg, 2125, South Africa
DWA.2009. environmental and social impact assessment. Loulo gold mine. Digby Wells
Environmental. Fern Isle, Section 10, 359 Pretoria Ave Randburg Private Bag X10046,
Randburg, 2125, South Africa
139
DWA. 2011. Report back on the environmental issues at Loulo. Loulo Gold Mine. Kenieba cercle.
Mali. Unpublished report. Digby Wells Environmental. Fern Isle, Section 10, 359 Pretoria Ave
Randburg Private Bag X10046, Randburg, 2125, South Africa
DWA (2012). Hydrogeological Investigation – Gara Underground Dewatering. Unpublished report.
Digby Wells Environmental. Fern Isle, Section 10, 359 Pretoria Ave Randburg Private Bag
X10046, Randburg, 2125, South Africa
Dickens C.W.S,. & Graham P.M. (2002). The South African Scoring System (SASS), Version
5, Rapid bioassessment method for rivers. African Journal of Aquatic Science. 27: 1–
10.
European Commission (2001). Commission Regulation as regards heavy metals, Directive
2001/22/EC, no: 466/2001.
European Commission (2006). No 1881/2006 of 19 December 2006 setting maximum levels for
certain contaminants in foodstuffs
Frankenbach, R.I. & Meyer, J.S. 1999. Nitrogen removal in a surface-flow wastewater treatment
wetland. Wetlands 19(2): 403-412.
Gerber A, & Gabriel MJM. (2002). Aquatic Invertebrates of South African Rivers: Field
Guide. Institute for Water Quality Services, Department of Water Affairs and Forestry,
Pretoria.
Good, B.J. & Patrick, J.R. 1987. Root-water-sediment interface processes. In: Aquatic plants for water
treatment and resource recovery (eds Reddy, K.R. & Smith, W.H.). Magnolia Publishing Inc.,
Orlando, Florida.
Hammer, D.A. 1989. Constructed wetlands for wastewater treatment - municipal, industrial and
agriculture. Lewis Publishers, Chelsea, MI. 381 pp.
Hammer, D.A. 1994. Guidelines for design, construction and operation of constructed wetland for
livestock wastewater treatment. In: Proceedings of a workshop on constructed wetlands for
animal waste management (eds DuBowy, P.J. & Reaves, R.P.). Lafayette, IN. pp. 155-181.
Health Canada (2007). Canadian standards for various chemical contaminants in foods, Ottawa,
Ontario. http://www.hc-sc.gc.ca/fn-an/securit/chem-chim/contaminants-guidelines-directiveseng.php.
International Network for Acid Prevention (INAP).2003. Treatment of Sulphate in Mine Effluents.
Lorax Environmental.
International Finance Corporation (IFC). 2010. Environment, Health and Safety Guidelines. Mining.
Table 1. Effluent guidelines – World Bank Group.
Joel Holliday. Geology map done by him….
Johnston, C.A. 1991. Sediment and nutrient retention by freshwater wetlands: effects on surface
water quality. Critical Reviews in Environment Control 21: 461-565.
Journal of Ecological Engineering (2005). Constructed wetlands for Wastewater Treatment. Ecol. Eng. 25,
475–477. www.elsevier.com/locate/ecoleng
140
Joy B. Zedler, CRC Press, Boca Raton, FL. (2000). Handbook for Restoring Tidal Wetlands
Kadlec, R.H. & Knight R.L. 1996. Treatment wetlands. CRC Press. Boca Raton, Florida. 893 pp.
Kayombo, S.; Mbwette, T.S.A; Katima, J.H.Y; Ladegaard, N.; and JØrgensen, S.E. (2004). Waste
Stabilization Pond (WSP) and Constructed Wetland (CW) Research Project. Prospective
College of Engineering & Technology (University of Dar Es Salaam) and The Danish
University of Pharmaceutical Sciences (Section of Environmental Chemistry – Copehhagen,
Denmark)
Kleynhans CJ. (1996) A qualitative procedure for the assessment of the habitat integrity
status of the Luvuvhu River. Journal of Aquatic Ecosystem Health 5: 41-54.
