IN SITU SOURCE CHARACTERISATION OF
DENSE NON-AQUEOUS PHASE LIQUIDS
(DNAPLs) IN A FRACTURED ROCK
ENVIRONMENT
Vierah Hulley
Submitted in accordance with the requirements for the Doctor of Philosophy 
degree in the Faculty of Natural Sciences and Agriculture, Department of 
Geohydrology at the University of the Free State, Bloemfontein, South 
Africa 
June 2013
 
 
 
 
 
 
 
 
Promoter: Professor K.T. Witthüser  
ACKNOWLEDGEMENTS 
This thesis is the result of continued support of several people to whom I want to 
express my gratitude. 
Without the funding and faith of the Investigation Site owners this study would never 
have resulted. Thank you for this wonderful opportunity and for trusting me to 
undertake and manage this study. The Investigation Site personnel added a vast 
amount to the understanding of historical site processes and subsequently to the 
identification of potential release areas.  
My gratitude is also extended to Dr. Smit who allowed me the freedom to pursue this 
research.  
I will be eternally indebted to the following Consultants who assisted in collecting field 
data, preparing of conference papers and on treating me as part of the team: Joe 
Fiacco, Paul Aucamp, Liam Smith, Ryno Weitz. This has been quite a ride! 
To my promoter, Prof. Kai Witthüser, thank you for your patience with my erratic 
submissions and long periods of silence. Thank you also for the advice on improving 
this thesis. 
I had the absolute pleasure of working with Prof. van Heerden and Dr Botes from the 
UFS Metagenomic Platform. Thank you for your enthusiasm on the findings from the 
site.  
The last few months have been particularly challenging for me. I would not have 
survived had it not been for my amazing support system in Houston: Anja, Karen and 
Raakhi; your generosity and friendship overwhelms me. 
Finally, to the most significant people in my life; my wonderful and supporting 
husband Francois and my adorable daughter, Lara – you are my inspiration, my 
motivation and my hope. This is for you.
i 
Table of Contents 
1 Introduction........................................................................................................ 1
1.1 Background and Motivation ......................................................................... 1
1.2 Objectives of Thesis .................................................................................... 2
1.3 Structure of Thesis ...................................................................................... 3
2 Literature Review ............................................................................................... 5
2.1 Introduction to DNAPLs ............................................................................... 5
2.1.1 DNAPL properties and behaviour in the subsurface ............................. 8
2.1.2 Transformation of DNAPLs ................................................................ 25
2.2 DNAPL Site Characterisation .................................................................... 38
2.2.1 Site characterisation and the significance of the conceptual site model 
(CSM) 38
2.2.2 DNAPL source zone characterisation in complex geo-contaminant 
settings 42
2.3 Defining the Level of Characterisation Required and Cost Implications ..... 52
3 DNAPL Source Zone Characterisation In a Fractured System ......................... 55
3.1 Background to the Investigation Site ......................................................... 55
3.2 Methodologies Used to Investigate the Site ............................................... 60
3.2.1 Collation and review of historical information and the identification of 
information gaps .............................................................................................. 64
3.2.2 Source zone characterisation ............................................................. 65
3.2.3 Near-field geological and hydrogeological characterisation ................ 76
3.2.4 Hydrogeochemical and biogeochemical characterisation ................... 81
3.3 Results ...................................................................................................... 85
3.3.1 Regional geology ............................................................................... 85
3.3.2 Local geology ..................................................................................... 89
3.3.3 Geohydrology ................................................................................... 108
3.3.4 DNAPL source zones delineation ..................................................... 117
3.3.5 Hydrogeochemical characteristics .................................................... 154
ii 
3.3.6 Biogeochemical characteristics ........................................................ 162
3.4 Conceptual Site Model ............................................................................ 169
4 Defining the Value and Level of Source Zone Characterisation Required in a 
fractured rock environment .................................................................................... 174
4.1 Source Zone Characterisation in a Complex Geo-contaminant Setting ... 174
4.2 Source Zone Characterisation Efficiency ................................................. 179
5 Conclusions and Recommendations .............................................................. 184
5.1 Novel Approaches to DNAPL Source Zone Characterisation in a Fractured 
Rock Environment ............................................................................................. 184
5.2 Source Zone Characteristics in a Fractured Rock Environment ............... 184
5.3 Source Zone Architectures of the Investigation Site ................................ 185
5.4 Recommendations for Future Work ......................................................... 186
6 References .................................................................................................... 187
Abstract ................................................................................................................. 208
Opsomming ........................................................................................................... 209
Key Words ............................................................................................................ 210
 
 
 
 
 
 
 
 
 
 
 
 
 
iii 
List of Figures 
Figure 2-1: Selected DNAPL production in the USA from 1920-1998 (modified from 
Cohen and Mercer, 1993; and Moran, 2006) ............................................................. 7
Figure 2-2: Evolution in the knowledge and understanding of DNAPLs as a primary 
groundwater contaminant (modified from Pankow and Cherry, 1996; Sale et al., 
2008) ......................................................................................................................... 9
Figure 2-3: Capillary pressure-saturation curve for porous media illustrating the 
relationship between initial saturation along main drainage (points A, B, C) and 
residual saturation along secondary wetting (from Kueper et al., 1993) ................... 11
Figure 2-4: Relative permeability – saturation curves for wetting phase (a) and non-
wetting phase (b) (from Vogler et al., 2001) ............................................................. 13
Figure 2-5: NAPL in the pore space of a granular porous ........................................ 14
Figure 2-6: Depiction of NAPL imbibition in the pore spaces. (a) shows small pores 
and pore wedges filled with NAPL before the larger pore bodies are filled. After these 
smaller spaces are filled, the NAPL is assumed to be immobile (red). (b) depicts the 
pore system is NAPL-filled with the mobile (free) NAPL shown as pink. Water is blue 
and solid particles are grey (from Lenhard et al., 2004) ........................................... 15
Figure 2-7: Simplified illustration of the existence of different phases of DNAPL within 
a heterogeneous setting (from Sale et al., 2008) ..................................................... 18
Figure 2-8: Illustration of matrix diffusion of dissolved phase contaminants adjacent to 
DNAPL source and along length of plume in a fracture. Matrix diffusion can attenuate 
the rate of plume advance in fractured rock (concentration vs distance plot), and can 
result in delayed breakthrough curves (concentration vs time figure) (from Kueper 
and Davies, 2009) ................................................................................................... 19
Figure 2-9: Significance of matrix diffusion (from Sale et al., 2008) ......................... 20
Figure 2-10: Evolution of a DNAPL release from early to late stages (from Sale et al., 
2008) ....................................................................................................................... 21
Figure 2-11: Pooling of DNAPL in a fractured network. If hydrostatic equilibrium is 
assumed, then the capillary pressure increases linearly with depth (from Kueper et 
al., 2003) ................................................................................................................. 23
Figure 2-12: Fracture aperture required to stop DNAPL migration versus the height of 
the accumulated DNAPL (from Kueper et al., 2003) ................................................ 23
iv 
Figure 2-13: Illustration of differing migration pathways for a low interfacial tension 
system (a) versus a high interfacial tension system (b) (from Pankow and Cherry, 
1996) ....................................................................................................................... 24
Figure 2-14: Sorption and volatilisation transformation pathways (from Cho et al., 
2004) ....................................................................................................................... 29
Figure 2-15: Biotic and abiotic transformation pathways (from Cho et al., 2004) ...... 29
Figure 2-16: Degradation pathways of chlorinated ethenes and ethanes. Bold arrows, 
reductive dechlorination; fine arrows, dichloroelimination; dashed arrows, 
dehydrohalogenation (from Hunkeler et al., 2005) ................................................... 31
Figure 2-17: Bayesian Belief Network for the reductive dechlorination .................... 32
Figure 2-18: Transition from regional to site-specific data during the progression of 
the site characterisation (from Sara, 2003) .............................................................. 39
Figure 2-19: Life cycle of the CSM (from United States Environmental Protection 
Agency, 2003) ......................................................................................................... 41
Figure 2-20: Hypothetical DNAPL source zone (from Kueper and Davies, 2009) ..... 42
Figure 2-21: Geo-contaminant settings (from  Air Force Center for Engineering and 
the Environment, 2007) ........................................................................................... 43
Figure 2-22: Variations in permeability and hydraulic conductivity for different 
geological media (from National Research Council, 2004) ...................................... 45
Figure 2-23: Lines of evidence to assess DNAPL presence (modified from Kueper 
and Davies, 2009) ................................................................................................... 51
Figure 2-24: Flowchart depicting iterative data collection process used in refining the 
DNAPL source zone boundaries (modified from Kueper and Davies, 2009) ............ 52
Figure 3-1: Land use surrounding the Investigation Site .......................................... 57
Figure 3-2: Potential source areas at the Investigation Site ..................................... 58
Figure 3-3: Available historical near-field and far-field borehole information (modified 
from Hulley et al., 2008) .......................................................................................... 62
Figure 3-4: Historical evidence of DNAPL contaminated in the subsurface at the 
Investigation Site (modified from SRK Consulting, 1995; SRK Consulting, 2001; 
ERM, 2008) ............................................................................................................. 63
Figure 3-5: Tasks and tools/methods for source zone characterisation at the 
Investigation Site ..................................................................................................... 68
Figure 3-6: Oil Red O® staining of residual DNAPL in soil boring 177 ..................... 69
v 
Figure 3-7: 2-D Lund surface geophysical traverses (Results from ERM, 2008; ERM, 
2012) ....................................................................................................................... 70
Figure 3-8: Location of all boreholes drilled at the Investigation Site (results from SRK 
Consulting, 1995; SRK Consulting, 2001, ERM, 2005; ERM, 2012) ........................ 83
Figure 3-9: Flow diagram for groundwater sampling at the Investigation Site (from 
ERM, 2008) ............................................................................................................. 84
Figure 3-10: North-south trending regional geological cross section (modified from 
ERM, 2008) ............................................................................................................. 87
Figure 3-11: East-west trending regional geological cross section (modified from 
ERM, 2008) ............................................................................................................. 88
Figure 3-12: Stratigraphic model ............................................................................. 89
Figure 3-13: 2-D resistivity profiles of Traverses 102 and 103 ................................. 91
Figure 3-14: 2-D resistivity profile of Traverse 109 .................................................. 92
Figure 3-15: Typical weathered zone profile at the site ............................................ 93
Figure 3-16: Particle size distribution of borehole EDC2S ....................................... 94
Figure 3-17: Picture taken of roto-sonic core (EDC2S) showing the transition from 
residual clay dolerite to highly weathered dolerite.................................................... 94
Figure 3-18: Base of weathering elevation ............................................................... 95
Figure 3-19: Photographic log of core hole BH63C ................................................ 101
Figure 3-20: Dolerite sill lower contact elevation .................................................... 104
Figure 3-21: C3 coal seam upper contact elevation ............................................... 105
Figure 3-22: Pre-Karoo basement elevation .......................................................... 106
Figure 3-23: Water strike elevation as a function of surface elevation .................... 108
Figure 3-24: Groundwater elevation as a function of surface elevation at borehole 109
Figure 3-25: Pressure transducer data for selected boreholes at the site (results from 
ERM, 2008) ........................................................................................................... 111
Figure 3-26: Groundwater flow direction and hydraulic gradient ............................ 112
Figure 3-27: Relative transmissivities of boreholes (results from ERM, 2012) ....... 113
Figure 3-28: 2-D resistivity anomalies detected at the Investigation Site (Results from 
ERM, 2008; ERM 2012) ........................................................................................ 119
vi 
Figure 3-29: Frequency distribution histograms of grouped chlorinated hydrocarbon 
compounds found at the Investigation Site ............................................................ 121
Figure 3-30: Target chlorinated hydrocarbon compounds detected during the passive 
soil gas survey ...................................................................................................... 122
Figure 3-31: Spatial distribution of chlorinated ethene vapours at the Investigation 
Site (results from ERM, 2010) ............................................................................... 123
Figure 3-32: Spatial distribution of chlorinated ethane vapours at the Investigation 
Site (results from ERM, 2010) ............................................................................... 124
Figure 3-33: Spatial distribution of chlorinated methane vapours at the Investigation 
Site (results from ERM, 2010) ............................................................................... 125
Figure 3-34: Inferred chlorinated hydrocarbon DNAPL release areas (results from 
ERM, 2010) ........................................................................................................... 126
Figure 3-35: Soil total chlorinated hydrocarbons concentrations, spatial distribution 
and composition at the Investigation Site (results from ERM 2012) ....................... 129
Figure 3-36: Chlorinated hydrocarbons distribution in the soils (in μg/kg) beneath the 
historical production, storage and handling area 1 ................................................. 132
Figure 3-37: 3-D solid models of grouped chlorinated hydrocarbons soil 
concentrations (μg/kg) ........................................................................................... 134
Figure 3-38: Plan view of inferred DNAPL source zones in soil ............................. 135
Figure 3-39: Portion of downhole geophysical log for core hole MLR01 showing the 
large density of fractures ....................................................................................... 137
Figure 3-40: (a) Schmidt net (equal area, contoured) stereonet and (b) Rose diagram 
of fractures within the source zones (uncorrected) ................................................ 137
Figure 3-41: 3-D strip logs showing fracture discs of core holes drilled in the historical 
production and handling facility ............................................................................. 138
Figure 3-42: Variation of fracture density with elevation at the historical production 
and handling facility ............................................................................................... 139
Figure 3-43: Location of transmissive fracture in PBH221 ..................................... 139
Figure 3-44: 3-D strip log of CBH42 showing the intervals and values used for 
calculating hydraulic aperture ................................................................................ 140
Figure 3-45: Apparent DNAPL thickness measured in boreholes .......................... 142
vii 
Figure 3-46: Logs of CBH42 and PBH213 showing the location of residual DNAPL 
(shown in blue as determined through the use of ribbon NAPL samplers and in red as 
observed using Oil Red O®) in relation to the litho-stratigraphy ............................. 144
Figure 3-47: Oil Red O®  staining at 22.7 m bgl in core hole CBH42 indicating the 
presence of residual DNAPL ................................................................................. 145
Figure 3-48: Locations of pooled, residual and inferred DNAPL in the fractured 
bedrock at the Investigation Site ............................................................................ 147
Figure 3-49: Location of bedrock source zone cross sections ................................ 148
Figure 3-50: Bedrock cross section A-B (modified from ERM, 2012) ..................... 149
Figure 3-51: Bedrock cross section C-D (modified from ERM, 2012) ..................... 149
Figure 3-52: Bedrock cross section E-F (modified from ERM, 2012)...................... 150
Figure 3-53: Bedrock cross section G-H (modified from ERM, 2012) ..................... 150
Figure 3-54: DNAPL composition in boreholes (Results from ERM, 2012) ............ 153
Figure 3-55: Plot of density vs viscosity for samples analysed from the Investigation 
Site 154
Figure 3-56: Variation in chlorinated ethenes, ethanes and methanes concentrations 
in SRK14S/D ......................................................................................................... 158
Figure 3-57: Total chlorinated ethenes plume (a) 1998, (b) 2002 and (c) 2012 ...... 159
Figure 3-58: Total chlorinated ethanes plume (a) 1998, (b) 2002 and (c) 2012 ...... 160
Figure 3-59: Total chlorinated hydrocarbon DNAPLs isoconcentrations filled contours 
plot (boreholes located in unweathered dolerite) ................................................... 161
Figure 3-60: Semi-quantitative depth discrete total dissolved hydrocarbon 
concentrations ....................................................................................................... 162
Figure 3-61: Genomic DNA extracted from selected borehole samples (a) pre-
optimisation and (b) post-optimisation. Lane M, MassRulerTM DNA Ladder 
(Fermentas); lane 1, MLR01S, lane 2, SRK02D; lane 3, SRK12D; lane 4, SRK34S; 
lane 5, SRK02S, lane 6, SRK21D.......................................................................... 165
Figure 3-62: δ13C vs natural logarithm of PCE concentration in boreholes ............. 167
Figure 3-63: δ13C vs natural logarithm of PCE concentration in boreholes ............. 167
Figure 3-64: Plan view conceptual site model of the Investigation Site .................. 171
Figure 3-65: Schematic cross sectional view of the conceptual site model for the 
Investigation Site ................................................................................................... 172
viii 
Figure 4-1: Schematic illustration of the extent of site assessments at the site ...... 175
Figure 4-2: Schematic representation of incorrect estimation of source zones based 
on traditional site characterisation approach (modified from Ramsey et al., 2002) . 176
Figure 4-3: Plot of cost of analysis for batches of samples using an off-site 
commercial laboratory ........................................................................................... 177
Figure 4-4: Hypothetical knowledge scenario at time t0 ......................................... 180
Figure 4-5: Proposed integrated appraoch to determine the level of source zone 
characterisation required ....................................................................................... 182
Figure 4-6: Proposed source zone characterisation methodology for the South 
African scenario .................................................................................................... 183
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ix 
List of Tables
 
Table 2-1: Dominant attenuation mechanisms for DNAPL groups (In Pretorius, 2007 
from Carey et al., 2000) ........................................................................................... 27
Table 2-2: Biotic and abiotic transformation reaction definitions (modified from Vogel 
et al., 1987) ............................................................................................................. 28
Table 2-3: Published δ13C values for TCE and PCE from various manufacturers .... 37
Table 2-4: Advantages and disadvantages of the Triad approach  (Interstate 
Technology and Regulatory Council, 2003) ............................................................. 40
Table 2-5: Characterisation methods and their potential for providing information 
about the source zone (from National Research Council, 2004) .............................. 46
Table 2-6: The degree of uncertainty regarding the presence of a source zone at a 
site based on the occurrence of various events (National Research Council, 2004) 50
Table 3-1: Facilities at the Investigation Site that are/were associated with production, 
handling and/or disposal of DNAPLs ....................................................................... 59
Table 3-2: Target compounds for the passive soil gas survey.................................. 69
Table 3-3: Constant values used for the Monte Carlo simulations to determine soil 
concentration threshold values ................................................................................ 72
Table 3-4: Lithological parameters measured at selected soil and bedrock cores .... 75
Table 3-5: Effectiveness of the borehole geophsyical tools used at the Investigation 
Site 96
Table 3-6: Summary of fracture pick analysis on selected core holes ...................... 97
Table 3-7: Geotechnical results of samples retrieved from soil and bedrock cores 107
Table 3-8: Hydraulic conductivity values for the dolerite, Karoo Ecca and pre-Karoo 
basement aquifers ................................................................................................. 114
Table 3-9: Groundwater vs surface water elevation for the stream 
boreholes/piezometers .......................................................................................... 115
Table 3-10: Descriptive statistics for grouped chlorinated hydrocarbon compounds 
found at the Investigation Site ............................................................................... 121
Table 3-11: Calculated soil chlorinated hydrocarbon concentration threshold 
minimum and maximum values for the Investigation Site ...................................... 128
Table 3-12: Method of detection of residual and pooled DNAPL within boreholes . 143
x 
Table 3-13: Summary of maximum depth of ribbon NAPL sampler installation ...... 144
Table 3-14: Summary of mole fractions (%) composition of groundwater samples with 
inferred chlorinated hydrocarbon DNAPL (results .................................................. 151
Table 3-15: Physical and chemical properties of pooled DNAPL samples from the 
Investigation Site ................................................................................................... 152
Table 3-16: Calculated length of time required for the dissolved phase plume to 
detach from the weathered source zone ................................................................ 156
Table 3-17: Pore water results for stream piezometer samples ............................. 157
Table 3-18: Chemical oxygen demand, biological oxygen demand and dissolved 
oxygen results ....................................................................................................... 163
Table 3-19: Toxicity screening of groundwater ...................................................... 163
Table 3-20: Field physical measurements ............................................................. 164
Table 3-21: Summary of DNA concentration and purity of genomic DNA extracts . 165
Table 3-22: Summary of compound specific isotope results .................................. 168
Table 3-23: Release areas and nature of release for the Investigation Site ........... 170
Table 4-1: Summary on the cost efficiency and the effort efficiency of novel 
approaches used at the Investigation Site ............................................................. 178
 
 
xi 
List of Appendices 
(available electronically) 
Appendix A: Core logs 
Appendix B: 2-D resistivity profiles 
Appendix C: Passive soil gas survey results 
Appendix D: Weathered zone analyses results 
Appendix E: Monte Carlo simulations to determine soil concentration threshold 
values 
Appendix F: Monte Carlo simulations to determine plume retardation factors 
Appendix G: FACTTM chemistry results 
Appendix H: EpiWeb v4.0.BioWIN simulation results 
 
 
 
 
 
 
 
xii 
CHAPTER 1 
1 INTRODUCTION 
1.1 Background and Motivation 
The combination of water scarcity and the reliance of communities on groundwater 
have led to stricter regulations within South Africa to protect and preserve 
groundwater from anthropogenic degradation. The management and protection of 
water resources (including groundwater) is legislated through the National Water Act 
(Republic of South Africa, 1998).   
Of the contaminants inventoried in South Africa, DNAPLs represented over 60% of 
the priority pollutants (Usher et al., 2004). DNAPLs (such as chlorinated 
hydrocarbons) also make up the most frequently detected contaminants in aquifers in 
the United States of America (USA). A study undertaken by the United States 
Environmental Protection Agency (United States Environmental Protection Agency, 
1990) found that synthetic organic compounds were found in 17-28% of their public 
drinking water treatment facilities where groundwater is used as the water source. A 
more recent study undertaken by the United States Geological Survey (USGS) found 
that 20% of water samples from aquifers sampled contained volatile organic carbons 
(Zogorski et al., 2006). 
The acronym “DNAPL” was first coined in the United States of America during 
litigation proceedings in New York State in the late 1970’s - early 1980’s and referred 
to a black, immiscible liquid that was heavier than water (Pankow and Cherry, 1996). 
Chemicals falling within this category include halogenated and non-halogenated 
hydrocarbons (for e.g. chlorinated hydrocarbons, etc.), pesticides, polychlorinated 
biphenyl mixtures (PCBs), phthalate compounds, substituted aromatic compounds 
(for e.g. nitrotoluenes, nitrobenzenes, benzyl chlorides, chloroanilines, etc.), mercury, 
creosotes and coal tars. DNAPLs are normally not released into the environment as a 
pure or single component chemical, but are often released as spent solvents that 
contain a fair fraction of other organic compounds. The effect of these components 
can potentially add to the problem of characterisation and remediation of DNAPL-
impacted areas by influencing the properties of the DNAPL (Interstate Technology 
and Regulatory Council , 2000). 
1 
The properties of DNAPLs allow them to have numerous beneficial uses in processes 
such as degreasing equipment and treating wood. The use of these compounds is 
not restricted to heavy industry but they are also commonly used in workshops, dry-
cleaners and photography workshops located all over South Africa. The potential of 
contamination of water resources from these chemicals is therefore ubiquitous. 
Consequently, a comprehensive understanding of the relationship between 
form/phases, geo-mechanisms, toxicity, degradation and exposure pathways of these 
priority pollutants is important as it directly affects the effectiveness of remedial 
solutions. Despite the low solubility of these compounds in water, components of the 
DNAPLs are soluble enough to present risks to human health and the environment. 
There are two (2) prevalent exposure pathways of these chemicals to human beings; 
viz. through ingestion of contaminated groundwater, and/or inhalation of toxic 
vapours (Cheremisinoff, 1990). 
In recent years, a number of site-specific investigations and remediation efforts have 
shown that DNAPL encountered in fractured bedrock poses a particular problem in 
locating and removing (for e.g. United States Environmental Protection Agency, 
2004). Fractured bedrock is the prevalent geo-contaminant setting in the South 
African scenario. Conventional methods of using the pump-and-treat technology 
alone has shown to be only partially successful in removing some of the 
contaminants from the DNAPL source zones (Mackay and Cherry, 1989; MacDonald 
and Kavanaugh, 1994). This lack of success can be partially attributed to the 
uncertainty associated with incomplete source zone characterisation. 
1.2 Objectives of Thesis 
This research was initiated due to the lack of adequate source zone characterisation 
of DNAPLs within fractured bedrock. The United States Environmental Protection 
Agency (2003); and the Strategic Environmental Research and Development 
Program and Environmental Security Technology Certification Program  (SERDP and 
ESTCP, 2001; 2006) prioritised the following as the highest research needs related to 
DNAPLs: 
• The assessment of source zone treatment 
• Source zone delineation and characterisation 
• Diagnostic tools to evaluate remediation performance. 
The Investigation Site (see Chapter 3) presents an ideal opportunity to test source 
zone characterisation methodologies of a DNAPL-impacted site on a mega-scale. 
2 
Additionally, the benefits to industry cannot be undermined. Failure of remediation 
technologies as a direct result of poor source zone characterisation is not uncommon. 
However, these have paramount cost implications for site owners.  
The objectives of this thesis are as follows: 
• Determine the architecture of DNAPLs in a fractured (heterogeneous) system.  
• Evaluate the effectiveness of source zone characterisation and delineation 
using multiple characterisation tools in a fractured system. This is a particularly 
unexplored process internationally. Focussed studies have been on sites with porous 
media due to the difficulty in locating source zones in fractured rock aquifers 
(National Research Council, 2004).  
• Demonstrate the effectiveness of utilising the Triad Approach to site 
characterisation in a fractured system as compared to a conventional site 
characterisation process. This process will be innovative for characterisation in South 
Africa. 
• Evaluate and document the process to be followed for a complex site with 
multiple contaminants and source zones within a fractured rock environment. 
• Develop criteria to allow Site Owners to optimise data collection that can 
inform decision-making. 
1.3 Structure of Thesis 
This thesis is divided into five (5) chapters. Chapter 1 provides the background, 
motivation and objectives of this thesis.  
Chapter 2 describes previous work undertaken by other authors to determine the 
influence of DNAPL and media properties on the fate, transport and transformation of 
DNAPLs in the subsurface. Chapter 2 also provides a literature review of site 
characterisation tools investigated by previous authors and work done to date in 
attempting to define the optimum level of characterisation required at a site. 
3 
Chapter 3 is a synthesis of the background to the Investigation Site and the 
evaluation of traditional and novel methods and approaches used in this research to 
characterise the DNAPL source zones. The results of the research are presented.  
Chapter 4 documents a qualitative process model to improve source zone 
characterisation efficiencies in the South African context. 
Chapter 5 presents the conclusions of this research and contains recommendations 
for future work. 
The Appendices to this research are provided electronically at the back of the 
document. 
4 
CHAPTER 2 
2 LITERATURE REVIEW 
Due to the prevalent use of DNAPLs (particularly halogenated hydrocarbons or 
solvents) during the 20th Century, groundwater contamination from these chemicals 
has become commonplace in industrialised regions of countries (for e.g. Cohen and 
Mercer, 1993; United States Environmental Protection Agency, 1993; Usher et al., 
2004; Moran, 2006; etc.). This chapter presents a description of the uses and 
attributes of DNAPLs that allows them to be beneficial while at the same time having 
toxic and costly ramifications to detect and remediate in the subsurface. This chapter 
also explores the properties of DNAPLs and their associated behaviour in geo-
contaminants settings, with an emphasis on fractured rock environments; the 
prevalent geo-contaminant setting in the South African scenario.  
2.1 Introduction to DNAPLs  
DNAPLs are chemicals (organic and inorganic) or groups of chemicals displaying the 
following properties:   
• fluid density greater than 1.01 g/cm3,  
• solubility less than 2% (or 20000 mg/l),  
• vapour pressure less than 40 kPa (300 torr) (Pankow and Cherry, 1996).  
Huling and Weaver (1991) and Cohen and Mercer (1993) identify over 70 DNAPLs 
that have been commonly used in a wide variety of applications ranging from product 
feedstocks to end products. The types of DNAPLs most commonly encountered at 
contaminated sites are chlorinated hydrocarbons (such as trichloroethane, 
tetrachloroethane, trichloroethene etc), followed by PCBs, coal tars and creosote 
(Kerndorff et al., 1992; Cohen and Mercer, 1993; United States Environmental 
Protection Agency, 1993; Usher et al., 2004).  A detailed description of the major 
groups of DNAPLs is provided in Cohen and Mercer (1993).  DNAPLs typically enter 
the subsurface as a result of activities related to the disposal, storage, usage, 
transportation and handling of these chemicals.  
Chlorinated hydrocarbon DNAPLs such as trichloroethene (TCE or trichloroethylene), 
tetrachloroethene (PCE or perchloroethene) and trichloroethane (TCA) are used in 
aerospace and electronics industries, dry cleaning, metal cleansing and degreasing, 
5 
the manufacture of pharmaceuticals, pesticide formulation and wood manufacturing
(Cohen and Mercer, 1993; Halogenated Solvents Industry Alliance, 2004). 
Chlorinated solvents are also encountered in everyday household consumer products 
such as drain/oven/pipe cleaners, household degreasers, pesticides, shoe polishes 
and deodorizers (United States Environmental Protection Agency, 1980).
DNAPL that is composed of two or more chemical compounds are referred to as 
multi-component DNAPL or a mixed DNAPL. This is the common scenario at 
chemical waste landfills, lagoons, chemical waste handling and reprocessing sites 
and other industrial facilities where various organic chemicals were released into the 
environment and where DNAPL mixtures are present (Cohen and Mercer, 1993).  It is 
not unusual to find that chlorinated solvents represent a significant portion of mixed 
DNAPL sites. A mixed or multi-component DNAPL often displays different physical 
and chemical properties compared to a single component DNAPL (such as 
tetrachloroethene or PCE at a dry cleaning site). The Investigation Site (Chapter 3) 
represents the ideal case study of a mixed DNAPL site consisting predominantly of 
chlorinated solvents or chlorinated hydrocarbon DNAPLs.  
Production of DNAPLs began in the 19th Century with coal tars being produced at 
manufactured gas plants (MGPs) in Europe and the USA.  Chlorinated solvents were 
first produced in Germany in the 1800’s and in the USA early in the 20th Century. The 
rise in industrial activity after World War II saw a corresponding increase in use of 
DNAPLs, particularly the chlorinated solvents whose properties had tremendous 
industrial value. Pankow and Cherry (1996) record that from 1940-1980, 
approximately 2 billion pounds (approximately 907 000 tons) of chlorinated solvents 
was being produced in the USA alone. Figure 2-1 shows the production of selected 
DNAPLs in the USA within the period 1920-1998.   
Despite the wide production and use of DNAPLs since the 19th Century, the 
significance of these chemicals as soil and groundwater pollutants was only realised 
in the 1970’s (Schwille, 1988). Rivett et al. (1990) and Kueper et al. (2003) state the 
following reasons for this late awareness: 
• The general lack of understanding of the significance of groundwater as a 
supply of potable water;  
• The lack of suitable analytical equipment and methodologies for testing these 
compounds in low enough detection limits. Even with the advent of suitable analytical 
methods for the detection of DNAPL since the mid-1950’s; the lack of knowledge of 
6 
these chemicals as groundwater contaminants allowed scientists to concentrate their 
efforts on other known contaminants such as alkyl benzene sulphonate (ABS) 
detergents and aldrin; 
• The practices of disposing of DNAPL wastes directly onto the ground and into 
shallow soil systems; 
• The extreme difficulty of suspecting and detecting the presence of DNAPL in 
groundwater due to their lack of any noticeable taste and odour and of their higher 
densities relative to water.  
500000
DNAPL
Methylene Chloride
Tetrachloroethene
1,1,1-Trichloroethane
400000 Trichloroethene
Carbon Tetrachloride
Aniline
300000
200000
100000
0
0 0
92 93 94
0 509 96
0
97
0 809 99
0 00
1 1 1 1 1 1 1 1 20
Year
Figure 2-1: Selected DNAPL production in the USA from 1920-1998 (modified from 
Cohen and Mercer, 1993; and Moran, 2006)
While production of most chlorinated solvents have dropped to levels similar to that of 
the mid-20th Century (Figure 2-1) as a result of human health and environmental 
concerns (Pankow and Cherry, 1996), the extensive and widespread use of these 
chemicals continue to be problematic due to their pervasive and non-conventional 
behaviour in the subsurface and their continued use as chemicals of choice by users. 
Doherty (2000) notes the use of PCE as the preferred solvent of approximately 85-
90% of approximately 30 000 dry cleaners and launderers in the USA.  
7 
Selected DNAPL Production in the USA (tons)
Knowledge of detecting DNAPLs as well as the understanding of the fate and 
transport of DNAPLs in the various geo-contaminant settings has evolved over the 
last half century (Pankow and Cherry, 1996 and Sale et al. 2008).  The dominant 
theories and practices pertaining to DNAPLs that have evolved are shown in Figure 
2-2. 
2.1.1 DNAPL properties and behaviour in the subsurface 
DNAPLs can migrate in the subsurface as volatiles in soil gas, dissolved in 
groundwater and as a mobile separate “free” phase. According to the United States 
Environmental Protection Agency (1992), the major factors that control DNAPL 
migration in the subsurface include: 
• The volume of DNAPL released 
• The area of infiltration at the DNAPL entry point to the subsurface 
• The duration of release 
• Properties of the DNAPL such as density, viscosity and interfacial tension 
• Properties of the soil/aquifer media such as pore size and permeability 
• General stratigraphy such as the location and topography of low permeability 
units 
• Micro-stratigraphic features such as root holes, small fractures and slickensides 
found in low permeability layers (silt/clay layers). 
The various properties affecting the migration of DNAPL in the subsurface is explored 
at length in Mercer and Cohen (1990), Fetter (1999), Cohen and Mercer (1993), 
Gebrekristos (2007), etc. The influence of the properties of the DNAPL in its 
movement in the subsurface and the influence that the changes in aquifer properties 
have on this migration are briefly described in Section 2.1.1.1. Section 2.1.1.2 
focuses on the influence of a fractured rock setting has on the movement of the 
DNAPL as this is the primary geo-contaminant setting of the Investigation Site 
(Chapter 3).  
8 
•Reigning paradigm is that •There is widespread 
groundwater quality is 
Mid-1960s – evidence ofchlorinated largely unaffected by solvents as major Late 1980s-
anthropogenic influences. 
Organic compounds were 1970s groundwater contaminants. Attempts to present
thought to be the easiest have contamination 
to naturally deplete in the contained, remediated, 
subsurface, and hence •Toxic organic compounds •Focussed efforts to 
received very little detected in groundwater 
andprevented become the characterise DNAPLs, 
attention.  supplies between 1966 
primary focus. Use of increased remediation 
and 1972. chlorinated solvents peaks efforts with proper 
•Increased production of in the 1970s and then conceptual model 
DNAPLs •Trihalomethanes were the first VOCs’ detected in begins to decline. development approach. 
drinking water by the mid- •The use of DNAPLs show 
Prior mid- 1970’s. The production of a steady decline, however, 
DNAPLs continues to rise 
1960s 1970s-1980s large quantities are still and the main disposal produced and used. 
method remains direct 
disposal on land
Figure 2-2: Evolution in the knowledge and understanding of DNAPLs as a primary groundwater contaminant (modified from Pankow and Cherry, 1996; Sale et al.,
2008)
9 
2.1.1.1 DNAPL migration in the subsurface and the significance of heterogeneity 
DNAPLs enter the subsurface through release via spills, leaks or direct disposal onto 
the ground surface. According to Poulsen and Kueper (1992), once the DNAPL has 
entered the subsurface, the interaction between the following forces determines the 
principle flow direction of the DNAPL: 
• Gravity (also referred to as buoyancy or hydrostatic pressure): plays an 
important role in DNAPL migration in the first few meters of instantaneous release. 
• Capillary pressure: resists the migration of non-wetting DNAPL from larger to 
smaller pore openings in water-saturated porous media. Capillary pressure and 
geological structures such as bedding planes and fractures are the predominant 
variables that determine DNAPL migration in the subsurface. 
• Hydrodynamic pressure (also referred to as hydraulic or viscous force): can 
either promote or resist DNAPL migration. The effect of hydrodynamic pressure on 
DNAPL migration increases with decreased gravitational pressure, decreased 
capillary pressure and increased hydraulic gradient (Cohen and Mercer, 1993). 
Capillary pressure is determined by the difference in fluid pressure between the 
wetting and non-wetting fluid.  While wettability is measured or defined by the contact 
angle, this definition is complicated by the fact that the contact angle is hysteretic, 
displaying different values which are dependent on whether a fluid is advancing or 
receding over a surface (Adamson and Gast, 1997; Hiemenz and Rajagopalan, 
1997). There are recorded cases of environmental systems where DNAPL is not the 
non-wetting phase such as the Hill Air Force Base in the USA where the system is 
partially oil-wetting (Jackson and Dwarakanath, 1999). Experiments undertaken by 
Mohammad and Kibbey (2005) to test the effects of dissolution on the contact angle 
indicate that wettability is subject to temporal change should the dissolution reduce 
the contact angle between the DNAPL and the medium surface. The scenario where 
the DNAPL has dissolved and the contact angle with the media surface has 
approached zero may have an effect on remediation times (Mohammad and Kibbey, 
2005). 
This relationship between a water-NAPL system with water being the wetting phase 
is described mathematically in the equation below: 
 = ∆ =    Equation 2-1 
Where: 
Pc = Capillary pressure 
10 
Pnw = the pressure of the non-wetting fluid (DNAPL) 
Pw = the pressure of the wetting fluid (water) (Cohen and Mercer, 1993; Pankow and 
Cherry, 1996). The quantification of the capillary pressure function is an empirical 
relationship between capillary pressure and saturation, S (Thomas, 1982) in the form:  
 = ∆ = () Equation 2-2
For media with a range in pore sizes, the capillary behaviour of the media is a 
function of the media saturation. This relationship, shown in Figure 2-3, is dependent 
on the flow dynamics (Hassanizadeh et al., 2002). The capillary pressure – saturation 
relationship is hysteric. The curve is double-legged with one leg valid during the 
drying cycle (i.e. drainage) and the other leg valid during the wetting cycle 
(imbibition). A higher degree of non-wetting saturation can only be achieved with 
higher capillary pressure. 
Figure 2-3: Capillary pressure-saturation curve for porous media illustrating the relationship
between initial saturation along main drainage (points A, B, C) and residual saturation along
secondary wetting (from Kueper et al., 1993)
Sw = wetting phase saturation, SNW = non-wetting phase saturation
The relationship between capillary pressure and interfacial tension, contact angle and 
pore size is described by the following Laplace equation: 
 = ()⁄ Equation 2-3
Where: 
σ = interfacial tension between NAPL and water (N/m) 
11 
ϕ = contact angle. In a water-wet system ϕ is < 70o, while in a DNAPL-wet system ϕ is 
>110o
r = pore throat radius (m) (Mercer and Cohen, 1990; Cohen and Mercer, 1993). 
Capillary pressure increases with high interfacial tension, a small pore throat radius 
and small contact angle. Interfacial tension is the most important physicochemical 
property controlling multiphase flow in the subsurface. Interfacial tension allows non-
wetting DNAPLs to form globules in water and water-saturated media (Cohen and 
Mercer, 1993). 
At the stage when sufficient pressure is produced in the DNAPL to overcome the 
combined capillary pressure and water pressure, infiltration of the DNAPL into the 
porous medium will occur. The pathway taken by the DNAPL is determined by the 
pore throat size distribution (Poulsen and Kueper, 1992). In experimental work 
undertaken by Harrold et al. (2003) on United Kingdom (UK) soil types and shallow 
aquifer materials it has been shown that under strongly water-wetting conditions, 
small changes in the interfacial tension can have major effects on the DNAPL head 
required to overcome the medium’s capillary pressure.  
The pressure at which a DNAPL will invade a media continuously once the threshold 
value has been reached is referred to as the entry pressure. For most media the entry 
pressure corresponds to a water saturation of 0.8 – 0.95 (Pankow and Cherry, 1996). 
Recent studies (for e.g. Gooddy et al., 2002) have demonstrated that experimental 
entry pressure measurements are lower than those predicted by capillary pressure 
theory. This implies that infiltration of DNAPL into the media is more likely than has 
previously been thought. 
Relative permeability is a function of phase saturation and the most common manner 
of expressing this relationship for homogenous porous media are the X – curve and 
the Corey curve (Diomampo et al., 2002).  The X - curve describes relative 
permeability as a linear function of saturation while the Corey curve relates relative 
permeability to the irreducible or residual liquid and gas saturation. The relative 
permeability - saturation relations for wetting and non-wetting phases are hysteric 
(Figure 2- 4).  The relative permeability - saturation relationships quantify the relative 
permeability of the porous medium to a particular fluid phase at different fluid 
saturation levels.  
12 
Two-phase flow through fractures (either smooth or rough-walled) can be modelled 
by a porous medium approach (Diomampo et al., 2002). In this approach fractures 
are treated as two-dimensional porous media and a pore spaced occupied by one 
phase is not available for flow for another phase.  In this case flow is governed by 
Darcy’s Law and phase interference is represented by the relative permeability 
variable. 
Three-phase or multiphase relative permeabilities are required to describe the 
simultaneous movement of NAPL, water and air at a point (Cohen and Mercer, 1993).  
Chang et al. (2009) examines the relationship between relative permeability, 
saturation and capillary pressure in multiphase flow systems. The study found that 
the degree of connectivity of micro channels occupied by the wetting phase fluid 
could influence the relative permeability. 
(a) (b)
Figure 2- 4: Relative permeability – saturation curves for wetting phase (a) and non-wetting
phase (b) (from Vogler et al., 2001)
When a DNAPL enters the subsurface it can partition into the different phases.  As 
the DNAPL sinks through the unsaturated zone, a portion is trapped in the porous 
media at residual saturation (Sr) due to interfacial tensions and dead-end pores. 
Residual DNAPL forms discrete blobs or ganglia (Figure 2-5). The distribution of the 
residual DNAPL is not uniform or easily predictable in the subsurface due to minute 
variations in the media pore size distributions. According to Cohen and Mercer (1993) 
a non-wetting fluid is discontinuous at residual saturation, while the wetting fluid is 
not. Factors affecting wettability include DNAPL and aqueous phase composition, the 
presence or absence of organic matter, mineralogy, saturation history, aqueous pH 
(Dawson and llangasekasre, 1999; Barranco et al., 1997; Barranco and Dawson, 
13 
1999) and dissolution (Mohammad and Kibbey, 2005). Changes in wettability may 
play a significant role in affecting capillary pressure, relative permeability, residual 
saturation and the fluid displacement potential (Barranco et al., 1997).  
In the unsaturated zone, NAPL spreads as a film (Wilson et al., 1990; Cary et al., 
1989) between the water and gas phases given a positive spreading coefficient (∑). 
Halogenated solvents typically have negative spreading coefficients and hence will 
not spread in the unsaturated zone as a result of the internal cohesion (Wilson et al., 
1990).  Residual saturation values in the unsaturated zone range from 0.10 to 0.20.  
Soil grain NAPL
Water
Figure 2-5: NAPL in the pore space of a granular porous
media (from Wilson et al., 1990)
In the saturated zone, residual saturation values are larger than those of the 
unsaturated zone for the following reasons (In Cohen and Mercer, 1993 from 
Anderson, 1988): 
• The fluid density ratio (NAPL:air vs NAPL:water above and below the water table, 
respectively); 
• The NAPL is the non-wetting fluid in most saturated media and is hence trapped 
in the larger pores; 
• NAPL is the wetting fluid (with respect to air) in the unsaturated zone and tends to 
spread into adjacent pores and leaves a lower residual content behind.  
Lenhard et al. (2004) found that NAPL in small pores in the unsaturated zone can be 
considered immobile because: 
• the pore dimensions are small 
• there are no continuous passages through the channels to conduct the NAPL 
• NAPL adjacent to water films that are adsorbed to solid particles do not contribute 
to NAPL advection, especially if the water film is thin 
14 
The formation of residual NAPL through imbibitions in the unsaturated zone is 
illustrated in Figure 2-6. 
Residual saturation values in the saturated zone range from 0.10 to 0.50. The 
mechanisms that facilitate the trapping of residual NAPL in the saturated zone are 
documented by Chatzis et al. (1983) and Wilson et al. (1990) and include the snap-off 
mechanism and the bypassing mechanism. Snap-off occurs in cases where there is a 
high aspect ratio between the pore body and the pore throat. This results in single 
droplets or blobs or residual NAPL. Bypassing occurs when the wetting fluid flow 
disconnects the non-wetting fluid. This causes ganglia to be trapped in clusters of 
large pores surrounded by smaller pores. As a result of the snap-off and bypass 
mechanisms, residual saturation increases with increasing pore aspect ratios and 
pore size heterogeneity and with decreasing porosity (Chatzis et al., 1983; Powers, et 
al., 1992).  
Residual DNAPL is immobile under normal subsurface conditions (assuming that the 
equilibrium conditions holding the residual phase DNAPL are unchanged).  However, 
residual DNAPL plays a significant role as long-term sources of contaminants for 
continued dissolution of contaminants into water and air in adjacent pores.  
(a) (b)
Figure 2-6: Depiction of NAPL imbibition in the pore spaces. (a) shows small
pores and pore wedges filled with NAPL before the larger pore bodies are filled.
After these smaller spaces are filled, the NAPL is assumed to be immobile (red).
(b) depicts the pore system is NAPL-filled with the mobile (free) NAPL shown as
pink. Water is blue and solid particles are grey (from Lenhard et al., 2004)
Residual DNAPL is exposed to air allowing for volatilisation across air-DNAPL 
interfaces; and to water allowing for dissolution into infiltrating water across DNAPL-
water interfaces in the unsaturated zone. Dissolved components in soil moisture 
become available for partitioning across air-water interfaces. A mathematical model 
15 
for the prediction of residual phase NAPL in the air-NAPL-water phase was 
developed by Lenhard et al. (2004). According to Kueper et al. (2003), the absence of 
chlorinated solvents such as TCE and PCE (DNAPL phase) in the unsaturated zone 
is not conclusive in determining whether a past release occurred at a site or whether 
the release failed to reach the water table. The absence of the DNAPL phase could 
be a result of vaporisation processes which can deplete residual chlorinated solvents 
within a few years in warm and dry climates. The presence of vapour, adsorbed or 
aqueous phases however remains. Hughes et al. (1990) and Jellali et al. (2001) show 
through field experiments using TCE that shallow groundwater contamination can 
form through vapour transport through advection in the unsaturated zone and 
dissolution into groundwater. Jellali et al. (2001) also show experimentally that mass 
transport of DNAPL in the unsaturated zone is also affected by mass transfer from 
groundwater to the unsaturated zone through volatilisation of dissolved components. 
DNAPL with a higher Henry’s Law Constant display higher volatilisation.  
When in contact with the capillary fringe, DNAPLs will tend to migrate laterally 
(Schwille, 1988). At this point the DNAPL will accumulate until gravitational forces at 
the base of the mass exceed the threshold entry pressure of the underlying saturated 
zone. The entry pressure for the DNAPL is relative to the pore size. DNAPLs tend to 
migrate into the largest pores first (Kueper et al., 1989; Kueper and McWhorter, 
1991). Once the threshold entry pressure is exceeded, the DNAPL will displace the 
water and continue migrating under capillary and gravitational forces. Zhang and 
Smith (2002) show that the fingering process in a homogenous porous media can be 
divided into two stages: the DNAPL finger initiation stage and the finger elongation 
stage. They also show that fingering of a small amount of DNAPL in homogenous, 
porous media will penetrate to a much larger depth than previously predicted using 
classic flow equations, consequently overestimating the average DNAPL content 
while underestimating the depth of DNAPL penetration. 
If the DNAPL encounters a bowl-shaped stratigraphic trap it may be immobilized as a 
reservoir of continuous immiscible fluid. If no stratigraphic traps are present, the 
DNAPL will migrate according to the slope of the capillary barrier layer, which might 
be a different direction from that of the hydraulic gradient (Cohen and Mercer, 1993). 
DNAPL in unconsolidated media can also pool or mound above fine-grained horizons 
such as silt and clay units (Kueper et al., 1993; Feenstra et al., 1996). The maximum 
pool height is inversely proportional to the permeability of the capillary barrier unit 
(Durnford et al., 1997; Kueper et al., 2003).  However, in most instances preferential 
16 
pathways such as fractures, root holes etc. are present within these low permeability 
units which allow for the migration of DNAPL (Kueper and McWhorter, 1991; O’Hara 
et al., 2000; Parker et al., 2004). The architecture of DNAPL in the saturated zone 
depends on the heterogeneity of the stratigraphic units (Illangasekare et al., 1995)
and can exist as pools, disconnected globules and ganglia, or immobilised in 
stratigraphic traps and residual saturation (Figure 2-7). Irrespective of whether the 
DNAPL is mobile or immobile in the saturated zone, it can still dissolve and 
persistently contaminate groundwater. The sparsely distributed ganglia and pools of 
DNAPL make it very difficult to find (Pankow and Cherry, 1996; Parker et al., 2003). 
Stratigraphic or geological heterogeneity also has a huge bearing on the mechanisms 
by which contaminants are stored and released from the source zone to 
downgradient plumes.   
DNAPL will dissolve over time within a hydraulically transmissive layer. Initially, the 
chemical gradient will be sustained via transport of the dissolved DNAPL constituent 
away from the source zone. This process occurs through the following mechanisms: 
• Transverse diffusion into the groundwater and subsequent horizontal 
advection along the top of the source 
• Advective transport through the source 
• Matrix diffusion into low permeability zones  
The DNAPL constituents are stored as dissolved phase in water and as a sorbed 
phase on or in solids in these low permeability zones. Sorption acts as a sink that 
accelerates DNAPL dissolution. The aqueous phase or dissolved phased plume is 
typically the most mobile form of DNAPL contamination and most investigation efforts 
are often focussed on determining the likely risks of the migrating phase on identified 
receptors.  
DNAPLs are only slightly soluble in water. This implies that they persist as long-term 
sources of contaminants. Most sites consist of complex or mixed contaminants. In 
these cases, the properties of an individual component in the mixture vary from those 
of the pure component. The term “effective solubility” is used to describe the solubility 
of a particular component in a complex mixture. As dissolution proceeds, the 
composition of the NAPL changes (Mackay et al., 1991). For example Melber et al. 
(2004) document that as the dissolution of creosote proceeds; the more soluble 
components will be rapidly lost. This causes the mole fraction (and hence the 
effective solubilities) of the other constituents to increase. The properties of a mixture 
are determined by the properties of its pure components and their concentrations in 
17 
the mixture. Hence, chemicals of interest behave ideally in the matrix containing 
them. Under these conditions, the concentration of a chemical in the aqueous phase 
is proportional to the mole fraction of the chemical in the phase corresponding to 
Raoult’s Law. The following expression can be used to predict the concentrations of a 
chemical in a complex mixture in the aqueous phase (Schwarzenbach et al., 1993):
 =   Equation 2-4 
Where: 
Cw = chemical’s concentration in the aqueous phase (mol/L) in equilibrium with the 
organic phase or the effective solubility  
X0 = mole fraction of the chemical in the NAPL phase 
S = aqueous solubility of the pure liquid chemical (mol/L) 
The mole fraction can be calculated using the following equation: 
 =   � Equation 2-5 
Where: 
MFx = mass fraction of the selected organic compound in the mixture 
MW0 = average molecular weight of the mixture 
MWx = molecular weight of the selected compound 
The octanol-water partition coefficient (Kow) is used to describe the degree to which 
an organic substance will preferentially dissolve in water or an organic solvent. The 
larger the octanol-water partition coefficient, the larger the tendency of the organic 
substance to dissolve in octanol rather than in water and will hence be less mobile in 
the environment  (Fetter, 1999).  
Figure 2-7: Simplified illustration of the existence of different phases of DNAPL within a
heterogeneous setting (from Sale et al., 2008)
18 
Matrix diffusion is the process whereby contaminants move into low permeability 
layers. Matrix porosities of clayey deposits range from 30-70% and between 5-25% 
for sedimentary deposits (Freeze and Cherry, 1979). Matrix diffusion into low 
permeability layers can sustain dissolved plumes in transmissive zones after the 
DNAPL source zone is depleted (Sudicky et al., 1985; Parker et al., 1994; Parker et 
al., 1997; Liu and Ball, 2002; Chapman and Parker 2005; Air Force Center for 
Engineering and the Environment, 2007) It can also attenuate the rate at which a 
plume advances in fractured rock and can result in delayed breakthrough curves 
(Kueper and Davies, 2009), commonly referred to as “pulses” (Figure 2-8).  
The storage of contaminants through matrix diffusion and the consequent back 
diffusion of these chemicals into transmissive zones is illustrated in Figure 2-9. Four 
primary factors control the DNAPL mass stored within the low permeability zone: 
• the duration of DNAPL present at the interface between the transmissive and 
the low permeability zones; 
• the solubility of the DNAPL constituents; 
• The amount of adsorption to solids; 
• The rates of biotic and abiotic degradation of contaminants in the low 
permeability zone (Air Force Center for Engineering and the Environment, 2007). 
Figure 2-8: Illustration of matrix diffusion of dissolved phase contaminants
adjacent to DNAPL source and along length of plume in a fracture. Matrix
diffusion can attenuate the rate of plume advance in fractured rock
(concentration vs distance plot), and can result in delayed breakthrough
curves (concentration vs time figure) (from Kueper and Davies, 2009)
19 
The total chemical mass in a DNAPL source zone consists of the following 
constituents: DNAPL, dissolved constituents, vaporised constituents (gas phase) and 
sorped phased (Cohen and Mercer, 1993).  
Immediately after a spill or release, free phase DNAPL will be the largest component 
of the contaminant mass in the source zone. Over time, the DNAPLs dissolve and 
develop an aqueous phase plume. Mass is also transferred into vapour and sorbed 
phases. During the late stage, little or no DNAPL will remain and the plumes are 
sustained via release of contaminant mass through back diffusion from low 
permeability zones. The mass stored in the aqueous phase and the sorbed phase is 
referred to as “non-DNAPL source mass” (Air Force Center for Engineering and the 
Environment, 2007). The evolution of a DNAPL (chlorinated solvent) release is 
illustrated in Figure 2-10. 
Diffusion of contaminants into a low Diffusion of contaminants out of a low 
permeability zone beneath a DNAPL pool and permeability zone after complete DNAPL 
the downgradient plume (arrows depict the dissolution (arrows depict the movement of 
movement of dissolved DNAPL) dissolved DNAPL) 
Figure 2-9: Significance of matrix diffusion (from Sale et al., 2008)
According to Sale et al. (2008) the key factors that control the rate which a release 
ages include: 
• The amount of DNAPL release; 
• The solubility of the constituent of DNAPL; 
• The rate of groundwater flow; and 
• The architecture of transmissive and low permeability zones. 
20 
EARLY STAGE MIDDLE STAGE LATE STAGE
Figure 2-10: Evolution of a DNAPL release from early to late stages (from Sale et al., 2008)
2.1.1.2 DNAPL migration in fractured and dual porosity media 
Fracture networks are ubiquitous and dominate the hydrogeological setting of South 
Africa, having been formed by the deformation of rocks and soils through processes 
such as tectonic forces, subsurface subsidence or thermal expansion.  Groundwater 
flow through a fracture network is strongly influenced by the geometry of fractures.  
Fracture flow is flow through a dual porosity system which includes the porous matrix 
and the fracture network. The degree of interconnection between the matrix and the 
fracture network defines the character of the flow domain and is a function of bulk 
hydraulic conductivity which is a function of the fracture network distribution, the 
matrix hydraulic conductivity, the fracture orientation, connectivity and apertures.   
In unsaturated fractured rock, the hydraulic conductivity of a fracture decreases with 
aperture size (Wang and Narashiman, 1985; Martinez et al., 1992) as a result of 
capillary forces draining the largest pores in the unsaturated zone first and 
transforming them into flow barriers. In saturated fractured rock, hydraulic 
conductivity of a fracture increases with the fracture’s aperture. The cubic law (Snow, 
1965) describes laminar flow in fracture between two smooth parallel plates and the 
hydraulic conductivity of the fracture is proportional to the square of the fracture 
aperture. In natural formations where the fracture aperture is variable and not 
constant, the cubic law is not valid. Variable apertures result in flow being channelled 
into preferential flow paths along zones with the largest interconnected apertures (for 
e.g. Tsang et al., 1991; Murphy and Thomson, 1993; Birkholzer and Tsang, 1997; 
Ge, 1997; Oron and Berkowitz, 1998; O’Hara et al., 2000).  
DNAPL can enter fractures in the unsaturated zone and in the saturated zone 
(Kueper et al., 2003).  Residual DNAPL and DNAPL pools will form in fractured 
bedrock.  
21 
The fracture entry pressure is directly proportional to the interfacial tension and 
inversely proportional to fracture aperture; resulting in preferential DNAPL migration 
through the larger aperture fractures of a fracture network. According to Kueper and 
McWhorter (1991), the critical aperture controlling NAPL flow through a variable 
aperture is the aperture corresponding to the entry pressure. Steele and Lerner 
(2001) state that the critical aperture corresponds to the smallest aperture 
constriction where NAPL first forms a fully connected pathway across a variable 
aperture network. If static equilibrium is assumed, the capillary pressure immediately 
above the fracture can be expressed as the height of the pooled DNAPL (Kueper and 
McWhorter, 1991). Figure 2-11 depicts a fractured network where the vertical extent 
of migration has been stopped due to the narrowing of the fracture aperture at point 
‘A’. In this case the capillary pressure at the top of the DNAPL pool is zero. According 
to Kueper and McWhorter (1991) and Kueper et al. (2003), only those fractures that 
have entry pressure less than the capillary pressure of the DNAPL-water system are 
invaded. Under hydrostatic conditions, progressively smaller aperture fractures with 
depth can be invaded. The relationship between the vertical height of the DNAPL 
pool (H) and the fracture aperture at point ‘A’ is given by the following equation: 
 =   ( − ) Equation 2-6 
Where: 
H = vertical height of pooled DNAPL 
σ = DNAPL-water interfacial tension 
Ф = contact angle 
ρN = DNAPL density 
ρW = groundwater density 
g = acceleration due to gravity 
e = fracture aperture 
A plot of the fractured aquifer required to stop vertical migration versus depth for a 
range of DNAPLs is shown in Figure 2-12. The graph shows that for the same height 
of the different DNAPL compositions, the decrease in aperture size for a 
corresponding chlorinated solvent must be very quick with depth compared to the 
same pool height of coal tar DNAPL. According to Kueper et al. (2003), vertical 
migration to considerable depths at sites with chlorinated solvent DNAPLs might have 
stopped by the time the site is investigated due to the very high densities and low 
viscosities.  Larger pool heights, lower viscosities and higher densities allow for 
quicker penetration of the bedrock fractures (Pankow and Cherry, 1996). 
22 
Figure 2-11: Pooling of DNAPL in a fractured network. If hydrostatic equilibrium is assumed,
then the capillary pressure increases linearly with depth (from Kueper et al., 2003)
Figure 2-12: Fracture aperture required to stop DNAPL migration
versus the height of the accumulated DNAPL (from Kueper et al.,
2003)
Other factors that influence the migration of DNAPL include: 
• Interfacial tension – A system with a low interfacial tension will show higher 
vertical movement through a low permeability unit as compared to a higher interfacial 
tension system (Figure 2-13) 
23 
Figure 2-13: Illustration of differing migration pathways for a low interfacial tension system (a)
versus a high interfacial tension system (b) (from Pankow and Cherry, 1996)
• Fracture density, orientation and connectivity – DNAPLs migrate through the 
pathways of least capillary resistance. Hence, the fracture orientation and 
connectivity will influence the direction in which the DNAPL migrates as compared to 
the hydraulic gradient. Significant fracture density and interconnectivity in a confining 
layer such as clay can allow for vertical migration of DNAPL as well as leakage of 
dissolved contaminants (Harrison et al., 1992). Small scale heterogeneities can 
influence flow paths of the DNAPL, requiring adequate fracture characterisation 
(Dennis et al., 2010). 
In aquifer systems with dual porosity where the matrix porosity is high, matrix 
diffusion (see Section 2.1.1.2.) plays an important role. Dissolved contaminants are 
subject to diffusion out of and into the rock matrix.  According to the United States 
Environmental Protection Agency (2001) and Parker (2007), discrete fracture 
pathways (as compared to the total fracture network) play a significant role in 
hydrogeological investigations in fractured rock. As only some subsets of open 
fractures will have active groundwater flow, the challenge in application of 
characterisation technologies is to locate the significant fractures and apply these 
technologies so that they take measurements that are representative of in situ
conditions. The location of the DNAPL plume is hydraulically downgradient of the 
DNAPL source zone. Depending on the biogeochemistry environment and the 
processes involved, the chemical composition of the plume may be different from the 
DNAPL in the source zone. According to the National Research Council (2004) 
groundwater plumes tend to have larger spatial extent and are more continuous when 
compared to the contaminant mass distribution within source zones. As a result of the 
architecture of DNAPL source zones, irregular and stratified plumes are created. 
24 
2.1.2 Transformation of DNAPLs 
Three processes affect solute contaminant transport: 
• Physical processes (dispersion, diffusion, dilution and volatilisation); 
• Geochemical processes (sorption, abiotic reactions); and 
• Biotic processes (biodegradation).  
According to the United States Environmental Protection Agency (1998) the DNAPL 
form does not easily attenuate and it is therefore important to determine whether the 
site has DNAPL in order to account for this during remediation planning and 
implementation. The dominant attenuation processes at DNAPL contaminated sites 
are shown in Table 2-1.  
DNAPLs can undergo non-destructive attenuation through processes such as 
dilution, dispersion, diffusion, sorption and volatilisation. These processes can reduce 
the contaminant concentrations in groundwater and in the case of volatilisation can 
also reduce contaminant mass. However, these processes rarely have an effect on 
the contaminant toxicity or mass. The mass and composition of DNAPLs can be 
affected through destructive attenuation mechanisms. These are discussed in more 
detail in the following sections. 
2.1.2.1 Transformation processes of halogenated DNAPLs 
Halogenated DNAPLs (such as chlorinated hydrocarbons/solvents) are the most 
common types of DNAPLs found at industrial sites. The Investigation Site (Chapter 3) 
contains chlorinated hydrocarbons as the predominant contaminants of concern; 
hence this section will focus on the transformation processes that affect these 
compounds. 
Biotic and abiotic processes lead to the transformation of these compounds in the 
subsurface. Biotic transformations can lead to more toxic daughter or intermediate 
products (such as cis-1,2-dichloroethene and vinyl chloride) and can further 
complicate characterisation (Cheremisinoff, 1990). Recently a large amount of focus 
has been paid to destructive attenuation mechanisms as a remediation technology 
(for e.g. United States Environmental Protection Agency, 1999; National Research 
Council , 2000; Newell and Aziz, 2004; Pretorius, 2007; Usher et al., 2008) and to the 
applications of molecular biological tools (Stroo et al., 2006).  It was previously 
thought (Robertson and Alexander, 1996) that biotransformation close to a 
25 
chlorinated solvent source was unlikely due to the high toxicity of the DNAPL 
contaminant concentrations. Several authors have however reported microbial 
activity at concentrations at or near the aqueous solubility of PCE (for e.g. Yang and 
McCarty, 2000; Cope and Hughes, 2001). According to laboratory studies undertaken 
by Yang and McCarty (2002) and Cope and Hughes (2001), microbial activity can 
increase DNAPL dissolution by a factor of 5 or more.  
Biotic transformations are typically faster than abiotic transformations; provided that 
there is sufficient substrate, nutrients and a microbial population that can facilitate the 
transformation (Vogel et al., 1987). The most common abiotic reactions are 
hydrolysis and dehydrohalogenation. While most abiotic reactions are slow; they can 
still be significant relative to groundwater movement rates. Intermediates formed 
during abiotic transformation tend to accumulate and commonly require biotic 
processes to stimulate the abiotic reactions (Wiedemeier et al., 1999). The 
transformation process affecting halogenated DNAPLs can be divided into two 
categories: (i) reactions that require external electron transfer (oxidation reactions 
such as epoxidation and reduction reactions such as coupling); and (ii) reactions that 
do not (substitution reactions such as hydrolysis and dehydrohalogenations). The 
definitions of the various types of transformations are provided in Table 2-2.   
If the primary transformation pathway for halogenated DNAPL contaminated 
groundwater is a physical process such as volatilisation, then the Henry Law’s 
Constant would be the determining factor of the evolution of the groundwater 
composition (Cho et al., 2004). An example of the transformation of groundwater 
impacted with 1,1,1-Tricholoroethane (1,1,1-TCA), 1,1-Dichloroethane (1,1-DCA) and 
1,1-Dichloroethene (1,1-DCE) is provided in Figure 2-14. This figure illustrates that if 
the physical process (sorption and volatilisation) are the primary transformation 
process occurring in mixed chlorinated solvents impacted groundwater, then the 
groundwater will become increasingly enriched in 1,1,1-TCA. However, with the same 
starting composition, the groundwater becomes enriched with 1,1-DCA under biotic 
transformation processes or 1,1-DCE under abiotic transformation process as shown 
in Figure 2-15 (Vogel and McCarty, 1987; Cho et al., 2004).  
26 
Table 2-1: Dominant attenuation mechanisms for DNAPL groups (In Pretorius, 2007 from Carey et al., 2000) 
DNAPL Group Example Non-Destructive attenuation Destructive attenuation mechanisms
mechanisms
Chlorinated Solvents PCE,  TCE, TCA × × × × × × ×× × ×
DCM, VC, DCE × × × × ×× × × × ×
Creosote Phenols and × × × ×× ×× × ×
phenolics 
Pesticides Chlorinated × × × × × × ×
pesticides, 
organophosphate 
Where: 
× = of primary importance 
×× = of secondary importance 
27 
SSoorrppttiioonn
DDiissppeerrssiioonn
DDiiffffuussiioonn
VVoollaattiilliissaattiioonn
AAeerroobbiicc ddeeggrraaddaattiioonn
((CCoonnttaammiinnaanntt aass eelleeccttrroonn ddoonnoorr))
AAnnaaeerroobbiicc ddeeggrraaddaattiioonn
((ccoonnttaammiinnaanntt aass eelleeccttrroonn aacccceeppttoorr))
RReedduuccttiivvee ddeehhaallooggeennaattiioonn
FFeerrmmeennttaattiioonn
CCoo--mmeettaabboolliissmm
OOxxiiddaattiioonn//rreedduuccttiioonn
Table 2-2: Biotic and abiotic transformation reaction definitions (modified from Vogel et al., 1987)  
Reaction Term Definition
A reaction in which two alkyl groups or aryl groups connect together 
Coupling E.g. 2CCl4 + 2e
-    CCl3CCl3 + 2Cl
-
Elimination of HX to form an alkene
Dehydrohalogenation E.g. CCl3CCl3  CCl2CH2 + HCl 
Reductive elimination of two halide substituents to an alkene
Dihalo-elimination E.g. CCl3CCl
- -
3 + 2e  CCl2CH2 + 2Cl
Reaction in which an epoxide is generated
Epoxidation E.g. CHClCl2 + H2O  CHClOCCl2 + 2H
+ + 2e-
A reduction reaction in which a carbon-hydrogen bond is broken and hydrogen replaces halogen substituent
Hydrogenolysis E.g. CCl4 + H + 2e
-  CHCl3 + Cl
-
Addition of a hydroxyl group
Hydroxylation E.g. CH3CHCl2 + H2O  CH3CCCl2OH + 2H
+ + 2e-
Reaction in which the solvent serves as the nucleophile (i.e. a reacting species providing an electron pair)
Solvolysis E.g. CH3CH2CH2Br + H2O  CH3CH2CH2OH + HBr 
28 
Figure 2-14: Sorption and volatilisation transformation pathways (from Cho et al., 2004)
Figure 2-15: Biotic and abiotic transformation pathways (from Cho et al., 2004)
29 
Recent developments in the use of compound specific isotope analysis (for e.g. 
Hunkeler et al., 2005; Liang et al., 2007) has allowed for distinguishing abiotic from 
biotic transformations. Compound specific isotope analysis (CSIA) is increasingly 
becoming a useful tool to assess in situ biodegradation and biodegradation pathways 
of chlorinated solvents. The theory and application of CSIA is described in more 
detail below (Section 2.1.2.2).  
Degradation pathways for chlorinated ethanes and ethenes via reductive 
dechlorination, dichloroelimination and dehydrohalogenation are shown in Figure 2-
16.  Under reducing or anoxic conditions, chlorinated ethenes and ethanes can be 
sequentially dechlorinated by microorganisms (Ballapragada et al., 1997; de Best et 
al., 1997; de Best et al., 1999; Adamson and Parkin, 2001; Major et al., 2002; 
Hunkeler et al., 2005). Phylogenetic groups of bacteria that are capable of metabolic 
reductive dechlorination for chlorinated ethenesinclude Dehalobacter,
Sulfurospirillum, Desulfuromonas, Desulfitobacterium, Clostridium and 
Dehalococcoides.  A limitation is that not all organisms can allow the complete 
dechlorination to ethene, causing the subsurface accumulation of toxic intermediates 
such as dichloroethene and vinyl chloride.  The most promising phylogentic group of 
bacteria in the reductive dechlorination of chlorinated ethenes has been identified to 
be Dehalococcoides (He et al., 2003).  According to McCarty (1997) reductive 
dechlorination of tetrachloroethene (PCE) to TCE only occurs under methanogenic 
conditions and will not occur in nitrate-reducing zones. Stiber et al., (1999) propose a 
Bayesian Belief Network (BBN) framework for the reductive dechlorination of TCE 
(Figure 2-17). A BBN is a graphical probabilistic technique. BBNs are useful for the 
modelling of causative networks. Various authors have observed the reductive 
dechlorination of PCE, TCE and carbon tetrachloride (CCl4) under anaerobic 
conditions to less-chlorinated products (de Best et al., 1999; Adamson and Parkin, 
2001; Major et al., 2002); however, the rate of dechlorination and the transformation 
products vary widely (Middeldorp et al., 1999; Istok et al., 2007). Push pull tests have 
been undertaken in order to measure the in situ rates of reductive dechlorination of 
chlorinated ethenes using injected reactive tracers such as trichlorofluoroethene 
(TCFE) as a reactive tracer for trichloroethene (TCE) and cis-dichlorofluoroethene 
(DCFE) for cis-DCE (Hageman et al., 2001; Ennis et al., 2005; Field et al., 2005; Istok 
et al., 2007; Taylor et al., 2007). These reactive tracers have similar chemical 
structures and reactivity to the targeted contaminants and are commonly referred to 
as “surrogates”. Other parameters that can be measured to determine whether biotic 
30 
degradation is occurring include: redox potential, dissolved oxygen, iron (II), 
methane, pH, sulphate, nitrate and chloride.
1,1,2-Trichloroethane and 1,2-dichloroethane may also be transformed by 
dichloroelimination (a biotic process) to vinyl chloride and ethane respectively (Belay 
and Daniels, 1987; Chen et al., 1996; Klecka et al., 1998). Dehydrohalogenation is an 
abiotic transformation process which leads to the formation of TCE from 1,1,2,2-
tetrachloroethane, dichloroethane from 1,1,2-trichloroethane and vinyl chloride from 
1,2-dichloroethane (Jeffers et al., 1989; Pagan et al., 1998).  
Figure 2-16: Degradation pathways of chlorinated ethenes and ethanes. Bold arrows,
reductive dechlorination; fine arrows, dichloroelimination; dashed arrows,
dehydrohalogenation (from Hunkeler et al., 2005)
31 
Figure 2-17: Bayesian Belief Network for the reductive dechlorination
of TCE in groundwater (from Stiber et al., 1999)
2.1.2.2 Compound specific isotope analysis (CSIA) 
The last two decades have seen a proliferation in the use of isotopes in solving 
hydrogeochemical problems (Bottcher et al., 1990; Clark and Fritz, 1997; 
International Atomic Energy Agency, 2002). Stable isotopes that are useful in 
environmental studies include hydrogen, carbon, nitrogen, oxygen, sulphur and 
chlorine. Stable isotopes can be subdivided into the light stable isotopes such as 13C, 
2H, 37Cl and the radiogenic stable isotopes that are produced by the decay of 
radioactive elements with long half-lives such as 87Sr which is produced from the 
decay of 87Rb. The deviation of the stable isotope value of the sample from the 
international standard (V-PDB for carbon, V-SMOW for hydrogen and the standard 
abundance of the heavier chlorine isotope (37Cl) is referred to as 24.47% of total 
chlorine) will be either negative or positive. A negative value means that the sample 
is depleted in its heavy isotope content while a positive delta indicates that the 
sample is enriched in its heavy isotope content (Clark and Fritz, 1997).  
Isotope ratios can change due to biological and chemical processes which cause 
isotope fractionation. Stable isotope fractionation is the term used to refer to the 
change in stable isotope ratios. Slater et al. (1999), Wang and Huang (2003), Mancini 
et al. (2003), Kopinke et al. (2005) show that changes in isotope ratio resulting from 
processes such as volatilisation, dispersion and sorption are significantly smaller 
32 
(<2‰) due to these processes being isotopically conservative when compared to the 
changes in isotope ratio as a result of biodegradation (>2‰). Large fractionations 
(>10‰) are associated with biotic and abiotic degradation as compared to dissolution 
and volatilisation (Slater et al., 1998). During these processes, molecules containing 
the lighter isotopes exclusively (for e.g. 12C, 1H, 35Cl) will react more rapidly when 
compared to molecules containing the heavier isotope (for e.g. 13C, 2H, 37Cl), due to 
the lighter isotopes having weaker bonds.  As the reaction continues there is a shift in 
the ratio of the lighter isotope to the heavier isotope. The extent to which isotope 
fractionation occurs during the process of producing the organic compounds induces 
an isotope composition fingerprint that can provide evidence for the identification of 
sources, types of transformation reactions and sinks of organic compounds 
(Meckenstock et al., 2004). 
Stable isotope fractionation can be expressed as the stable isotope fractionation 
factor (α). This expression is described below: 
 = ⁄ =  + � +  Equation 2-7
Where:  
hE= heavy isotope of a given element E 
R = stable isotope ratio of the compound; and 
the subscripts a and b may represent a compound at time zero (t0) and at a later point 
(t) in a reaction; or a compound in a source zone, versus a down gradient well 
(Hunkeler et al., 2008). 
Stable isotope fractionation during biotic and abiotic degradation can also be (and 
commonly is) quantitatively described using the Raleigh equation, originally 
developed for homogenous batch systems (Clark and Fritz, 1997): 
 = (−) Equation 2-8
Where: 
R0 is the initial isotope value of the compound  
f is the ratio (C/C0) of the compound at time t and zero. 
In flow-through studies the application of the Raleigh equation leads to an 
underestimation of the amount of biodegradation due to physical and chemical 
heterogeneities (Kopinke et al., 2005). The Raleigh equation can be rearranged to 
form the following expression: 
 = −⁄ Equation 2-9
33 
The relationship between the change in isotopic composition and contaminant 
reduction is expressed as the isotopic enrichment factor (ε) and is defined as: 
 = (  ) ×  Equation 2-10
The larger the fractionation during the reaction, the more negative is the 
corresponding value of epsilon (ε) (Hunkeler et al., 2008). Loglinear regression of the 
stable isotope ratio versus decreasing concentration results in a straight line. The 
slope of this line denotes ε. 
Depending on the underlying reaction mechanism, isotope fractionation occurs 
according to a kinetic isotope effect (KIEE) for the element (E) at the reacting bond(s). 
Hence, isotope enrichment factors are characteristic for a given degradation pathway. 
The differing energy of activation between the ground state of the isotopic reactants 
versus their transition states are as a result of the different KIEs (Melander and 
Saunders, 1980). This is because the different KIEs reflect the different reaction rates 
of bonds containing light and heavy isotopes or the distinction in their frequency of 
vibrational energy; i.e. 

 =     Equation 2-11 
Where: 
l is the light isotope and 
h is the heavy isotope. 
KIEs are large if bonds are broken or formed in the rate limiting step, the nature of the 
bond broken and the reaction mechanism.  
Isotope enrichment factors are calculated assuming the observed isotope 
fractionation in the bulk organic compound is a result of intermolecular isotopic 
competition between molecules containing exclusively n isotopically light atoms of 
element E. While the bulk ε values are used to assess the fractionation of entire 
molecules, it is not suitable to comparing isotope effects and reaction mechanisms 
among different compounds. For this type of comparison, the KIEs at the reacting 
bond are necessary (Elsner et al., 2005). Isotope enrichment factors can be 
converted into apparent kinetic isotope effects (AKIE) by calculating the isotope 
enrichment factors at the reactive position of the molecule (εreactive position): 
 +⁄.∆+ =   +  .    Equation 2-12 
34 
The reactive position is the position in the molecule where the heavy isotope is 
located and where the initial enzymatic bond transformation takes place (Cook, 1991; 
Melander and Saunders, 1980). The AKIE is often lower than the KIE and can be 
calculated with the following equation (Elsner et al., 2005):  
 ≈ +  Equation 2-13 
Recent years have seen an increase in the use of compound specific isotope 
analysis (CSIA) that have been used to quantify biodegradation of several organic 
compounds such as MTBE, BTEXs, chlorinated solvents and PAHs (Sturchio et al., 
1998; Meckenstock et al., 1999; Hunkeler et al., 1999; Hunkeler et al., 2002; Griebler 
et al., 2004; Meckenstock et al., 2004; Elsner et al., 2005; Fischer et al., 2008).  
The uses of CSIA include the determination of: 
• which contaminants are being degraded;  
• the extent of biodegradation;  
• geochemical processes that are involved (for e.g. aerobic, sulphate reducing, 
etc.);  
• contaminant degradation pathways;  
• first order biodegradation rates; and 
• contaminant mass loss. 
Comparison of the site-specific ε against published values can provide information on 
the type of reaction and organism involved at the site. The isotope enrichment factor 
can be used to determine the progress of degradation at sampling time t (Hunkeler et 
al., 2005).   
Until recently CSIA analysis of chlorinated hydrocarbons has been limited to the use 
of C isotopes. However, chlorine and hydrogen isotopes are being increasingly used 
(for e.g. by Shouakar-Stash et al., 2003; Abe et al., 2009). Dual or two-dimensional 
isotope approaches are useful: 
• in rationalising the variations observed in the enrichment factors 
• to identify reaction mechanisms of organic contaminants 
• to quantify the relative contribution of two reaction pathways 
• to distinguish between variations between commercial products and 
• to distinguish between different sources at the same field site (Hunkeler et al., 
2001; Mancini et al., 2003; Sturchio et al., 2007). The dual isotope approach 
35 
works on the premise that through measurement of isotope ratios for two 
elements of a compound simultaneously (for e.g. δ13C and δ2H or δ13C and δ37Cl) 
the observed correlations between isotope fractionation are characteristic of the 
reaction mechanism. According to Abe et al. (2009), such a correlation is 
independent of contaminant concentrations.  
Available literature values for δ13C values for PCE and TCE from different 
manufacturers are provided in Table 2-3. 
36 
Table 2-3: Published δ13C values for TCE and PCE from various manufacturers  
Compound Results from Shouakar-Stash et al. (2003) Results from van Wamerdam et al. (1995) Results from Beneteau et al. (1999)
n Mean δ13CVPDB ‰ STDEV 1σ n Mean δ
13CVPDB ‰ STDEV 1σ n Mean δ
13CVPDB ‰ STDEV 1σ
TCE DOW 92 7 -31.57 0.01 2 -31.9 0.05 - - -
TCE DOW 95 4 -29.33 0.1 - - - 3 -29.84 0.07
TCE PPG 93 4 -27.37 0.09 2 -27.8 0.01 - - -
TCE PPG 95 4 -31.12 0.06 - - - 3 -31.68 0.01
TCE ICI 93 4 -31.01 0.09 3 -31.32 0.03 - - -
TCE StanChem 
93 3 -29.19 0.14 - - - - - -
PCE DOW 3 -23.19 0.1
PCE ICI 4 -37.2 0.03
PCE PPG 2 -33.84 0.03
PCE Vulcan 3 -24.1 0.04
37 
2.2 DNAPL Site Characterisation 
2.2.1 Site characterisation and the significance of the conceptual site 
model (CSM) 
Site characterisation is an important process of gaining insight into the nature, extent, 
complexities and risks associated with the contamination at a particular site. The site 
characterisation process is undertaken as a scientific, iterative, phased approach 
(Kiersch, 1958; Cohen and Mercer, 1993; Soesilo and Wilson, 1997; Sara, 2003; 
Gebrekristos, 2007; Hulley, 2009). Site characterisations often begin on a regional 
scale or far-field and then progressively become localised or near-field (Figure 2-18). 
The early phases of site characterisation utilises intrusive, non-intrusive and 
laboratory investigations to develop a conceptual site model (CSM) which provide a 
holistic overview of the site complexities, the source areas, the pathways of 
contamination and the receptors (Sara, 2003; United States Environmental Protection 
Agency, 2003). The CSM is revised as the investigation progresses and becomes an 
important and dynamic planning tool. 
Generic DNAPL conceptual site models have been developed for various geological 
settings (Huling and Weaver, 1991; Kueper et al., 2003; Pretorius, 2007). DNAPL 
CSM are utilised to assess: 
• Site characterisation properties and site-specific data; 
• The potential for separate phase DNAPL migration; 
• The potential for vapour transport of DNAPL chemicals; 
• The potential for dissolution and transport of DNAPL chemicals;  
• The distribution of chemicals in the subsurface; 
• Cross-contamination risks associated with characterisation and remedial 
activities; 
• Planning of detailed site characterisation activities; 
• Planning  of remediation activities; 
• Consideration of alternative remediation activities. 
Non-intrusive investigations such as reconnaissance site visits and surface 
geophysics add important information to the investigation when combined with the 
intrusive investigations (MacDougall et al., 2002). Non-intrusive investigations are 
often rapid and cost-effective and provide a guide for where drilling and direct 
sampling (intrusive investigation) should occur (MacDougall et al., 2002).  
38 
Figure 2-18: Transition from regional to site-specific data during the progression of the site
characterisation (from Sara, 2003)
The last decade has seen a paradigm shift from the traditional site characterisation 
approach to a more dynamic approach or the Triad approach (Interstate Technology 
and Regulatory Council, 2003). According to Interstate Technology and Regulatory 
Council (2003) the central principle of the Triad approach is the management of 
decision uncertainty. Decision uncertainty during a site investigation is managed 
through the development of an accurate conceptual site model.  Data uncertainty 
during the process is minimized though the use of various tools to address data gaps 
during the relevant phases of the investigation. The sources of data uncertainty 
(which subsequently may lead to decision uncertainty) as well as the tools that can 
be applied to minimise error or uncertainty are further elaborated on by the Interstate 
Technology and Regulatory Council (2003). 
The Triad approach was developed as an initiative by the United States Department 
of Energy in an effort to make site investigations and site remediation more cost 
effective (Burton et al., 1993). The Triad approach is three pronged and incorporates: 
• Systematic project planning: A coordinated effort is made to identify and 
manage factors and issues that may contribute to uncertainty and decision errors. 
Cost effective strategies to manage the factors and variables are developed. 
Systematic planning is also used to identify decision end-points and to estimate 
acceptable levels of uncertainty for the decisions that are required for the site. 
• Dynamic work plan strategies: These types of strategies use real-time 
decision-making in the field to limit the number of mobilisations back to the field to fill 
data gaps or take remediation actions. A dynamic work plan strategy is where 
39 
decisions are made and the work plans guide sampling and analysis are adjusted in 
response to data generated while the field crew is still on site.  
• Real-time measurement technologies: Real-time measurements are those 
that are produced within a rapid time-frame so that real-time decision-making and 
maturation of the CSM can occur in real-time. Rapid-time measurement technologies 
include in situ detection techniques, on-site analytical tools, mapping data in real-
time, rapid sampling platforms and rapid turn-around from a laboratory. 
The Triad approach has both advantages and disadvantages (Table 2-4) and the 
best results are obtained by combining traditional approaches with the Triad 
approaches in order to manage uncertainty, realise long term cost savings and 
effectively characterise and remediate a site. A combination of these methodologies 
and techniques are applied to the Investigation Site. 
 
Table 2-4: Advantages and disadvantages of the Triad approach  (Interstate Technology and 
Regulatory Council, 2003) 
Advantages Disadvantages
Better investigation quality Higher up-front costs
Faster investigations, restoration and Change in approach to data quality
redevelopment 
Lower life-cycle costs Lack of tools to manage decision uncertainty
Improved stakeholder communication Greater need for training about Triad
More effective clean-ups Negative bias towards field-generated data
40 
Figure 2-19: Life cycle of the CSM (from United States Environmental Protection Agency, 2003)
41 
2.2.2 DNAPL source zone characterisation in complex geo-contaminant
settings
2.2.2.1 General 
A successful and cost-effective remediation strategy is dependent on how well the 
source areas are characterised. This is particularly pertinent for DNAPL-
contaminated sites. DNAPL physicochemical properties (refer to Section 2.1.1.) make 
them difficult to characterise. The site characterisation process is often compounded 
through the presence of mixed DNAPLs and complex geological and hydrogeological 
settings.  
The DNAPL source zones together with geological heterogeneities play an important 
role in dissolution kinetics and plume geometry. DNAPLs source zones can sustain 
contaminant plumes over long distances and long periods of time. A hypothetical 
DNAPL source zone is depicted in Figure 2-20. In reality source zones have more 
heterogeneity than that shown below with DNAPL occurring in lenses and ganglia as 
compared to pooling of DNAPL. 
Subsurface conditions are generally heterogeneous, often complex and arise from a 
range of different geophysical and geochemical processes. The heterogeneities 
produced by varying degrees of fracturing, compaction and cavity formation within the 
different geological settings provides a challenge for making broad or generalised 
statements regarding the source zone characteristics and the efficacy of remediation 
technologies (National Research Council, 2004). 
Figure 2-20: Hypothetical DNAPL source zone (from Kueper and Davies, 2009)
42 
Five general geo-contaminant settings are shown in Figure 2-21, which broadly 
illustrate differing conditions. Spatial variations in permeability and porosity (Figure 2-
22) differentiate the five settings (National Research Council, 2004). The setting 
encountered at the Investigation Site includes a combination of Type IV and Type V 
settings (i.e fractured media with varying matrix porosity). This is typical of one of the 
major aquifer systems in South Africa where the dominant groundwater flow occurs in 
fractured and dual porosity systems such as the Karoo Supergroup (Woodford and 
Chevallier, 2002) and Table Mountain Group. The Karoo Supergroup is the dominant 
geology type at the Investigation Site (Chapter 3). According to Woodford and 
Chevallier (2002) the fractures within the Karoo Supergroup are commonly sub-
horizontal (<500) in attitude. The Karoo Supergroup is intercepted by numerous 
dolerite dykes, sills and plugs that cut through the sedimentary rocks. The dolerite 
rocks are generally fine or medium grained. However, they also coincide with faults 
and are fractured (Botha et al., 1998), making them unreliable as DNAPL barriers. 
According to Woodford and Chevallier (2002), the fractures occasionally extend 
several tens of meters from the dolerite intrusions into the Karoo Supergroup country 
rock.  
 As the Type IV and V geo-contaminant settings are encountered at the Investigation 
Site, the following paragraphs elaborate on the variations and characteristics of these 
settings.   
Figure 2-21: Geo-contaminant settings (from Air Force Center for Engineering
and the Environment, 2007)
43 
In Type IV settings the primary transmissive feature is secondary permeability 
resulting from fractures. Little or no void space exists in the matrix that typically has a 
very low permeability of less than 10-17m2 (K < 10-10 m/s. The bulk permeability of the 
media is however dependant on the fracture network (including the frequency, 
aperture size and degree of interconnectivity). The bulk permeability of Type IV 
settings media is considered to range from 10-15 – 10-11 m2 (K = 10-8 – 10-4 m/s). In 
Type IV settings, advection is limited to fractures. Due to the low matrix porosity, little 
mass is stored in low permeability zone. In this scenario, the primary source is 
DNAPL. Over time, the DNAPL is depleted from the more transmissive fractures and 
are dominated by DNAPL in dead-end fractures. According to Parker et al., 1997, a 
common feature of the Type IV setting is large plume dimensions due to high 
contaminant migration velocity. The primary challenge in a Type IV setting is the 
complexity of the fracture network.  
In a Type V setting the bulk permeability of the media ranges from 10-16 – 10-13 m2 (K 
= 10-9 – 10-6 m/s), while the permeability in the rock matrix is less than 10-17 m2 (K < 
10-10 m/s).  Typically the porosity of the fractures is <1% compared to the total unit 
volume of the rock mass. In a Type V setting the porosity of the rock matrix ranges 
from 1-40%. The low permeability zones in Type V settings initially attenuate the 
DNAPL constituents that partition into groundwater through diffusion from the 
fractures into the rock matrix. When the DNAPL is depleted (late stage in plume 
evolution), reverse diffusion from the rock matrix sustains dissolved phase 
concentrations in groundwater flowing in fractures. In scenarios where the matrix has 
a large sorptive capacity, the low permeability zones will act as a contaminant sink 
and accelerate the rate of DNAPL depletion.  
The National Research Council (2004) lists the following as critical information that 
are crucial for source zone characterisation: 
• Understanding the source presence and nature; 
• Characterising hydrogeology; 
• Determining source zone geometry, distribution, migration and dissolution 
rate in the subsurface; and 
• Understanding the biogeochemistry. 
Site specific heterogeneities do not allow the source characterisation methodology to 
be duplicated in their entirety from one site to the other. However, depending on the 
44 
site conditions a number of characterisation methods can be used to provide 
information about the source zone. 
Figure 2-22: Variations in permeability and hydraulic conductivity for different
geological media (from National Research Council, 2004)
Table 2-5 shows some characterisation methods that can be used to address the 
crucial points listed above. Also shown is the potential of the methods to provide 
information about the source zone. The complexities of undertaking DNAPL site 
characterisations are largely overcome by using the “Toolbox Approach” prescribed 
by previous authors (Kueper et al., 2003; Fiacco et al., 2005; Gebrekristos, 2007; 
Pretorius, 2007; Gebrekristos et al., 2008; Hulley et al., 2008) who utilize a variety of 
non-intrusive and intrusive approaches to adequately characterise a DNAPL-
contaminated site. When combined with the Triad approach the Toolbox approach 
can allow for successful DNAPL site characterisation and remediation. Kram et al.
(2001), United States Environmental Protection Agency (2003), Griffin and Watson, 
(2002), Kueper and Davies (2009) amongst others provide information on the various 
available and emerging source delineation and site characterisation tools.  
45 
Table 2-5: Characterisation methods and their potential for providing information about the 
source zone (from National Research Council, 2004) 
Method/Tool Source Material Hydrogeology Source zone Biogeochemistry
delineation
Historical Data Maybe Maybe Maybe No
Regional No Yes No Maybe
geology
Geophysical No Yes No No
tools
Direct push Maybe Yes Yes Yes
Core analysis Maybe Yes Maybe Yes
Downhole Maybe Yes No No
methods
Piezometers No Yes No No
Pump tests No Yes No No
Groundwater Maybe No Maybe Yes
analysis
Solid (matrix) No No No Yes
characterisation
Microbial No No No Yes
analysis
Soil vapour Maybe No Maybe No
analysis
DNAPL analysis Yes No No No
Partitioning No Maybe Yes No
tracer tests 
Ribbon NAPL Yes No Yes No
samplers
Dyes Maybe No Maybe No
* Methods highlighted above were applied to characterise the source material, hydrogeology, source zone and biogeochemistry at the 
Investigation Site (Chapter 3) 
2.2.2.2 Characterisation and delineation of DNAPL source zones  
The objectives for DNAPL source zone delineation varies from site to site and may 
include one of more the following: 
• To ensure that the flow paths and quality of groundwater downgradient of the 
source zone are monitored for the presence of dissolved phase contaminants to 
assess the protection of current and potential receptors. 
46 
• To facilitate the proper design of containment systems involving groundwater 
extraction and/or physical barriers. 
• To facilitate implementation of DNAPL mass removal technologies. 
• To establish boundaries for institutional controls (Kueper and Davies, 2009). 
According to Kueper and Davies (2009) it is not feasible to determine the exact 
location and extent of individual DNAPL migration pathways within the source zone. 
Uncertainty in delineating the spatial extent of the source zone is derived through 
taking a finite number of local scale measurements at discrete locations. They 
recommend the delineation of “Confirmed/Probable” as well as “Potential” DNAPL 
source zone. Overestimation of the size of Confirmed/Probable source zone could 
overstate the costs for technology application and may result in a particular 
technology being prematurely screened out. In contrast to this, underestimation of the 
size of the Confirmed/Probable source zone could lead to underestimation of costs 
and the poor performance or the chosen technology. 
As with conventional site characterisation processes, the characterisation of DNAPL 
source zones is an iterative process requiring multiple lines of evidence.  Figure 2-23
shows the line of evidence that can be used to assess the presence of DNAPL at a 
site. These include direct and indirect methods. Table 2-6 provides the degree of 
uncertainty of the presence of source zone(s) based on various events. At sites 
where a DNAPL cannot be isolated from the source zones, indirect methods are used 
to infer the presence of DNAPL. Such methods include the measuring of high 
dissolved or vapour contaminant concentrations relative to saturated aqueous or 
vapour concentrations; or the measurement of high concentrations of the 
contaminant in soil cores. According to Mackay et al. (1991) and Cohen and Mercer 
(1993); aqueous concentrations greater than 1% of DNAPL solubility is indicative of 
the presence of DNAPL. In a field experiment undertaken by Broholm et al. (1999), 
the dissolved-phase plume measured from the DNAPL source was greater than 10% 
of the effective solubility of the DNAPL components across the lateral and vertical 
extent of the source zone.  
The United States Environmental Protection Agency (1992) state that   
concentrations of DNAPL-related chemicals in soils greater than 10 000 mg/kg are 
indicative of DNAPL presence. The National Research Council (2004) proposes a 
cautionary approach when inferring the presence of DNAPL as a result of the 
heterogeneous distribution of DNAPL in the subsurface. They state that short of 
collecting actual chemical samples from a DNAPL sample, the nature of the source 
47 
material in terms of key physical and chemical parameters may not be fully 
understood. According to Kueper and Davies (2009) the two primary lines of 
evidence that a DNAPL source zone exists includes visual observation of a DNAPL 
sample retrieved from a monitoring well and/or if you have chemical concentrations in 
soil exceeding the value corresponding to a threshold DNAPL saturation. Other 
converging secondary lines of evidence can be used to determine the presence of 
DNAPL in the subsurface (Figure 2-23). These include: 
• Chemical concentrations in soil exceeding the value corresponding to equilibrium 
partitioning relationships. These threshold chemical concentrations can be 
calculated using the following equation: 
 =   ( +  + )    Equation 2-14 
Where: 
C Ti  = soil concentration (mg/kg) threshold for component i (calculated). C Ti  represents 
the maximum amount of contaminant i that can be present in a porous media sample 
in the sorbed, aqueous and vapour phases with a DNAPL phase present. 
Ci = effective solubility (mg/l) of component i (calculated) 
ρb = dry bulk density (kg/l) (site specific measurement) 
Kd = soil-water partition coefficient (l/kg) (calculated using Kd=Kocfoc, where Koc is the 
organic carbon-water partition coefficient (l/kg) and foc is the fraction organic carbon 
present in the soil) 
θw = air-filled porosity (no unit) (site specific measurement).  
KH = Henry’s Law constant 
θa = air-filled porosity (site specific measurement) 
The above calculation for the threshold chemical concentration can be applied below 
the water table by setting θa = 0. 
• Site Use/Site History can be ascertained by methods such as employee 
interviews, company purchase and sale records, aerial photographs and building 
plans. Former lagoons, underground tanks, floor drains and leach fields are 
sometimes coincident with the location of DNAPL source areas. 
• The location of vapour-phase plume may be coincident with the current or former 
presence of DNAPL in the unsaturated zone. Mapping of the vapour-phase plume 
is usually useful in deciding where to collect additional data. Active and passive 
techniques can be used. This line of evidence is not applicable to DNAPLs that 
lack significant vapour pressures such as coal tars and creosotes. 
• Hydrophobic dyes such as Oil Red O® partitions into the DNAPL imparting a red 
colour to the organic liquid. Hydrophobic dye techniques include a shake test in 
which soil or water is placed into a jar/bag with a small amount of dye, and down-
48 
hole samplers that force a dye impregnated absorbent ribbon against the 
borehole wall. Only pooled DNAPL can migrate towards the ribbon sampler and 
hence false negatives does not preclude the presence of residual DNAPL in the 
adjacent formation. 
• Evaluating groundwater quality data: 
o Sampled groundwater showing greater than 1% of effective solubility of 
DNAPL (see Equation 2-6 for the calculation of effective solubility) indicate 
that the sampled groundwater may have come into contact with DNAPL. 
The distance to the possible DNAPL locations cannot be determined from 
the magnitude of the concentration alone. Sampled groundwater 
concentrations downgradient of a DNAPL source zone can be significantly 
less than the effective solubility because of hydrodynamic dispersion, 
wellbore dilution, non-optimal monitoring well placement and degradation 
processes.  
o The presence of a persistent plume extending from suspected release 
locations in the downgradient direction is evidence of a continuing source. 
If significant back diffusion is occurring in the subsurface, the plume may 
persist even if the DNAPL has been depleted. This line of evidence is 
therefore most applicable to high permeability settings. 
o The presence of contaminated groundwater in locations that are not 
downgradient of known or suspected sources may be evidence of DNAPL 
presence hydraulically upgradient of the monitoring point. 
o Abrupt reversals of groundwater contaminant concentration levels with 
depth or increasing concentrations with depth can be associated with 
DNAPL presence. Concentration trends can be best detected using small 
interval sampling techniques. 
o Groundwater downgradient of a multi-component DNAPL may exhibit a 
temporal decline in the concentrations of the higher effective solubility 
compounds and a stable or increasing trend in time of the lower effective 
solubility compounds. Compound specific biodegradation may result in 
certain compounds decreasing and others (such as low molecular weight 
daughter products) increasing within the plume.  
o Detection of highly sorbing and low solubility compounds which have low 
mobility in groundwater may be associated with a nearby DNAPL source. 
This line of evidence can be useful in delineating the extent of the DNAPL 
in the downgradient direction. 
49 
• Other methods that are considered a line of evidence for DNAPL includes the use 
of partitioning interwell tracer tests (PITTs), which involves the injection and 
withdrawal of a tracer that has the ability to partition into the DNAPL. PITTs are 
typically employed after some level of source zone characterisation has been 
completed. The use of measurement probes such as the membrane interface 
probe or the hydrosparge probe with direct push techniques is becoming popular. 
Most of these devices provide a relative measure of the total concentration. 
Table 2-6: The degree of uncertainty regarding the presence of a source zone at a site based 
on the occurrence of various events (National Research Council, 2004) 
Event DNAPL Source
Known or probable historical release of DNAPL High certainty
Process or waste practice suggests probable DNAPL release High certainty
DNAPL visually detected in subsurface, monitoring wells, etc. High certainty
Chemical analyses indicate DNAPL presence (≥ saturation) High certainty
DNAPL chemicals used in appreciable quantities at site Likely; some uncertainty
Chemical analysis suggests possible source zone Likely, some uncertainty
According to Sale et al. (2008) the reasons why source delineation investigations 
miss a portion (s) of the source include: 
• The use of limited datasets with insufficient resolution 
• The heterogeneous distribution of DNAPL and the other phases 
• Not being able to gain information regarding the distribution of the DNAPL, the 
sorbed phase, vapour phase and the mass stored in low permeability zones. This 
arises from the heavy reliance on groundwater data obtained from boreholes with 
long screen intervals in transmissive zones 
• The failure to investigate low permeability zones at older sites  
• Hastily implementing a source zone remediation system without adequate site 
characterisation 
• Difficulties in resolving where the source ends and the plume begin. 
50 
NO a. Visual Observation YES
Inconclusive.
Use other lines of 
evidence (d-h) NO b. DNAPL saturation YES
c. Soil partitioning
d. Site Use 
NO DNAPL SOURCE Site History NO DNAPL SOURCEYES
ZONE ZONE
e. Vapours
f. Dye testing
h. Groundwater
g. Other (e.g. PITT)
Figure 2-23: Lines of evidence to assess DNAPL presence (modified from Kueper and Davies, 2009) 
51 
Data Collection and 
Interpretation
Has enough data been 
NO collected to assess DNAPL 
presence
YES
No further data collection is required in the 
Does a DNAPL source zone NO context of DNAPL source assessment and 
exist? delineation
YES
Is there a need to delineate NO
the DNAPL source zone?
YES
Map the confirmed/probable 
and potential DNAPL 
source zones
Collect more data Is more accurate 
YES delineation of source NO
zone(s) required?
Figure 2-24: Flowchart depicting iterative data collection process used in refining the DNAPL
source zone boundaries (modified from Kueper and Davies, 2009)
2.3 Defining the Level of Characterisation Required and Cost 
Implications 
The goal of a site characterisation is to define the nature and extent of the 
contamination. While in most cases this is achieved concurrently, a distinction can be 
made between these two goals (SERDP and ESTCP, 2006). 
The level of site characterisation required at a site is dependent on the following 
criteria: 
• The complexity of the site conditions (for e.g. the types of contaminants and their 
characteristics and distribution, the hydrogeological setting, the biogeochemical 
setting of the subsurface); 
52 
• The clean-up targets or goals (this is dependent on factors such as the regulatory 
drivers, the pathways and receptors, community pressure). 
The Strategic Environmental Research and Development Program and 
Environmental Security Technology Certification Program (SERDP and ESTCP, 
2006) recommend  the use of high resolution sampling of cores in order to determine 
the location and distribution of DNAPL contaminants in great detail as the mass 
distribution is typically controlled by small-scale features. High resolution core 
sampling is hence essential to determine the nature or style of contaminant mass 
distribution within source zones. In some cases minor heterogeneities can result in 
extremely complex migration pathways and localized entrapment of DNAPL.  
The sampling scale in this case is a function of the geological setting and the age and 
type of the contaminant. However, high resolution sampling and analysis is costly if 
the sampling scale is large. Site screening tools should therefore be an integral part 
of the characterisation method and should aim to reduce the uncertainty associated 
with source zones and facilitate the use of more rigorous delineation methodologies. 
Kram et al. (2002) use three unit model scenarios with predetermined parameters 
and compare these to selected site characterization techniques and approaches.  
Costs for three (3) different model scenarios were determined based on pre-defined 
assumptions and the application of different characterisation techniques. Each unit 
model scenario consisted of approximately 22 250 m3, with identical depths to 
groundwater, depth of resolution and volume of DNAPL released. Each unit model 
scenario consisted of different soil types. According to Kram et al. (2002) the most 
cost-effective technique is the FLUTeTM approach with the most expensive technique 
being the PITT. 
Miansney and McBratney (2002) defined efficiency of different methods used in 
predicting water retention and hydraulic conductivity of soil samples in terms of effort, 
cost or value of information. The study found that due to large spatial variations of soil 
hydraulic properties, the use of cheaper qualitative and semi-quantitative 
methodologies are more efficient than the use of costlier methods with higher levels 
of precision.  The value of the use of less expensive, qualitative methodologies in 
contaminated land management is further elaborated by a guideline document 
produced by Barnes (2009). The use of these tools (termed “rapid measurement 
tools” in the document) is presented within the framework of risk management for 
53 
contaminated sites in conjunction with the use of laboratory-based analysis 
(quantitative methodologies). 
54 
CHAPTER 3 
3 DNAPL SOURCE ZONE CHARACTERISATION IN A 
FRACTURED SYSTEM 
The mid-twentieth century saw an increasing number of chemical industries being 
developed in South Africa; largely as a result of sanctions against the country and 
increasing national demand. The Investigation Site was one of the earliest chemical 
facilities in the country. As the site is confidential, information related to site 
ownership, locality etc. are excluded from this thesis. 
The site is a large industrial complex, consisting of several production facilities and 
covering an area of approximately 400 ha. It is located within a heavy industrial zone 
and is proximal to residential areas, agricultural lands, open municipal land and 
underground mining and other chemical production facilities (Figure 3-1).  
This chapter provides background into the Investigation Site and evaluates traditional 
and novel technologies for adequate source zone characterisation in a fractured 
system.   
3.1 Background to the Investigation Site 
The production of chemicals is a highly integrated process involving a combination of 
steps: 
• Receiving of raw products via road, rail and/or pipeline  
• Storage of the raw products in warehouses and/or tank farms (underground 
and/or above ground tanks) 
• Transport of the raw materials as input into the process  
• Blending of raw materials 
• Transport of product via pipes, road for storage on-site 
• Temporary storage of waste products on-site 
• Treatment and disposal of effluent and solid waste (hazardous, non-
hazardous/inert) 
• Transport of product via pipes, road or rail off-site. 
During operations, chemicals can enter the subsurface as a result of poor 
housekeeping (such as uncontained spills during loading/off-loading), through 
55 
seepage from disposed wastes in lagoons and waste sites and via leaking storage 
tanks. 
Manufacturing operations at the Investigation Site began in the mid-1960’s and 
included the bulk chemical production of a wide range of chemical products 
consisting of polyvinyl chloride, peroxides, carbon fluorocarbon, 1,2-dichloroethane, 
polyethylene, tetrachloroethene, carbon tetrachloride, cyanide, carbon fluorocarbons 
and chlorine.  Production facilities built in the mid-1960s that handled and 
manufactured DNAPLs were decommissioned and demolished in the late-1990s 
(Table 3-1). The on-site disposal of chemicals also ceased with increased awareness 
of the risks associated with the inappropriate disposal of chemicals and with stricter 
environmental controls. A new production facility where vinyl chloride and 1,2-DCA 
are stored and handled as part of the manufacturing process was built in the late-
1990s to updated international standards. Figure 3-2 shows the facilities that are or 
were associated with the production, storage and disposal of DNAPLs at the 
Investigation Site. These facilities can be considered potential source areas, based 
on the processes that were or are occurring at the site. 
56 
Figure 3-1: Land use surrounding the Investigation Site
57 
Figure 3-2: Potential source areas at the Investigation Site
58 
Table 3-1: Facilities at the Investigation Site that are/were associated with production, handling and/or disposal of DNAPLs 
Map Legend (Figure 3-2) Plant Production/function Area (ha) Duration of Operations Probable DNAPL of Concern
Historical DNAPL production and Activated carbon with mercuric chloride used to 3.5 1966-1996 Mercury; 1,2-DCA; coal tar 
handling facility 1 produce acetaldehyde  and vinyl chloride residue; vinyl chloride  
Historical DNAPL production and Carbon tetrachloride, PCE, 1,2-DCA, TCE, 7.5 1967- 1996 PCE; CCl4; chloroform; TCE;
handling facility 2 Arcton 11/12, Frezone 22 and HCl, chloride 1,2-DCA; 1,1,2,2-TetCA; 
vinyl chloride; 1,1,2-TCA, 
mercury 
Storage of 1,2-DCA 1967-1998 1,2-DCA 
Redundant hazardous waste site Redundant disposal site   8 1966-1978 Mercury; 1,2-DCA; coal tar 
1 residue; PCE; CCl4; TCE; 
vinyl chloride 
Redundant hazardous waste site Redundant disposal site   5 1978-2004 Mercury; 1,2-DCA; coal tar 
2 residue; PCE; CCl4; TCE; 
vinyl chloride 
Effluent  Dams Storage of storm water and effluent on southern 4 1966-current Mercury; 1,2-DCA; coal tar 
section of site residue; PCE; TCE; CCl4
Operational plant using DNAPL 1,2-DCA manufactured and used to produce 13 1996-current 1,2-DCA; vinyl chloride
other chemicals 
59 
3.2 Methodologies Used to Investigate the Site 
The Investigation Site has been under characterisation by different Consulting 
Companies since the early 1990’s. Groundwater monitoring reports showed the 
presence of dissolved phase and free phase DNAPL. Additionally, assessments at 
facilities that used mercury (Table 3-1) indicated the presence of elemental mercury 
(H0) in soil and building rubble within the footprint of the demolished facilities. The 
long history of various studies and monitoring at the site has allowed for a vast 
amount of historical information to be available. Staff members who began work at 
these facilities in the 1970’s and whom are still employed at the site were interviewed. 
The author gathered the following information from these employees during 
interviews: 
• Samples from the storage tanks containing 1,2-DCA were obtained by opening 
the valves for approximately 5 minutes and allowing the product to drain into the 
ground (the storage area has no secondary containment) prior to sampling. This 
occurred at least once every day during the history of these tanks being used for 
the storage of 1,2-DCA (1967-1998). 
• Disposal of off-specification material in the basements or pump houses of the 
production facilities was common practice. 
• The waste sites were unlined. 
The traditional approach in determining the extent of the DNAPL contamination on 
site relied heavily on gathering information through the drilling of boreholes below 
potential source areas. As a result approximately 40 borehole pairs up to a maximum 
depth of 40 metres below ground level were drilled in the 1990s. At the time very little 
was understood regarding the fate and transport of DNAPLs (Figure 2-2) and much 
less was understood on the significance of geology on the fate and transport of 
DNAPLs in bedrock.  
A site-specific conceptual site model (CSM) is developed as an outcome of this study 
through source zone characterisation. The methodologies used in the 
characterisation process are an integration of traditional and novel approaches.  For 
the purposes of this investigation, traditional site characterisation approaches are 
defined as following the following criteria: 
• Having a fixed work plan 
• Samples for analysis are sent to a fixed laboratory 
• Field methods that are employed are quantitative 
60 
• High level of uncertainty which can be reduced sequentially over an extended 
period of time through repeated mobilisations to site 
• Well placement is based on available knowledge on the potential release areas or 
source zones.    
In comparison, novel source zone characterisation approaches may have aspects of 
traditional approaches (such as confirmatory sampling using a fixed laboratory), but 
also include dynamic work plans and the use of rapid measurement tools and in situ
methodologies to reduce uncertainty within a shorter time frame. 
61 
Figure 3-3: Available historical near-field and far-field borehole information (modified from Hulley et al., 2008)
62 
Figure 3-4: Historical evidence of DNAPL contamination in the subsurface at the Investigation Site (modified from SRK Consulting, 1995; SRK Consulting, 
2001; ERM, 2008) 
63 
3.2.1 Collation and review of historical information and the identification 
of information gaps 
A significant amount of geological and geohydrological information was available 
from the Investigation Site, neighbouring industrial facilities and mines at the 
commencement of this research.  The data included information on the geology of the 
area.  Approximately 2000 coal exploration boreholes had been advanced in a 10 km 
radius from the site providing an extensive database with existing information. 
Additional information was collated from the following sources: 
• Previous reports prepared by Consultants for the mines and industries 
• 1:50 000 topographical sheets, 
• 1:250 000 hydrogeologcial maps obtained from the Department of Water Affairs 
and the National Groundwater Database data on groundwater wells in the region 
• 1:250000 regional geological maps published by the Council for Geoscience.  
 A far-to-near field approach was adopted to evaluate the available data. The reason 
for this was to first evaluate regional geological controls that could affect the 
migration of DNAPL. When working at DNAPL sites, drilling within sources without 
understanding the exact dimensions of the source areas can lead to cross-
contamination of aquifers or the mobilisation of the contamination. The boreholes 
with available geological information (existing historical datasets) are shown in Figure 
3-3.  
The following were identified as information gaps that required further investigation: 
• Obtain near-field (site-specific) geological and hydrogeological information in 
order to: 
o Determine the local geology underlying the Investigation Site; 
o Establish the depth of overburden and the weathering profile; 
o Establish the depth and configuration of intrusive dolerite sills into the 
Karoo Ecca Sequence; 
o Define the nature and depth of the pre-Karoo basement rock 
o Define the nature of the fracture network 
o Understand the types of aquifers at the Site and the interrelationship 
between the aquifers; 
o Determine the aquifer parameters 
• Characterisation of the source zone through the delineation of the sources or 
release areas and determining the source material characteristics 
64 
• Biogeochemical characterisation of the plume.   
Careful planning of all activities associated with the characterisation preceded any 
work being undertaken. This involved the development (by the author) of the project 
execution and communication plans for internal and external stakeholders. The 
Consultants developed the Safety, Health and Environmental plans which included 
emergency preparedness plans, personnel occupational health monitoring 
requirements, task work procedures and risk assessments and a DNAPL contingency 
plan should DNAPL be encountered during drilling.  
3.2.2 Source zone characterisation 
Source zones contaminated with elemental mercury (Figure 3-4) were investigated 
extensively by Consultants using test pits and DraegerTM tube analysis and will not be 
elaborated on further in this research. This section will focus on characterising source 
zones consisting of organic DNAPLs. 
DNAPL source zones were delineated through the application of the following 
approaches: 
• Rapid, cost-effective methods to confirm release zones and the presence of 
DNAPL  
• Rapid, cost-intensive methods to confirm the presence of free phase DNAPL in 
the weathered aquifer. 
In the context of this study the “weathered zone” includes the overburden and 
weathered portion of the dolerite or Karoo Ecca sediments located within the upper 
portion of the site, while the “fractured rock” refers to the unweathered portion of the 
stratigraphy. The subdivision is made on the basis of the different technologies used 
to characterise the source zones. The term “fractures” used in the context of this 
study includes planar structural features or discontinuities such as joints, cracks, 
fractures and bedding planes. Hence, the weathered zone in this context is also 
“fractured” based on the large amount of jointing identified at its base. The tasks and 
tools associated with source zones characterisation in the weathered and fractured 
zones are shown below (Figure 3-5).  
65 
3.2.2.1 Weathered source zone characterisation 
Confirmation of source zones in the unsaturated zone was done through a semi-
quantitative method. The GORE Module (also known as the GORESORBER™ 
Module) was used to collect chlorinated hydrocarbon vapour concentrations in the 
unsaturated zone. The GORESORBER™ modules consist of several engineered 
granular adsorbent materials that are encased in a chemically inert, hydrophobic, 
micro-porous GORE-TEX™ membrane. Modules were placed approximately 100 
meters apart in a grid that covered known, probable and potential source areas. The 
sampling event was undertaken during winter in order to prevent false negatives due 
to the effect of temporal variability effects on gas partitioning (Washington, 1996; 
McHugh et al., 2007). Subsurface borings were to an approximate depth of 500 to 
800 mm below ground level (bgl) advanced using a handheld hammer drill or a 35 
mm diameter hand auger, depending on subsurface geological conditions. Each hole 
was sealed with inert materials to minimize vapour exchange with the atmosphere, 
and allowed to equilibrate with subsurface vapours for a period of approximately 
twelve to fourteen days.  The unique serial number, sample locality and date were 
recorded on site. Once removed, the modules were couriered to W.L. Gore 
Associates, Inc. (Gore) in the USA for laboratory analysis of chlorinated 
hydrocarbons using US EPA Method 8260 and the gas chromatograph/mass 
spectroscopy (GC/MS) technique. The target compounds analyzed during the 
passive soil gas survey are summarized in Table 3-2.  
The only successful geophysical technique that was employed within the boundaries 
of the Investigation Site was the ABEM LUND (2-D) Resistivity Imaging System1. The 
purpose of the geophysical surveys was to: 
• Map the bedrock sub outcrop surface including the depth of the overburden and 
the weathering. The bottom of weathering could be stratigraphical lows where 
DNAPL could collect. 
• Define the geological units including the depth and configuration of the units. 
• Identify preferential groundwater migration pathways. 
1Controlled Source Array Magneto Telluric (CSAMT, Stratagem EH4 system by Geometrics 
Inc.) surveys were undertaken outside the site boundary to confirm the depth to basement. 
CMSAT surveys were restricted to outside the site due to interferences from the facilities on-
site.  Drums containing chlorinated solvents were suspected of being buried in a redundant 
production area, based on information provided during interviews with site Employees. 
Ground penetrating radar (GPR) was used in order to try and isolate this source. No useful 
results were obtained.
66 
• Identify potential contaminant plumes. 
Electrode spacing within source zones was reduced from 10 to 5 meters. Resolution 
depths varied between 20 m bgl to 70 m bgl. The location of the 2-D resistivity 
surveys are shown in Figure 3-7. Typically, these traverses should be undertaken 
with a grid configuration with traverses along parallel lines in order to maximize data 
coverage. However, due to the large amount of surface and sub-surface 
infrastructure at the site this was not possible.  
Drilling in the weathered zone was undertaken using an ATV 600 C roto-sonic drilling 
equipment for soil core retrieval. This is the first application of a roto-sonic drill rig for 
an environmental investigation in South Africa. Sonic drilling is a soil penetration 
technique that applies the principles of Binghams’ findings on the fluidization of 
porous materials (Bingham, 1916) in combination with the law of inertia. This 
technique was applied to the site as it allows for the fast retrieval of core without the 
use of any fluids2, hence enabling the tracking of seepage zones. The soil cores 
retrieved from each boring were screened for the presence of chlorinated 
hydrocarbons using the MiniRAE 3000 PID on 0.25 m intervals. Where high PID 
readings (> 150 ppm) indicated the possible presence of DNAPL, field shake tests 
using the hydrophobic dye Oil Red O® was employed to verify the presence of 
DNAPL (Figure 3-6).    
An on-site field laboratory was set up and managed by the Consultants. The purpose 
of this laboratory was to ensure that samples were analysed in real-time in order to 
facilitate rapid decision-making regarding the delineation of release areas. The 
laboratory consisted of two gas chromatographs that were dedicated to analysing 
samples generated from the sonic drilling. Gas chromatography is widely accepted 
as a primary analytical tool for site characterisation due to its capability to separate, 
detect and quantify target analytes in a complex mixture of thermally stable organic 
compounds. The gas chromatographs were operated by trained and experienced 
laboratory technicians.  
2 Attempts with direct-push equipment (Geo-probe) failed at the site as a result of the high 
subsurface clay content.  
67 
Identify potential historical review
release areas
Undertake passive GoresorberTM
soil vapour surveys  modules
Advance boreholes in Roto-sonic 
delineated release drilling
Undertake active soil Photo-
vapour surveys of ionisation
cores detector
Confirm presence of Oil Red O
DNAPL using dye for 
samples with vapour 
measurements >150 ppm
Analyse soil samples On-site field 
laboratory -
chemical 
analysis
off-site 
laboratory -
geotechnical 
parameters
Identify potential Historical 
groundwater review
anomolies in fractured Resistivity 
Advance boreholes  Percussion 
drilling
Delineate source zones 1. Visual
inspection
2. Ribbon NAPL
samplers
3. hydrophobic 
dyes
4. soil gas survey 
Determine properties 1. Ribbon NAPL 
of source matrix  samplers
2. downhole
geophysics
3. packer tests
Analyse source matrix, off-site laboratory 
source material and analysis of 
aqueous samples geotechnical, 
chemical and 
physical 
properties
Figure 3-5: Tasks and tools/methods for source zone characterisation at the Investigation
Site
68 
FFrraaccttuurreedd ((bbeeddrroocckk)) ZZoonnee WWeeaatthheerreedd ZZoonnee
Table 3-2: Target compounds for the passive soil gas survey
Compound Method Detection Limit (µg)
Total petroleum  hydrocarbons
Total petroleum hydrocarbons 0.01
Chlorinated Ethenes
Tetrachloroethene 0.01
Trichloroethene 0.01
1,1-dichloroethene 0.03
cis-1,2-dichloroethene 0.01
trans-1,2-dichloroethene 0.03
vinyl chloride 0.23
Chlorinated ethanes
1,1-dichloroethane 0.01
1,1,1-trichloroethane 0.01
1,2-dichloroethane 0.01
1,1,2-trichloroethane 0.01
1,1,1,2-tetrachloroethane 0.01
1,1,2,2-tetrachloroethane 0.01
Chlorinated methanes
Chloroform 0.01
carbon tetrachloride 0.01
Chlorinated benzenes
chlorobenzene 0.01
1,4-dichlorobenzene 0.01
1,3-dichlorobenzene 0.01
1,2-dichlorobenzene 0.01
Figure 3-6: Oil Red O® staining of residual DNAPL in soil boring 177
69 
Figure 3-7: 2-D Lund surface geophysical traverses (Results from ERM, 2008; ERM, 2012) 
70 
Soil samples taken at 0.25 to 0.5 m intervals were submitted under chain of custody 
QA/QC procedures to the field laboratory. The methanol was extracted using a 
combination of shake and centrifugal techniques. Samples were analysed by EPA 
SW846 Method 8620 (gas chromatograph/mass spectrometer, GC/MS) for the same 
VOCs analysed as part of the passive soil vapour survey (Table 3-2). The laboratory 
quality control system adhered to the requirements specified in EPA SW846 Method 
8620 and included daily QC checks, GC/MS run logs, laboratory control samples, 
matrix spike/matrix spike duplicate samples, continuing calibration verification and 
blank samples. Sample results were submitted electronically as electronic tables 
(Excel spreadsheets) and electronic data deliverables. 
The results obtained from the analysis of the samples from the weathered zone were 
treated statistically using Monte Carlo simulations to determine their threshold 
concentrations (based on Equation 2-16). The constants used in the simulations are 
provided in Table 3-3. The randomizing factors for dry soil bulk density, water-filled 
porosity, air-filled porosity and fraction of organic content were based on the site 
specific results obtained. 
Solid models for the chlorinated hydrocarbons were developed using RockWorks15. 
Frequency distribution histograms were generated for each compound as well as for 
the sum of groups of compounds. Based on this, data was filtered to exclude the 
maximum concentration outliers. The data was modelled using the built-in software 
inverse distance anisotropic algorithm. High fidelity smoothing was applied to the 
solid models. 
Soil samples taken from sonic-drilled boreholes (EDC 2S, BH 221 and BH 61) were 
sent to Soil Lab (Pty) Ltd in Pretoria (South Africa) for the following analyses: 
• particle size (grain size distribution) 
• permeability tests 
• organic content 
• porosity tests 
• dry density, and  
• moisture content. 
Grain size distribution tests were taken at intervals of 0.5 m for BH 221 and BH 61, 
and at visible lithological changes for EDC 1S. The lithological parameters analysed 
for the soil samples is provided in Table 3-4. 
.  
71 
Table 3-3: Constant values used for the Monte Carlo simulations to determine soil concentration threshold values 
Chlorinated Solubility (mg/l) Reference Henry’s Law Reference Organic carbon- Reference
hydrocarbon Constant (unit less) water partition
DNAPL coefficient (L/Kg)
PCE 206 Horvath et al., 1999 0.72363 Gossett, 1987 94.94 Meylan et al., 1992
TCE 1280 Horvath et al., 1999 0.402698 Leighton and Calo, 60.7 Meylan et al., 1992
1981 
1,1-DCE 2420 Horvath et al., 1999 1.067048 Gossett, 1987 31.82 Sabljic et al., 1995
Cis-1,2-DCE 6410 Horvath et al., 1999 0.166803 Gossett, 1987 39.6
Trans-1,2- 4520 Horvath et al., 1999 0.16683 Gossett, 1987 39.6
DCE 
VC 8800 Horvath et al., 1999 1.13655 Gossett, 1987 21.73
1,1,2,2-TetCa 2830 Horvath et al., 1999 0.015004 Leighton and Calo, 94.94 Meylan et al., 1992
1981 
1,1,1-TCA 1290 Horvath et al., 1999 0.703189 Gossett, 1987 43.89 Meylan et al., 1992
1,1,2-TCA 4590 Horvath et al., 1999 0.033688 Leighton and Calo, 60.7 Meylan et al., 1992
1981 
1,1-DCA 5040 Horvath et al., 1999 0.229763 Gossett, 1987 31.82
1,2-DCA 8600 Horvath et al., 1999 0.048242 Leighton and Calo, 39.6 Meylan et al., 1992
1981 
CCl4 793 Horvath, 1982 1.128373 Leighton and Calo, 43.89 Meylan et al., 1992
1981 
CHCl3 7950 Mackay et al., 1980 0.150041 Gossett, 1987 31.82 Meylan et al., 1992
72 
3.2.2.2 Fractured (bedrock) source zone characterisation: 
Potential source zones in the fractured bedrock were identified based on the 
presence of free phase DNAPL found at the site during previous investigations, 
through anomalies identified through geophysical surveys (see Section 3.2.) and by 
drilling directly within inferred release areas (for example, within the footprint of the 
historical plants) determined through the soil vapour investigations and the results 
from testing of soils in the overburden and weathered zones.   
Rotary percussion drilling was planned to commence in the release areas and to then 
work on an outward on a grid system dependent on whether DNAPL is intercepted on 
not. This method was not possible. Site clearance (as part of the site safety 
requirements) for drilling was undertaken on a per borehole basis and the drill rig was 
moved across the site depending on where permission was granted. The primary 
release area/source zone was targeted from drilling from the most likely point of 
direct release to the ground to outward to delineate the lateral extent. Drilling targeted 
the different fractured bedrock stratigraphic zones in order to delineate the vertical 
extent of the DNAPL.  
Rotary percussion boreholes were advanced within source zones to depths of 
between 6 m and 85 m bgl. VOC headspace screening of recovered drill chips were 
undertaken at 1 m intervals using a MiniRAE 3000TM PID equipped with an 11.7 eV 
lamp. Recovered cores were also tested using Oil Red O®. Drilling was permanently 
stopped when free phase DNAPL was intercepted or other lines of evidence (such as 
high PID readings, strong odours, positive Oil Red O®) indicated the probable 
presence of DNAPL in the borehole. The final depths of bedrock boreholes drilled are 
provided in Appendix A. A bottom loaded bailer was immediately lowered to the base 
of each bedrock borehole to determine the presence/absence of DNAPL on 
completion of the hole. Drilled boreholes were inspected daily using the bail-down 
test thereafter to check whether any DNAPL had mobilised into the borehole from an 
intercepted fracture.  
The presence of residual DNAPL in the fractured bedrock was identified using ribbon 
NAPL samplers (Flexible Liner Underground Technologies Ltd, FLUTeTM membrane 
system). This marks the first use of ribbon NAPL samplers in South Africa. These 
samplers allows for depth discrete mapping of residual DNAPL within the fractured 
73 
zone. The FLUTeTMs were also used to develop vertical transmissivity profiles by 
measuring the descent velocity with depth  
Bedrock samples were sent to the Centre for Applied Geosciences (ZAG) at 
Tübingen University in Germany for laboratory analyses of the following parameters: 
• Matrix porosity using a helium porosimeter test 
• Dry bulk density 
• Rock density 
• Matrix permeability using a permeameter test, and 
• Fraction of solid-phase organic carbon (fOC) using EPA method 410.4 / SM 5220D. 
The lithological parameters analysed for the bedrock samples is provided in 
Table 3-4.  
Free phase DNAPL samples retrieved from the source zones were submitted to 
laboratories in the USA (Alpha Laboratory and Stone Environmental Laboratory) to 
determine the chemical composition. Samples were also submitted to the SABS in 
Pretoria and Tufts University in the USA for analysis of density, viscosity, interfacial 
tension and water content. The following methodologies were employed by Tufts 
University in order to determine the physical properties: 
• Samples were centrifuged in 40 mL centrifuge tubes using a Beckman 
Coulter Centrifuge to separate the NAPL from any aqueous phase and/or fines that 
may be present in the samples 
• The samples were centrifuged at 1500 rpm for 10 min. Samples were allowed to 
equilibrate in a constant temperature room (22 ± 0.10 C) for a minimum of 24 
hours before performing analysis 
• Density measurements were done using 10 mL pycnometer (nominal size). 
• Water content measurements were done with a DL38 Karl Fisher Titrator. 
• Interfacial tension measurements were done using drop shape analysis (IT 
Concept Tracker) using the pendant drop technique. 
• Viscosity measurements were done over a range for shear rates (5‐100 s‐1), using 
an AR‐G2 Rheometer with a stainless steel concentric cylinder geometry. 
Source zones in bedrock were also inferred through the evaluation of the effective 
solubility of the chlorinated hydrocarbons detected in the groundwater sampled at 
these wells, adapted from the method in Kueper and Davies (2009). 
 
74 
Table 3-4: Lithological parameters measured at selected soil and bedrock cores 
Borehole Depth (m Lithology Analysed for
Number bgl)
BH 221 3.0-3.3 Weathered Organic content
sandstone
BH 221 4.3-4.5 Residual Moisture content, dry density, 
sandstone coefficient of permeability, organic 
content
BH 221 6.0-6.20 Residual Moisture content, dry density, 
sandstone coefficient of permeability, organic 
content
BH 221 9.0-9.2 Weathered Moisture content, dry density, 
mudstone coefficient of permeability, organic 
content
EDC 2S 0.55 Residual dolerite Moisture content, dry density, 
coefficient of permeability, organic 
content, porosity
BH 61 2.0-2.15 Residual dolerite Moisture content, dry density, 
coefficient of permeability, organic 
content, porosity
BH 61 2.8-3.0 Residual dolerite Moisture content, dry density, 
coefficient of permeability, organic 
content
BH 61 3.3-3.45 Residual dolerite Moisture content, dry density, 
coefficient of permeability, organic 
content
CAP 1 12-12.5 Weathered Matrix porosity using a helium 
dolerite porosimeter, Dry bulk density, 
Rock density, Matrix permeability 
using a permeameter, Fraction of 
solid-phase organic carbon ()
CAP 1 22-22.5 Unweathered Matrix porosity using a helium 
dolerite porosimeter, Dry bulk density, 
Rock density, Matrix permeability 
using a permeameter, Fraction of 
solid-phase organic carbon ()
SWD 1 20-20.5 Unweathered Matrix porosity using a helium 
dolerite porosimeter, Dry bulk density, 
Rock density, Matrix permeability
using a permeameter, Fraction of 
solid-phase organic carbon () 
SWD 1 35-35.5 Unweathered Matrix porosity using a helium 
sandstone porosimeter, Dry bulk density, 
Rock density, Matrix permeability 
using a permeameter, Fraction of 
solid-phase organic carbon ()
SWD 1 76-76.5 Unweathered Matrix porosity using a helium 
andesitic lava porosimeter, Dry bulk density, 
Rock density, Matrix permeability 
using a permeameter, Fraction of 
solid-phase organic carbon () 
75 
3.2.3 Near-field geological and hydrogeological characterisation  
Geological and hydrogeological fieldwork was undertaken with the Consultants 
contracted to do the work and under supervision of the author. Drilling was 
undertaken in a phased approach with the initial drilling of a limited number of core 
holes (6) and percussion boreholes (9) on-site in order to refine the preliminary 
conceptual model that was based on historical data. Once the source zones in the 
weathered aquifer was delineated, additional diamond drilled core holes (13) and 
percussion drilled holes (103) were advanced into the bedrock.  
The geology of the site was investigated by: 
• Using existing databases of construction logs 
• Using a core drill rig to advance boreholes to the basement. 
The factors that determined the locality of the core holes included: 
• 2-D resistivity anomalies (for e.g. low resistivity zones) which could suggest the 
presence of a transmissive fracture zones or groundwater with elevated specific 
conductivity (for e.g. contaminated groundwater) 
• Logistical considerations. 
Diamond core drill runs were at 3 m lengths. Hydrophobic dye (Oil Red O®) was 
sprinkled onto the recovered core to determine whether DNAPL was present. 
Additionally, fractures and joint sets within the core were screened for VOCs using a 
handheld MiniRAE 3000TM PID equipped with an 11.7 eV lamp. Where the PID 
readings exceeded 50 ppm and/or a positive dye test was recorded, the borehole 
was advanced by up to 3 m to create a sump to allow for DNAPL collection. Drilling 
was then stopped. 
All core holes were logged and the logs are provided in Appendix A. Core Logging 
included geotechnical, structural and lithostratigraphic parameters: 
• Core recovery – amount of core recovered (measured in the core box) as a 
percentage of length drilled as determined by the driller (core run length). A 
significant loss of core is usually the result of core destruction by the drill bit 
(grinding) in poor, weak or soft ground conditions. A minor core loss or gain is 
usually a function of slightly more or less core collected in the core barrel 
measured against the length drilled (core run length); 
• RQD (Rock Quality Designation) – the collective sum of pieces of core greater 
than 100mm length per core run length expressed as a percentage (%). A low 
76 
RQD reflects broken core usually due to the presence of jointing, fractures and 
other structural discontinuities; and 
• Joint or fracture frequency – a measure of the number of all joints and fractures 
including open and closed (sealed joints) per core run length expressed as 
number of joints per meter. 
• Joint or fracture discontinuity 
• Geology: 
o Stratigraphic code (provided in Appendix A) 
o Rock type – main description of rock type i.e. sandstone, shale, lava, 
dolerite 
o Grain size of main rock type: 
 very coarse  - >6mm i.e. conglomerate; 
 coarse  – 2 to 6mm i.e. grit; 
 medium– 0.6 to 2m i.e. sandstone, dolerite, lava; 
 fine - 0.2 to 0.6mm i.e. siltstone, lava; and 
 very fine - <0.2mm i.e. shale, clay. 
o General weathering of the rock type:
 extreme weathering; 
 very weathered; 
 moderately weathered ; 
 slightly weathered; and 
 unweathered. 
o Colour of main rock type; 
o Hardness of main rock type: 
 very hard– extremely difficult to break; 
 hard – difficult to break; 
 moderate   – can be broken with hammer; 
 soft – can be scratched with a knife; and 
 very soft   – easily crumbles or deforms. 
o Bedding dip – bedding / core angle of a sedimentary rock i.e. 0-5 is 0° to 5° 
from the horizontal – a flat or sub-horizontal bedding; 
o True width – a measure of the true thickness of the geological unit. 
Rock coring is significantly slower (approximately 1m/hour) than air percussion 
methods and is therefore more costly.  
Borehole geophysical methods, such as electrical resistivity or natural gamma logs, 
provide continuous stratigraphic column information relevant to a specific location or 
77 
station. The purpose of the borehole geophysical logging at the Investigation Site 
was to: 
• Collect fracture orientation data to enhance the regional geologic conceptual 
model; 
• Identify potential transmissive fractures/fracture zones; and 
• Collect proxy geology data for comparison with the detailed geologic logs to 
determine if borehole geophysical logging can be used as a more cost effective 
means of interpreting geologic conditions. 
The following borehole geophysical logging tools (the probe system is denoted in 
parenthesis) were used to log diamond core boreholes advanced at the site: 
• Acoustic Televiewer (ATV) – acoustic imaging of the borehole that is interpreted to 
identify the depth, orientation and aperture of bedrock fractures. This tool can only be 
used in the presence of water. 
• Fluid Temperature (FT4/GTC3) – measures groundwater temperature within a 
borehole; inflections in the groundwater temperature profile may be attributable to 
groundwater entering or leaving the borehole via transmissive bedrock fractures. 
• Fluid Conductivity (FT4) - measures groundwater specific conductivity within a 
borehole; inflections in the groundwater conductivity profile may be attributable to 
groundwater entering or leaving the borehole via transmissive bedrock fractures. 
• Caliper (GTC3) – provides a continuous measurement of the borehole diameter. 
Enlargements in borehole diameter may represent bedrock fractures, fractured 
zones, weathered zones, or zones of softer rock. 
• Natural gamma (GTC3) – measures natural gamma activity due to the presence of 
minerals containing radioactive elements such as potassium, which can be used to 
interpret geologic stratigraphy (e.g., in sedimentary sequences, shales contain 
potassium-bearing clay minerals such as illite or montmorillonite whereas sandstones 
contain relatively few clay minerals). 
• Self Potential (RR2) – measures self-potential anomalies that are typically attributed 
to the presence of graphite, high concentrations of base metal sulphides such as 
pyrite), or fluid flow in porous media. 
• Resistivity (RR2) – measures the electrical resistivity of the rock, which is dependent 
on the mineral content (e.g., metal sulphide or oxide minerals are highly conductive 
whereas other minerals are typically weakly conductive), water content, and/or water 
quality. Given that the majority of minerals are poor conductors, low resistivity 
anomalies typically represent zones of increased porosity, fracturing, or groundwater 
conductivity. 
78 
• Dual density (DD6) – measures the density of geologic formations, which can serve 
as a proxy for rock type, by lowering a gamma source into a borehole and measuring 
the gamma rays arriving at dual detectors placed at fixed distances from the gamma 
source. 
“Fracture pick” analysis was undertaken by Reeves Wireline Services on the 
diamond drill core holes using the acoustic televiewer data. In this method, the planar 
structures (fractures, joints, bedding places) present in the sidewall of the borehole 
are identified. The planar structures are classified as one of the following:  
• Visible layering including bedding 
• Major fractures 
• Intermediate fractures 
• Minor fractures; and 
• Open fractures 
Rotary percussion drilled boreholes were used to determine the hydrogeological 
regime at the site.  
Ten (10) In-Situ Inc. Level Troll 700™ data-logging pressure transducers were 
installed in existing boreholes around the site by the Consultants (ERM, 2005) in 
order to collect dynamic groundwater elevation data. This was to support detailed 
evaluation of horizontal and vertical hydraulic gradients as well as seasonal 
variations in hydraulic gradients. On average each borehole was monitored over a 
two year period. The rainfall at the site was also monitored over this period. 
The hydraulic conductivity for aquifers was determined by two methods viz, packer 
testing and through the insertion of ribbon NAPL samplers in some boreholes.  
A straddle packer system was used to isolate each geological unit and pressure tests 
were conducted on selected boreholes to evaluate the hydraulic parameters of each 
interval. The tests were carried out using leak-proof upper and lower packers in order 
to isolate specific lengths of the borehole. The spacing of the packers was 
approximately 1.50 meters. The tests consisted of pumping water into the isolated 
zone at three different pressures: a, b and c in the following sequence: 
1st: 10 minutes at low pressure………………………………………...a 
2nd: 10 minutes at medium pressure……………………………….b 
3rd: 10 minutes at high pressure……………………………………....c 
79 
4th: 10 minutes at medium pressure…………………………………b 
5th: 10 minutes at low pressure………………………………………….a 
The flow of water into the isolated zone was measured by means of a flow meter for 
each of the pressure stage periods. The duration of the pressure test for each interval 
was recorded. The starting pressure for each interval was 50 kPa and this was 
increased by 50 kPa increments. The permeability (K) of each discrete test interval 
was estimated using the relationship between the flow rate and the applied hydraulic 
pressure using the following equation: 
 = (⁄)     Equation 3-1 
Where: 
r = radius of borehole 
R = assumed radius of influence  
L = length of the test interval  
Q = flow rate  
Pi = net injection pressure  
Drawdown tests were conducted on selected boreholes by removing water at a 
constant rate of 1.2 L/minute over a period of 1 hour, using a peristaltic pump. Ribbon 
NAPL samplers were only installed in boreholes with water level difference before 
and after the drawdown tests was < 1.5 m3. This was based on the assumption that 
the borehole intersected sufficient transmissive fractures for installation of the ribbon 
NAPL samplers. The transmissivity measurements are taken by recording the liner 
position with time, the tension on the liner and the excess head driving the liner.  
All the newly-drilled wells were gauged by the Consultants using a SolinstTM
electronic dip-meter/interface probe.  
The location of all investigation boreholes (weathered and fractured zones) are 
shown in Figure 3-8.  
Shallow piezometers (up to 7 m bgl) drilled along the boundary of the intermittent 
streams were gauged to determine whether these streams were gaining or losing 
3 This assumption was based on failed attempts at installing ribbon NAPL samplers in low 
transmissivity boreholes. 
80 
streams. The piezometers were designed so that the screened sections were set 
below one m bgl to the base of each hole. A plain riser was installed to rise 
approximately 500 mm above ground level. Each piezometer was equipped with 63 
mm uPVC screen and solid casing. A 2-5 mm silica gravel pack was installed across 
the screened interval and a bentonite seal was installed within 0.1 m of ground level 
in order to minimise the potential for ingress of surface water. All piezometers were 
capped.  
3.2.4 Hydrogeochemical and biogeochemical characterisation 
A low-flow peristaltic pump was used to retrieve groundwater samples. The field 
sampling procedure is illustrated by means of a flow diagram (Figure 3-9). The low-
flow pump was installed within the screen zones. The position of the pump varied 
depending on whether or not the screen depth was known. Groundwater was 
pumped at flow rates of 0.1-3 L/minute (depending on the volume of water in the 
well). The field parameters (Temperature, pH and electrical conductivity) were 
monitored using an overflowing bucket system until the parameters stabilize. The 
overflow bucket system consisted of field probes that were placed within a jug. This 
was allowed to overflow into a bucket, ensuring recycling of water within a jug. The 
volume of water purged depended on how quickly the field parameters stabilized. In 
order to determine the sampling set-up, the well volume at each well was first 
calculated, and thereafter a pump was selected. 
Retrieved groundwater samples were analysed for chlorinated hydrocarbons. 
Additionally, selected samples were analysed for toxicity, an empirical measure of 
biodegradability of samples, compound specific isotope analyses and microbial 
analysis (focus on PCE degradation).  The wells purged during the additional sample 
collection were also measured for dissolved oxygen and redox potential. Samples 
collected for compound specific isotope analysis (CSIA) were sent to the University of 
Waterloo (37-Cl isotope analysis) and Microseeps (13-C isotope analysis), where the 
ratio of stable isotopes were determined using an Isotope Ratio Mass Spectrometer 
(IRMS). The method for sample collection and preservation for CSIA as described by 
Hunkeler et al., 2008 was used.    
The sampled boreholes were analysed for toxicity by measuring the percentage 
inhibition using a Hach Eclox luminometer. Toxicity screening was undertaken on site 
and was a rapid inexpensive methodology to provide an overview of the site 
81 
toxicological profile. The field kit luminometer contains Vibrio fischeri luminescent 
bacteria. Individual samples were placed in the luminometer and the amount of light 
made by the bacteria after exposure to the sample is compared to the amount of light 
made by the bacteria after exposure to a control. The control used was 2% NaCl 
solution. 
The biochemical oxygen demand (BOD) determination is an empirical test which 
determines the relative oxygen requirements of wastewaters, effluents and polluted 
waters under aerobic conditions. The test measures the molecular oxygen utilised for 
the biochemical biodegradation of organic material (carbonaceous demand) during a 
specific period of time. Chemical oxygen demand (COD) is used to indirectly 
measure the amount of organic compounds in water. The Chemical oxygen demand 
(COD) test uses a strong chemical oxidant in an acid solution and heat to oxidize 
organic carbon to CO2 and H2O (Boyles, 1997). The biodegradability index of a 
sample can be determined from the BOD/COD ratio. The following applies: 
BOD/COD > 0.4………………………………biodegradable 
0.2<BOD/COD<0.4………………………medium biodegradable 
BOD/COD< 0.2………………………………non-biodegradable/very little biodegradability. 
Samples for mibcrobial analysis were collected in sterilized glass bottles. The 
analyses were undertaken at the University of the Free State Metagenomics Platform 
under the supervision of Professor E. Van Heerden. The purpose of the analysis was 
to detect the presence of genomic DNA. Water samples were filtered through a 
0.22μl polyethersulfone filter and demonic DNA was extracted from each filter using 
FastDNA SPIN for Soil Extraction (MP Biomedicals) kit. The DNA extracts were 
visualised on a 1% agarose minigel with ethidium bromide and UV-illumination after 
electrophoresis in TAE buffer (90 V for 1 hour). DNA concentration and purity were 
assessed using a Nano Drop spectrophotometer 3300 fluorospectrometer. 
82 
Figure 3-8: Location of all boreholes drilled at the Investigation Site (results from SRK Consulting, 1995; SRK Consulting, 2001, ERM, 2005; ERM, 2012) 
83 
Figure 3-9: Flow diagram for groundwater sampling at the Investigation Site (from ERM, 2008)
84 
3.3 Results 
3.3.1 Regional geology 
 The regional geology of the area is dominated by the following lithologies (Figure 3-
10 and Figure 3-11, cross section locations are indicated on Figure 3-3): 
• Pre-Karoo basement: The regional pre-Karoo basement is characterised by 
andesitic lavas of the Pretoria Group Hekpoort formation, various quartzite 
marker units of the Witwatersrand Group (i.e. Orange Grove and Turffontein 
quartzites), the Black Reef quartzite and the top contact of the Chuniespoort 
dolomites. The outcrop patterns of the marker horizons show circular deformation 
around the domal Vredefort impact structure. These folds consist of a steeply 
dipping overturned syncline, tight anticline and open syncline plunging at 4-8° 
NW. The basement suboutcrops 5km north and 14km west of the Investigation 
Site and gently deepens southeastwards under Karoo cover rocks. 
• Karoo Supergroup: The Karoo Supergroup is characterized by a gently undulating 
(elongated domes and basins) mixed succession (in the order of 40m thick) of 
mudstone, shale, siltstone, sandstone and coal seams (collectively called the 
Ecca formation), with a basal tillite (Dwyka formation). The undulating features of 
the succession mimics the surface topography of the underlying pre-Karoo 
basement, with Karoo sequence domes and basins overlying basement highs 
and lows, respectively. An overlapping unconformable relationship is evident 
between the lower units of the Karoo succession against the basement highs with 
the basal (C1), middle (C2) and occasionally, the top (C3) coal seams pinching 
out against the basement highs. The coal seams average between 4-6m thick 
reaching a maximum in basement lows of 10-12m thick. 
The lower mixed succession (lower Ecca formation) contains variable, but often 
significant, mudstone and shale units which range from less than 1m to greater 
than 10m thick.  The lower succession of the Ecca formation suboutcrops below 
unconsolidated overburden to the north of the Investigation Site. The upper part 
of the formation (sandstone dominated) suboutcrops south and west of the 
Investigation Site. 
• Unconsolidated overburden generally varies between 2- 7 meters thick. This 
increases to 10-18 meters NNE of the Investigation Site and >24 meters in the 
area of the primary River. 
• Regional structures/features:  
o Late Karoo sub-horizontal dolerite sill intrusion, generally between 30-60 
m thick is a major geological feature in the region resulting in extensive 
85 
areas of dolerite sub outcrop. Sills may change elevation within the Ecca 
formation along steep linear “roll” structures.  
o The main linear structural trends are NNW-SSE and E-W. A NS trend is 
prevalent in the south of the region. These features are characterized by 
linear drainage features, faults and airborne magnetic anomalies, the 
latter often locally confirmed as dolerite dykes in outcrop. The NNW trend 
is associated with dolerite dyke intrusion with minor faulting and is 
dominant in the east of the area. The E-W trend is associated with both 
faulting and/or dolerite dyke intrusion.  
o Major dolerite dykes are dominant north of the primary River.  
o A possible E-W fault with a down throw of 55m to the north may be 
present close to the primary River between boreholes NVL0787 and 
NVL5037 
o The depth of weathering is highly variable but generally varies between 2-
18 meters in all lithologies with a maximum record showing 35 meters in 
the shale (Ecca Formation) 
86 
Figure 3-10: North-south trending regional geological cross section (modified from ERM, 2008)
87 
Figure 3-11: East-west trending regional geological cross section (modified from ERM, 2008)
88 
3.3.2 Local geology  
In general the local geology underlying the Investigation Site consists of overburden 
and weathered rocks with high to moderate permeability and bedrock characterised 
by low permeability. The general site stratigraphy is shown below (modelled using 
RockWorks15 software, with the Inverse Distance Weighted algorithm). Further 
details on the lithostratigraphy as well as the structural discontinuities are provided 
below. Stratigraphic profiles across the source zones are shown in Section 3.3.4.  
Vertical exaggeration: 25
Figure 3-12: Stratigraphic model
The key to the stratigraphic codes shown in the stratigraphic model above are as 
follows: 
OVB: Overburden, unconsolidated material includes aeolian sand, silt and clay 
DO: Dolerite intrusions 
SST10: Dominantly sandstone unit, subordinate siltstone (massive), shale and 
mudstone 
SH10: Dominantly shale or mudstone unit, subordinate siltstone (banded). Rare thin 
sandstone and coarse sediment units 
C4: Coal unit, C4 or leader seam 
U5: Alternating shale, siltstone and sandstone on centimetre / decimetre scale 
C3: Coal unit, C3 or Top seam, subordinate clastic sedimentary partings 
89 
U4: Alternating shale, siltstone and sandstone on centimetre / decimetre scale 
C2: Coal unit, C2 or Middle seam, subordinate clastic sedimentary partings 
U3: Alternating shale, siltstone and sandstone on centimetre / decimetre scale 
U1: Alternating shale, siltstone and sandstone on centimetre / decimetre scale 
DW: Dwyka Tillites 
LAV: Pre-Karoo Basement Lava 
2-D Lund resistivity methods allowed for discrimination between low and high 
resistivity sub-surface features. Resistivity profiles are provided in Appendix B. In 
general, the correlation of geological zones identified through core logging with the 
resistivity profile data is summarised as follows: 
• The high resistivity zones (> 70 Ohm-m) in the geophysical profiles can be 
correlated with unweathered rock types (i.e. dolerite sill and Pre-Karoo 
Ventersdorp lava) 
• Moderately resistive zones (20-70 Ohm-m) correlate with Karoo Ecca 
sedimentary rocks (20-50 Ohm-m for sandstone and 10-20 Ohm-m for shale), 
partly weathered dolerite sill or thinly developed dolerite sill 
• Low resistivity zones (5-20 Ohm-m) reflect a variety of different geological 
features (e.g. overburden, clay-rich residual weathering profiles, possible 
structural breaks such as faults or large fractures in otherwise competent 
bedrock), zones of increased groundwater content (preferential groundwater flow 
paths), or potentially contamination zones.  
• Very low resistivity zones (<5 Ohm-m) reflect possible conductive groundwater, 
overburden, in situ clay rich profiles or very weathered Ecca sedimentary rocks 
(likely to be shale dominant). 
The resistivity profiles are generally characterised by horizontal stratification (for e.g. 
Traverse 102 and 103, Figure 3-13, with very low resistivity at the surface and 
increasing resistivity with depth reflecting a horizontal sequence of overburden, in situ
clay, weathered bedrock and bedrock (dolerite sill). The depth to the base of the 
overburden and in situ clay is recognisable due to the resistivity contrast between the 
two features. However, in areas where the Ecca sedimentary sequence outcrops at 
surface, this contrast is not easily discernible due to the Ecca sedimentary rocks 
being reflected as having low resistivity. Within specific resistivity profiles there are 
marked horizontal variations. In these instances (e.g. Traverse 109, Figure 3-14) 
zones of high and low resistivity are often separated by a steeply orientated break, 
reflecting a change in rock type (between dolerite and Ecca sedimentary rocks). 
90 
Figure 3-13: 2-D resistivity profiles of Traverses 102 and 103
91 
Figure 3-14: 2-D resistivity profile of Traverse 109
From the core and borehole logs for the site (Appendix A) it was possible to 
distinguish between the upper sand dominated soil, fill and transported material and 
underlying in-situ clay dominated weathered bedrock. The former is classified as 
overburden and the latter is part of the weathered bedrock profile. The overburden 
consists of sandy transported material up to a depth of approximately 6.5 m bgl, 
thereafter the material becomes predominantly clayey as a result of the weathering of 
dolerite. Overburden is more developed in the southern portion of the site boundary 
and adjacent to the intermittent stream to the south.  
A sandy and silty clay layer, varying up to 8m thick, is present in all boreholes drilled 
at the site area, except along the northern boundary. It mostly represents an in situ 
extreme weathering of dolerite and Ecca shale/siltstone. A typical profile of the 
weathered zone at the Investigation Site is shown below (Figure 3-15). Weathered 
Ecca sedimentary formation (consisting of residual sandstone or residual 
mudstone/shale) underlies the colluvium. The colluvium or the weathered Ecca 
sedimentary formations overlies weathered dolerite. Particle size distribution analysis 
determined from the analysis of Borehole EDC2 indicates a decrease of clay content 
with depth and a corresponding increase in gravel content with depth (Figure 3-16). 
This marks the transition from clayey residual dolerite or Ecca sedimentary rocks to 
highly weathered bedrock dolerite (Figure 3-17). The weathering in the dolerite varies 
92 
laterally and vertically across the site. Zones of deep (>35 m) and shallow (<15 m) lie 
adjacent to each other. Elongated pockets of deeper weathering show a primarily 
NW-SW with a second subordinate N-S trend which tend to coincide with synclinal 
and anticlinal flexure axes in the underlying dolerite and Ecca sedimentary rocks. The 
surface weathering base slopes to the SE across the site with a gradient of 0.0125 
(0.718°).  An isopach map showing the base of weathering elevation (or the upper 
contact of bedrock) at the Investigation Site is shown in Figure 3-18.  
Figure 3-15: Typical weathered zone profile at the site
93 
0-0.55
1.4-1.6
2.2-2.4
% Gravel
3.3-3.5
4.6-5.0 % Sand
5.1-5.3 % Silt
 6.6-6.8
 9.0-9.3 % Clay
0 20 40 60 80
Percentage (%) 
Figure 3-16: Particle size distribution of borehole EDC2S
Figure 3-17: Picture taken of roto-sonic core (EDC2S) showing the
transition from residual clay dolerite to highly weathered dolerite
The geotechnical properties of the weathered or residual lithologies are shown in 
Table 3-5. The fraction of organic carbon in this zone is higher than that observed for 
the bedrock. Residual and weathered Ecca sedimentary rocks have lower 
permeability (classified as very low permeability; Lambe and Whitman, 1969) than 
that observed for weathered and residual dolerite (classified as medium permeability; 
Lambe and Whitman, 1969). Soil permeability analysis of the residual and weathered 
Ecca sedimentary and dolerite samples show variations in permeabilities. The 
residual Ecca sedimentary samples exhibited permeability values ranging from 2.5 x 
10-6 m/d to 5 x 10-6 m/d, while the weathered dolerite samples exhibited permeability 
values ranging from 2.6 x 10-6 m/d to 2.1 x 10-3 m/d.  
94 
Depth (mbgl) 
Figure 3-18: Base of weathering elevation
95 
Bedrock characteristics are determined through logging of holes during diamond core 
drilling as well as through downhole geophysical analysis in selected core holes. The 
effectiveness of the downhole geophysical tools used has been differentiated by 
those that showed the best responses for the varying lithological units and geological 
features (Table 3-5). The responses observed using natural gamma, the acoustic 
televiewer, density probe, calliper, and resistivity probe corresponded closely to the 
geological/core logs. The following planar structures were identified using the 
“fracture pick” analysis on the diamond drill core holes using the acoustic televiewer 
data: 
• Visible layering including bedding 
• Major fractures 
• Intermediate fractures 
• Minor fractures 
• Open fractures. 
A summary of the fracture pick analysis is provided in Table 3-6. Fractures within the 
bedrock are most developed in weathered zones. In general, sub-horizontal (i.e <300) 
fractures or bedding planes dominate over moderate (i.e. 300 - 600) to steeply (i.e. 600
– 900) dipping fractures/bedding planes. 
Table 3-5: Effectiveness of the borehole geophsyical tools used at the Investigation Site 
Borehole Geophysical Tool Effective
Natural Gamma (GRTH) Yes
Self potential (SP) No
Acoustic Televiewer (TIMM and AMPM) Yes
Density (DENB and DENL) Yes
Resistivity (FE1 and FE2) Yes
Fluid Temperature (TEMP) No
Calliper (CATH) Yes
Fluid Conductivity (CONN) No
A core retrieved through diamond core drilling is shown in Figure 3-19  and shows the 
general bedrock succession underlying the Investigation Site.  
96 
Table 3-6: Summary of fracture pick analysis on selected core holes 
SWD1 SWD02 MLR01 CAP01 ASH01
Dolerite Sill Depth: 0-32.88 m bgl Depth: 0-33.59 m bgl Depth: 0-22.42 m bgl Depth: 0-35.51mbgl Depth: 23.49-79.64 m bgl
· Upper weathered zone · Moderately intense · Intense multidirectional · Moderately intense · Weathered zone – 
(to 11.5 m bgl) - intense multidirectional dipping sub-horizontal to multidirectional dipping intense multidirectional 
ENE to ESE dipping sub- sub-horizontal fractures in moderate sub-horizontal fractures sub-horizontal dipping 
horizontal fractures upper part (0-23 m bgl). dipping fractures and NW in upper part (0-23 m bgl). fractures and NE to E and 
· Central part of sill (11.5- Open multidirectional to NE, E and W dipping · Central part of sill (23- WNW moderate dipping 
26.0 m bgl) – a few minor sub-horizontal dipping steep fractures. 30.5 m bgl) – no fractures fractures in weathered 
sub-horizontal fractures fractures at 9-14.6 m bgl · Open N, NE and E recorded zone (to 36 m bgl) with 6 
and 3 major NE dipping · Central part of sill (23- dipping steep fractures at · Lower part of sill (basal open fractures between 
sub-horizontal fractures 31.5 m bgl) – a few minor 14.75-16.0mbgl contact zone) (30.5- 27.75-31.75 m bgl (with 1 
(14-17 m bgl) sub-horizontal fractures 35.51 m bgl) – very steep SW dipping 
· Lower part of sill · Lower part of sill (basal intense multidirectional fracture) 
(subsurface weathered contact zone) (31.5- dipping sub-horizontal · Central part of sill (36- 
basal contact zone) (26.0- 33.59 m bgl) – very minor to intermediate 62 m bgl) – weakly to 
32.88 m bgl) – very intense fractures. Open SW moderately intense ENE 
intense easterly dipping sub- moderately dipping to ESE sub-horizontal to 
ENE to ESE dipping horizontal minor to fracture at 32.2 m bgl moderate dipping 
sub-horizontal minor intermediate fractures · No steep dipping fractures and SW to NW 
fractures and a SSE · No steep dipping fractures recorded in and E to NE steep dipping 
dipping (27.25 m bgl) and fractures recorded dolerite sill fractures 
NW dipping (29.75 m bgl) · Multiple dolerite intrusion 
steep major fractures zone (64.25-67.27 m bgl) – 
intense sub-horizontal to 
steep E dipping fractures 
present between 62- 
69 m bgl. 
97 
Karoo Ecca Depth: 32.88-66.20 m bgl Depth: 33.59-64.42 m bgl Depth: 22.42-42.00 mbgl Depth: 35.51-62.87 m bgl Depth: 0-23.49-76.45 m 
Sedimentary bgl · Dominated by NE to SE · Dominated by E to NE · Dominated by S to SW · Dominated by NE and 
Sequence dipping sub-horizontal dipping sub-horizontal dipping sub-horizontal SW sub-horizontal to · No data 
bedding planes with 2 bedding planes. bedding planes moderate dipping 
open fractures at 40.5- · No moderate to steep · Moderate to steep NW to bedding planes / 
40.75 m bgl (dipping SE dipping fractures recorded NE dipping fractures 27- fractures 
and SW) in a shaly (apart from 1 open 32 m bgl · No steeply dipping 
sandstone of the SH10 fracture – see below) · U5 grit and sandstone fractures recorded 
shale unit · Open WNW moderate (32.14-33.92 m bgl) – 3 · More intense 
· Occasional NW, ENE dipping fracture at 46.6 m open southerly and NW development of 
and SE moderate dipping bgl in shale (SH10 dipping sub-horizontal sub-horizontal to 
fractures unit) fractures moderate dipping 
· C3 coal seam (45.15- · Open SE dipping bedding planes/fractures 
47.80 m bgl) – intense NE sub-horizontal fracture at (46.5-47.5 m bgl) along 
to SE moderate dipping 49.4 m bgl close to shale contact zone between C4 
fractures including 3 open (SH10) / C3 coal seam leader seam and shale 
fractures in lower portion contact (SH10) 
of the C3 seam (46.5- · More intense E to NE · Open N dipping 
47.5 m bgl) dipping sub-horizontal sub-horizontal fracture/ 
bedding planes in lower bedding plane at 
part of C2 coal seam (55- 46.5 m bgl in shale 
58 m bgl) (SH10) 
· Open NE moderate 
dipping fracture at 
47.5 m bgl in siltstone/ 
shale (SH10) 
98 
Pre-Karoo Depth: 66.20-89.08 m bgl Depth: 64.42-81.38 m bgl Depth: 42.00-79.28 m bgl Depth: 62.87-90.38 m bgl No data
Ventersdorp · Subsurface weathered · Moderately intense · Upper subsurface · Weak to moderately 
Lava zone (66.20-71.0 m bgl) – multidirectional weathered zone (42.00- intense multidirectional 
NE to SSE moderate sub-horizontal to steep 42.57 m bgl) – intense sub-horizontal to 
dipping minor and major dipping fractures narrow zone of ENE and moderate dipping 
fractures including a NE throughout the lava unit SW shallow, moderate fractures throughout the 
moderate dipping open · Open SW dipping and steeply dipping minor lava unit 
fracture at 69.5 m bgl sub-horizontal fractures at and major fractures 
· Minor fracturing 72.4 m bgl and 80.2 m bgl · Sub-horizontal and steep 
throughout most of the minor fracturing 
remainder of the lavas throughout most of the 
· Narrow intense zone of remainder of the lavas 
sub-horizontal to steep with steep major fractures 
multidirectional dipping at 48.25 m bgl (NNE 
minor fractures at 81.0- dipping), 53.75 m bgl 
81.5 m bgl corresponding (SSW dipping) and 
to a subsurface 66.75mbgl 
weathered zone in · Narrow intense zone of 
amygdaloidal lavas moderate to steep minor 
fractures at 72.62 m bgl 
coinciding with contact 
between upper massive 
lava and lower 
amygdaloidal lava 
99 
Highly fractured dolerite
Open vertical fracture in 
unweathered dolerite (~15 
mbgl) 
Massive dolerite with
low fracture frequency 
(19-30 mbgl) 
Dolerite - Ecca Group
   contact (~32 mbgl) 
Karoo Ecca Sandstone
100 
Carbonaceous Shale 
(Karoo Ecca) 
Coal
(Karoo Ecca) 
Highly weathered 
and fractured 
andesitic pre-
Karoo basement 
(Ventersdorp 
Lava) 
Figure 3-19: Photographic log of core hole BH63C
101 
The dolerite sill has intruded the lower portion of the SST10 sandstone unit (within 
the Ecca sedimentary sequence) in the Investigation Site area.  The elevation of the 
sill lower (basal) contact has been modelled as an isopach map using GIS software 
(Figure 3-20). In the site area it generally slopes to the SE with a gradient of 0.0142 
(0.81°). However, abrupt changes in elevation of the lower contact indicate the 
presence of dolerite rolls in the site and ENE, W and SW of the site. The dolerite sill 
is generally medium grained becoming fine grained within a few metres of the lower 
contact. The contact is characterised by a very fine grained chilled margin up to 0.5 m 
thick at depths of 31-34 m bgl. The dolerite displays orthogonal jointing (sub-vertical: 
70°-90° and sub-horizontal: 0°-20°). The dip directions are multidirectional in the 
shallow weathered zone and parallel to the basal contact in the basal contact zone 
i.e. the fractures dip east within the dolerite sill for most of the Investigation Site. A 
very high frequency of open weathered joints occurs in the surface weathered zone. 
A high frequency of partly open to closed joints is present in the lower portion of the 
sill in a zone up to 7m wide above the lower contact. Joint development in the dolerite 
sill is likely to be in response to the following events in the structural history of the 
rock: 
• An orthoganol extensional joint system (3 joint sets orientated perpendicular to 
each other with 2 sub-vertical joints and a sub-horizontal joint orientation) as a 
consequence of tensional stresses developed during cooling, solidification and 
contraction in the dolerite sill.  
• Extensional joint sets developed in the fold/flexure zones parallel and normal to 
the flexure axis i.e. NW-SE to NNW-SSE and NE-SW to ENE-WSW. 
A NNW-SSE trending dolerite dyke, with evidence of multiple dolerite intrusion at ~ 
65.25-66 m bgl, is present below the dolerite sill under a portion of the historical 
waste sites (Figure 3-20). This feeder zone does not cross-cut through the dolerite sill 
and is therefore assumed to be a concurrent feeder zone rather than a later dyke. 
The dolerite dyke has displaced Ecca sedimentary rocks upwards in a local horst 
structure.  
The dominant sedimentary units underlying the site (generally occurring below the 
dolerite sill), in terms of thickness are fine grained rocks consisting of the SH10 shale 
unit, U4 mixed shale, siltstone, sandstone unit and Dwyka tillite unit. The sandstone 
units higher in the succession, i.e.SST10 and SST11 are not significantly developed 
having been displaced by the intrusion of the dolerite sill. The C3 and C2 coal units 
and lower units in the succession only occur along the west, south and east periphery 
102 
of the site as they are pinched out against the basement palaeo-high under the site. 
Sub-vertical joints are recognised occasionally in the sandstone and siltstone units. 
Sub-horizontal joints are not easily recognisable as they are likely to be coincident 
with bedding and contact planes in the sedimentary sequence. The upper contact of 
the C3 coal seam has been modelled (Figure 3-21) and illustrates the presence of 
basins and swells in the Ecca sedimentary sequence which generally reflects the 
underlying pre-Karoo basement relief. A NW-SE swell axis runs under the site and 
continues SE of the site. The dip of the sedimentary sequence is radial outwards from 
the site with dips in the order of 1.20-2.69°. 
Unweathered crystalline Ventersdorp andesitic lavas are found at depths ranging 
from 30 -40 mbgl in the northern portion of the site to 60 – 70 mbgl along the south 
western and south eastern boundaries of the site. The lavas display alteration, 
bleaching and quartz veining in the upper portion of the sequence (generally 0 – 10 m 
thick), immediately below the contact with the Dwyka tillite (Karoo Ecca sedimentary 
sequence). This is a pre-Karoo paleo-weathering feature. The lavas are fractured and 
jointed. The joints are closed to partly open and slightly to moderately weathered. 
Fractures in the lavas are generally moderately (200 – 450) to steeply dipping (700 – 
900) with chlorite and slickenside surfaces. The basement lavas are characterised by 
an undulating topography with highs and lows (Figure 3-22) and generally dip 2.230
south easterly.  
The geotechnical properties of selected rock core samples are provided in Table 3-7. 
The unweathered dolerite sample contains no measurable amount of organic carbon 
nor any measurable primary porosity or permeability. The Ecca sandstone sample 
has trace levels of organic carbon. It has a low primary porosity (3.7%) and very low 
horizontal and vertical permeabilities. The bedrock core analytical data indicates that 
groundwater flow and solute transport within the bedrock units are likely dominated 
by flow within secondary porosity features such as fractures, joints, bedding planes 
and faults. The Ecca sedimentary sequence has too low values for porosity and 
permeability to support advective transport within the rock matrix; resulting in very 
limited diffusion of contaminant mass into the rock matrix.  
The results obtained pertaining to fracture characteristics as they relate to the 
mobility of DNAPL within source zones are discussed in greater detail in Section 
3.3.4.2.
103 
Figure 3-20: Dolerite sill lower contact elevation
104 
Figure 3-21: C3 coal seam upper contact elevation
105 
Figure 3-22: Pre-Karoo basement elevation
106 
Table 3-7: Geotechnical results of samples retrieved from soil and bedrock cores 
Borehole Depth Lithology f Dry bulk Rock Porosity Non - Capillary Vertical Horizontal MoistureID OC density density Pore space Permeability Permeability Content
(mbgl) (%) (g/cm3) (g/cm3) (%) (%) (Darcy) (Darcy) (%)
BH 221 3.0-3.3 weathered sandstone 0.59 ND ND - - - ND
BH 221 4.3-4.5 residual sandstone 5.44 1.77 ND - - 6.02E-06 ND 19.8
BH 221 6.0-6.2 residual sandstone 6 1.647 ND - - 3.06E-06 ND 22
BH 221 9.0-9.2 weathered mudstone 9.41 1.548 ND - - 3.39E-06 ND 13.5
14.7 (Capillary 
EDC2 0.55 residual dolerite ND 1.636 ND Pore space) 21.9 1.26E-02 ND 10
6.5(Capillary 
BH 61 2.0-2.15 residual dolerite 5.3 1.798 ND Pore space) 28.4 2.59E-03 ND 17.6 
BH 61 2.8-3.0 residual dolerite 5.61 1.639 ND - - 3.21E-06 ND 21.4
BH 61 3.3-3.45 residual dolerite 6.62 1.624 ND - - 1.28E-05 ND 23.3
CAP 1 12-12.5 Weathered Dolerite < 0.05 2.86 2.87 - - - - ND
CAP 1 22-22.5 Unweathered Dolerite < 0.05 2.91 2.98 - - - - ND
SWD 1 20 – 20.5 Unweathered Dolerite < 0.05 2.93 2.93 - - - - ND
Unweathered Ecca 
SWD 1 35 – 35.5 Sandstone 0.05 2.71 2.61 3.7 - 1.12E-02 1.60E-02 ND
Unweathered 
SWD 1 76 – 76.5  Basement Lava < 0.05 2.78 2.79 - - - - ND 
Where: 
- : no measurable porosity/permeability 
ND : not determined
107 
3.3.3 Geohydrology 
Multiple water strikes were encountered in the upper 0.5-7.5 meters at the site 
(Figure 3-15). Groundwater was typically encountered between 0.5 and 1.0 m bgl 
(Figure 3-23).    
Typically, depth to groundwater is a function of surface elevation at the site (Figure 3-
24). Pressure transducer data collected at the site indicate a seasonal response to 
groundwater elevation, with an increase noticed during the wet months and a 
decrease in groundwater elevation observed during dry months (Figure 3-25). The 
transducer data also indicates correlated groundwater elevation patterns within the 
weathered and unweathered portions of the dolerite sill.  
Regional topographic data as well as a review of groundwater elevation data indicate 
the presence of a north-south trending hydrogeological basin divide at the site. 
Groundwater flows to the northeast in the northern portion of the site and from north 
west to south east in the southern portion (Figure 3-26). The hydraulic gradient is 
0.01 in both flow directions. 
1490
water strike
1485
1480
1475
1470
1465
1460
1455
1460 1465 1470 1475 1480 1485 1490
surface elevation (mamsl) 
Figure 3-23: Water strike elevation as a function of surface elevation
108 
water strike elevation (mamsl) 
1490
1485
1480
y = 0.9886x + 15.302 
1475 R² = 0.9774 
1470
1465
1460
1455
1450
1450 1455 1460 1465 1470 1475 1480 1485 1490
Surface elevation at borehole (mamsl) 
Figure 3-24: Groundwater elevation as a function of surface elevation at borehole
 
The following hydrostratigraphic units are identified at the site: 
• In situ clay and dolerite hydrostratigrahic unit: This represents all groundwater 
present within the surface weathered (clays) and unweathered zone of the 
dolerite sill. Groundwater flow within the weathered portion is controlled by 
intragranular flow (controlled significantly through the thickness, spatial extent 
and continuity of sandy water-bearing units). Groundwater flow within the 
unweathered portion of the dolerite aquifer is within the fracture networks. Given 
the higher porosity of the weathered compared to the unweathered zones, 
groundwater is largely yielded from the weathered portion of the dolerite aquifer. 
The weathered and unweathered sections act as a single hydrogeologic unit, but 
the physical flow properties within the aquifer vary depending on the amount of 
weathering. This is a complex system of unconfined and semi confined units. 
• Karoo Ecca hydrostratigrahic unit: The Karoo Ecca hydrostratigraphic unit is 
spatially representative across the far-field and near-field scale, but shows 
variability in its hydraulic properties. At the site, it is spatially continuous, except 
where intruded by the dolerite sill. This is a semi-confined unit. Groundwater 
strikes are commonly observed at the contact zone between the dolerite sill and 
the Karoo Ecca sedimentary rocks.     
• Pre-Karoo Basement hydrostratigraphic unit: This hydrostratigraphic unit is 
represented by the andesitic basement lavas of the Ventersdorp. Static water 
levels measured for boreholes drilled within this aquifer indicated that it is a semi-
confined unit. 
109 
Groundwater Elevation (mamsl) 
Hydraulic conductivity results from packer testing and from ribbon NAPL samplers 
installation indicate variable hydraulic conductivity values for the dolerite and the 
Karoo Ecca hydrostratigraphic units (Table 3-8). Hydraulic conductivity results 
obtained from packer testing represent the bulk hydraulic conductivity for fractures 
and matrix over the length of the packer. Ribbon NAPL samplers were selectively 
installed in high transmissivity boreholes (Figure 3-27) and the measurement was 
depth discrete to regions of high ratios of fractures. Fractures in both the dolerite and 
the Karoo Ecca hydrostratigraphic units have high conductivities, with the hydraulic 
conductivity for the fractures within the Karoo Ecca unit ranging between 0.54 – 5.48 
m/day and those for the dolerite aquifer ranging between 0.15 – 0.45 m/day. Bulk 
hydraulic conductivity for the Karoo Ecca hydrostratigraphic unit was calculated to be 
a maximum of 0.0078 m/day while the bulk hydraulic conductivities for the dolerite 
aquifer and the pre-Karoo basement aquifer was determined to be <0.0 m/day in all 
cases using the Packer testing methodology (Table 3-8).  Analysis of vertical 
gradients between the shallow overburden boreholes and the weathered bedrock 
boreholes and between the weathered bedrock boreholes and competent bedrock 
boreholes were inconclusive. Vertical gradients across the site were not consistently 
upward or downward oriented (Figure 3-26). Calculated gradients ranged between 0 – 
0.06 meters per meter indicating that vertical groundwater flow is not a significant 
component of the overall site hydrogeology.  
The results from gauging of the intermittent stream boreholes /piezometers (Figure 3-
8), taken during the dry season, are provided in Table 3-9.  The intermittent streams 
located to the north as well as to the south of the site show portions where they are 
losing streams and portions where they are gaining streams. Generally, it appears 
that groundwater discharges to the intermittent stream located to the north, while 
surface water recharges groundwater in the intermittent stream located to the south 
of the site.  
110 
1490 100
SRK25S
SRK13S 90 SWD01-S MLR01-S
1485 SRK42S MLR01-D CAP01-D
SRK13D ZAM01-M
SRK06S ZAM01-M ZAM01-S
SRK40S SRK18S 80
SRK41D SRK06S ZAM01-S ASH01-S ASH01-D
1480 SRK41S SRK22S MLR01-D SRK10S SRK12D
SRK06D SRK40D SRK06D MLR01-D SRK12S SRK17S
SRK25D SRK20S 70 SRK17D SRK20D
SRK20S SRK23D
1475 SRK20D ASH01-D 60 SRK23S SRK13S
SRK12S/D SRK13D SRK18S
SRK32S SRK23S/D
ASH01-S SRK21S SRK21D
SRK32D
1470 50 SRK22S SRK25SSRK35D
SRK17S SRK25D SRK32SSRK35S
SRK17D SRK32D SRK35S
CAP01-D
SRK36D 40 SRK35D SRK36S1465
SRK36S SRK38S SRK36D SRK04S
SSRRKK3388DS SRK38D SRK04D SRK06S
SRK37S SRK04S 30 SRK06D SRK37S
1460 SRK39S SRK04D SRK37D SRK38SSWD01-S
SRK39D SRK38D SRK39S
20 SRK39D SRK40S
SRK40D SRK41S
1455 SRK41D SRK42S
10
Rainfall
1450 0
2005/11/10 2006/02/18 2006/05/29 2006/09/06 2006/12/15 2007/03/25 2007/07/03 2007/10/11
Time
Figure 3-25: Pressure transducer data for selected boreholes at the site (results from ERM, 2008)
111 
Groundwater elevation (mamsl)

Figure 3-27: Relative transmissivities of boreholes (results from ERM, 2012)
113 
Table 3-8: Hydraulic conductivity values for the dolerite, Karoo Ecca and pre-Karoo basement 
aquifers
Borehole Depth zone (m Methodology to determine
ID bgl) Aquifer Type K (cm/sec) K
CBH42 9.14-12.19 dolerite 5.90E-4 Ribbon NAPL sampler
CBH42 12.19-15.24 dolerite 5.21E-4 Ribbon NAPL sampler
PBH221 15.24-18.23 Karoo Ecca 1.48E-3 Ribbon NAPL sampler
PBH222 18.23-21.34 Karoo Ecca 5.67E-4 Ribbon NAPL sampler
PBH223 27.43-30.48 Karoo Ecca 6.34E-3 Ribbon NAPL sampler
PBH228 15.0-18.23 dolerite 1.74E-4 Ribbon NAPL sampler
PBH229 18.23-21.34 dolerite 3.13E-4 Ribbon NAPL sampler
PBH230 21.34-24.48 Karoo Ecca 1.22E-3 Ribbon NAPL sampler
PBH231 24.38-27.43 Karoo Ecca 6.25E-4 Ribbon NAPL sampler
SWD01 10.0-18.0 dolerite <0.00 packer
SWD01 18.0-25.0 dolerite <0.00 packer
SWD01 25.0-32.5 dolerite <0.00 packer
SWD01 32.5-37.0 Karoo Ecca <0.00 packer
SWD01 37.5-45.0 Karoo Ecca <0.00 packer
SWD01 45.0-52.0 Karoo Ecca 9.03E-6 packer
SWD01 52.0-59.0 Karoo Ecca 3.01E-6 packer
SWD01 59.0-65.0 Karoo Ecca <0.00 packer
SWD01 65.0-72.0 Pre-Karoo Basement <0.00 packer
SWD02 14.0-22.0 dolerite <0.00 packer
SWD02 22.0-29.0 dolerite <0.00 packer
SWD02 29.0-33.5 dolerite <0.00 packer
SWD02 33.5-38.0 Karoo Ecca <0.00 packer
SWD02 38.0-46.0 Karoo Ecca <0.00 packer
SWD02 46.0-54.0 Karoo Ecca <0.00 packer
SWD02 54.0-62.0 Karoo Ecca <0.00 packer
SWD02 62.0-81.38 Pre-Karoo Basement <0.00 packer
CAP01 15.0-23.0 dolerite <0.00 packer
CAP01 23.0-31.0 dolerite <0.00 packer
CAP01 Dolerite/Karoo Ecca 31.0-39.0 Chill margin <0.00 packer
CAP01 39.0-47.0 Karoo Ecca 1.97E-6 packer
CAP01 47.0-55.0 Karoo Ecca <0.00 packer
CAP01 55.0-62.0 Karoo Ecca <0.00 packer
CAP01 62.0-69.0 Pre-Karoo Basement <0.00 packer
CAP01 69.0-90.38 Pre-Karoo Basement <0.00 packer
114 
Table 3-9: Groundwater vs surface water elevation for the stream boreholes/piezometers 
Co-located Surface
Depth to Groundwater elevation water elevation (mamsl,
Borehole ID groundwater (m bgl) (mamsl, GW) SW) Δ (GW-SW) Losing/Gaining stream
PBH 298-6 1.26 1457.614 1457.786 -0.172 losing
PBH299-6 1.11 1457.407 1458.035 -0.628 losing
PBH299-25 1.11 1457.407 1458.035 -0.628 losing
PBH300-6 1.18 1456.347 1456.780 -0.433 losing
PBH301-6 1.00 1455.893 1456.191 -0.298 losing
PBH301-25 0.74 1456.153 1456.191 -0.038 losing
PBH302-6 4.50 1452.859 1455.958 -3.099 losing 
PBH303-6 1.09 1454.710 1454.525 0.185 gaining 
PBH304-6 1.26 1455.237 1454.931 0.306 gaining
PBH307-6 1.45 1457.220 1454.024 3.196 gaining
PBH308-6 1.94 1456.308 1453.476 2.832 gaining
PBH308-25 1.71 1456.338 1453.476 2.862 gaining
PBH309-6 1.05 1455.090 1452.798 2.292 gaining
PBH310-6 0.17 1452.900 1451.611 1.289 gaining
PBH310-25 0.23 1453.120 1451.611 1.509 gaining
PBH311-6 1.39 1451.585 1449.880 1.705 gaining
SMPA11 2.00 1456.83 1457.93 -1.101 losing
SMPA13 0.96 1457.99 1458.25 -0.260 losing
SMPA14 0.98 1456.98 1457.19 -0.206 losing
SMPA15 0.96 1456.39 1456.38 0.011 gaining
SMPA16 0.73 1456.00 1455.99 0.006 gaining
SMPA17 1.10 1455.81 1455.85 -0.037 losing
SMPA18 0.86 1455.03 1455.02 0.009 gaining
115 
SMPA19 0.98 1454.65 1454.46 0.189 gaining
SMPA21 1.51 1454.01 1453.94 0.077 gaining
SMPA22 1.27 1453.73 1453.69 0.043 gaining
SMPA23 1.06 1453.37 1453.49 -0.116 losing
SMPA24 1.15 1453.22 1453.33 -0.110 losing
SMPA25 1.14 1452.69 1452.89 -0.195 losing
SMPA26 1.31 1451.97 1452.39 -0.424 losing 
SMPA27 1.61 1451.06 1451.19 -0.127 losing
SMPA28 1.27 1450.21 1450.34 -0.131 losing
SMPA29 1.55 1449.94 1450.31 -0.371 losing
SMPA30 1.63 1449.52 1450.02 -0.494 losing
SMPA31 1.29 1448.67 1448.63 0.044 gaining
SMPB03 3.51 1454.56 1454.33 0.230 gaining
SMPB04 3.29 1453.61 1453.60 0.011 gaining
SMPB05 2.54 1453.34 1452.90 0.446 gaining
SMPB06 1.29 1453.21 1451.96 1.250 gaining
SMPB07 1.79 1451.24 1451.33 -0.093 losing
SMPB08 1.48 1450.50 1449.30 1.199 gaining
SMPB09 1.74 1449.13 1448.34 0.791 gaining
SMPB10 1.17 1448.80 1447.71 1.093 gaining
SMPB11 1.39 1447.52 1447.71 -0.188 losing
SMPB12 1.12 1446.65 1447.22 -0.568 losing
116 
3.3.4 DNAPL source zones delineation 
Local geological and geohydrological characteristics of the source zone are provided 
in Sections 3.3.2. and 3.3.3. respectively. This section focusses on identifying the 
source zone characteristics in terms of release areas, DNAPL distribution in the 
weathered zone and DNAPL characteristics/architecture in the fractured bedrock 
zone. 
3.3.4.1 Weathered source zone delineation 
2-D resistivity anomalies, representing areas of low resistivity (high conductivity) and 
potentially DNAPL source areas or contaminated groundwater, were identified in the 
weathered and fractured bedrock zones (Figure 3-28). These areas mark sharp 
contrasts to the resistivity in the surrounding rocks and cannot be explained through 
their lithological characteristics. The locations of these anomalies correspond to 
historical release areas identified through collection of historical data. 
Results from the passive soil-gas are provided in Appendix C. Frequency distribution 
histograms of the grouped chlorinated hydrocarbon DNAPLs are shown in Figure 3-
29. All distributions are skewed to the left of the distribution curves. Results indicate 
that PCE and CHCl3 are the most prevalent chlorinated hydrocarbons, observed in 
77% and 62% respectively of the samples analysed (Figure 3-30). TCE and CCl4
were detected at 25% of sample locations.  The isomers 1,1-DCE, cis-1,2-DCE, and 
trans-1,2-DCE were detected in 7 to 12% of the sample locations. Vinyl chloride was 
only encountered in 4% of sample locations. The majority of the samples analysed 
(439 of 460 samples) have chlorinated ethene concentrations ≤ 100 µg (absorbed 
mass). The remaining sample concentrations (absorbed mass) for chlorinated 
ethenes ranged between 100 and 800 µg. The result from a single location was high 
for the chlorinated ethanes, with the rest of the localities (459 out of 460) having 
concentrations ≤ 100 µg (absorbed mass). Absorbed mass concentrations for 
chlorinated benzenes and chlorinated methanes were low with 99.9% of the samples 
showing concentrations of ≤ 100 µg (absorbed mass). The maximum concentration 
detected for the chlorinated benzenes was 6.51 µg. The concentration of absorbed 
gaseous mass from one sample location was between 100 and 200 µg for chlorinated 
methanes.  
117 
The spatial variability of the grouped targeted hydrocarbons can be characterised by 
the calculation of the coefficient of variation (Table 3-10). The coefficient of variation 
is a normalized measure of variability that is independent of the measurement scale 
and can thus be used to compare data sets. The chlorinated benzenes and the 
chlorinated ethanes showed a higher amount of spatial variability than the other 
grouped hydrocarbons (i.e. TPH, chlorinated ethenes and the chlorinated methanes). 
Only trace levels of chlorinated benzenes are found at the site. 
The spatial distribution, using inverse distance weighting (IDW), of chlorinated 
ethenes with concentrations greater than 72 µg (absorbed mass) is shown in Figure 
3-31. The chlorinated ethenes are predominantly located close to the waste sites and 
the historical organics plant. In contrast the spatial distribution of the chlorinated 
ethanes (predominantly consisting of 1,2-DCA), using the inverse distance weighting 
method (IDW), shows an inferred release point is adjacent to the effluent dams on the 
southern part of the Investigation Site (Figure 3-32). This anomaly is likely to be a 
false positive and is attribute to the spills that occurred during the bailing of free 
phase DNAPL (composition 1,2-DCA and VC) from that locality (EDC 1)  in 2007. 
The spatial distribution of chlorinated methane is shown in Figure 3-33. Chlorinated 
vapours occur at smaller concentrations of absorbed mass and are found in the 
historical production and handling facility 1. 
The soil-gas vapours of chlorinated hydrocarbon DNAPLs in the unsaturated zone 
are spatially distributed at the historical production and handling facilities and the 
redundant hazardous waste sites. Based on the results obtained from the passive soil 
vapour analysis and the 2-D resistivity anomalies, release areas are inferred at the 
hazardous waste sites and the historical plant area (Figure 3-34).  
118 
Figure 3-28: 2-D resistivity anomalies detected at the Investigation Site (Results from ERM, 2008; ERM 2012)
119 
120 
Figure 3-29: Frequency distribution histograms of grouped chlorinated hydrocarbon 
compounds found at the Investigation Site 
Where: 
Total chlorinated hydrocarbons = total chlorinated benzenes + total chlorinated ethenes + total chlorinated ethenes + 
total chlorinated