Kyambadde, K. J., G. F. and D. L., G., 2004a. A comparative study of Cyperus papyrus and
Miscanthidium violaceum-based constructed wetlands for wastewater treatment in a tropical
climate. Water Research, 38: 475-485.
Lawrence D.M; Treloar P.J; Rankin A.H; Harbridge P., and Holliday J (2013). The Geology and
Mineralogy of the Loulo Mining District, Mali, West Africa: Evidence for Two Distinct Styles of
Orogenic Gold Mineralization. ©2013 Society of Economic Geologists, Inc. Economic
Geology, v. 108, pp. 199–227
Lawrence D.M (2010). Petrography studies at Gounkoto. GDD 2010 sample suite. Loulo Gold Mine
Loulo Gold Mine. 2012. Gara ore reserve statement. Underground department. December 2012
Loi no. 2012-015/AN du 27 Février 2012 portant code miner au Mali
Loi no. 02-006/AN du 31 Janvier 2002 portant code minier au Mali
Loi no. 01-020 du 20 mai 2001 relative aux pollutions et nuisances;
Loi no. 08-033 du 11 Aout 2008 relative aux installations classées pour la protection de
l’environnement ;
MacDonald, D.D., C.G. Ingersoll, and T.A. Berger. (2000a). Development and evaluation of
consensus based sediment quality guidelines for freshwater ecosystems. Arch. Environ.
Contam. Toxicol. 39:20-31.
MacDonald, D.D., L.M. Dipinto, J. Field, C.G. Ingersoll, and E.R. Long. (2000b). Development and
evaluation of consensus-based sediment effect concentrations for polychlorinated biphenyls.
Environ. Toxicol. Chem. 19:1403-1413.
Malzbender, D. & Earle, A. (2007). Water Resources of the SADC: Demands, Dependencies and
Governance Responses. Institute for Global Dialogue’s (IGD) and the Open Society Initiative
for Southern Africa’s (OSISA) “Research Project on Natural Resources Dependence and Use
in Southern Africa: Economic and Governance Implications”. African Centre for Water
Research (ACWR). Cape Town.
Margalef R. (1961). Information Theory in Ecology. General Systems 3: 36–71.
141
McMillan PH, (1998). An integrated habitat assessment system (IHAS v2) for the rapid
biological assessment of rivers and streams. Division of the Environment and
Forestry Technology, Report No. ENV-P-I 98132. CSIR, Pretoria.
Mitchell, M.J; Driscoll, C.T; Kahl, J.S; Likens, G.E; Murdoch, P.S; and Pardo, L.H. 1996. Climatic
Control of Nitrate Loss from Forested Watersheds in the Northeast United States.
Environmental Science and Technology 30: 2609-2612
Morgan (2014). Hydrological assessment of a constructed wetland. Unpublished report. Digby wells
Environmental. Fern Isle, Section 10, 359 Pretoria Ave Randburg Private Bag X10046,
Randburg, 2125, South Africa
Paterson & Cooke.2011. Thickening testwork on Yalea and Gara Underground Discharge Samples.
Report No. SEN-RDG-8241.1 R01 Rev1. Unit 2, 33 Kyalami Boulevard, Kyalami Business
Park, Kyalami, 1684, South Africa
unpublished
PO Box 14, The Woodlands, 2080, South Africa.
RDA, 1989 (Recommended dietary allowance (10th ed.). Washington, DC: National Academic Press)
and WHO
Reddy, K.R. & De-Busk, T.A. 1987. State-of-the-art utilisation of aquatic plants in water pollution
control. Water Science and Technology 19(10): 61-79.
Reed, S. C., Crites, R. W. & Middlebrooks, E. J. 1995. Natural systems for waste management and
treatment. Second edition. McGraw-Hill Inc. New York, NY. 433 pp.
Rosgen D (1996) Applied river morphology. Wild-land hydrology, Pagosa Springs, Colorado
Shklomanov, I.A. (1998). World Water Resources. Prepared in the framework of the International
Hydrological Programme. State Hydrological Institute. St Petersburg, Russia.
Sikora, F.J., Tong, Z., Behrends, L.L., Steinberg, S.L. & Coonrod, H.S. 1995. Ammonium removal in
constructed wetlands with recirculating subsurface flow: removal rate and mechanisms (eds
Kadlec, R.H. & Brix, H.). Water Science and Technology 32(3): 193-202.
Tate, R., 2014. Loulo Aquatic Ecological Report. Unpublished report. Digby wells
Environmental Fern Isle, Section 10, 359 Pretoria Ave Randburg Private Bag X10046,
Randburg, 2125, South Africa
Thieme ML, Abell R, Stiassny MLJ, Skelton P (2005). Freshwater Ecoregions of Africa and
Madagascar: A Conservation Assessment. World Wildlife Fund.
Thirion CA, Mocke A, Woest R (1995). Biological monitoring of streams and rivers using
SASS4. A Users Manual. Internal Report No. N 000/00REQ/1195. Institute for Water
Quality Studies. Department of Water Affairs and Forestry. 46.
Turkmen, A., Turkmen,M., Tepe, Y., & Akyurt, I. (2005).Heavymetals inthree Commercially valuable
fish species from Iskenderun Bay, North-ern East Mediterranean Sea, Turkey.Food
Chemistry,91, 167–172.
Treatment Wetlands (2004), Robert H. Kadlec and Robert L. Knight, Lewis Publishers, Boca Raton,
Fl.
142
TWP.2012 – 3D design of the proposed thickener
United States Environmental Protection Agency (2000). Guiding Principles for Constructed Treatment
Wetlands: Providing for Water Quality and Wildlife Habitat (2000), United States
Environmental
Protection
Agency,
EPA
843-B-00-003.
www.epa.gov/owow/wetlands/constructed/guide.html
Available
online
at
United States Environmental Protection Agency (2000). Constructed Wetlands Handbooks (Volumes
1-5): A Guide to Creating Wetlands for Agricultural Wastewater, DomesticWastewater, Coal
Mine Drainage and Stormwater in the Mid-Atlantic Region (1993-2000), Available online at
www.epa.gov/owow/wetlands/pdf/hand.pdf
United States Food and Drug Administration (1993). Guidance document for nickel in shell fish.
DHHS/PHS/FDA/CFSAN/Office of seafood, Washington D.C.
United States Environmental Protection Agency (USEPA) (2006). Rapid Bioassessment Protocols for
Use in Streams and Wadeable Rivers: Periphyton, Benthic Macroinvertebrates, and Fish,
Second Edition
US Army Corps of Engineers - Hydrologic Engineering Centers River Analysis System (HECRAS). Department of The Army Corps of Engineers Institute for Water Resources Hydrologic
Engineering Center 609 Second Street Davis, CA 95616-4687
Vymazal and L. Kröpfelová, 2009. Removal of organics in constructed wetlands with horizontal subsurface flow: A review of the field experience. Science of the Total Environment, 407: 39113922.
Wetland International (2003). The Use of Constructed Wetlands for Wastewater Treatment. Malaysia
Office. 3A31. Block A, Kelana Center Point. Jalan SS7/19. 47301 Petaling Jaya. Selangor,
Malaysia. Email: [email protected] – www.wetlands.org
WHO, (1993). 41
st
Report of the Joint FAO/WHO Expert Committee on Food Additives. WHO
Technical Report Series, no. 837. Geneva.
WHO. (1999). Joint FAO/WHO. Expert committee on food additives. In: Summary and conclusions,
53rd meeting, Rome, 1–10 June.
Zhou Q, Zhang J, Fu J, Shi J, Jiang G (2008) Biomonitoring: An appealing tool for
assessment of metal pollution in the aquatic ecosystem. Analytica Chimica Acta 606:
135–15.
143
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement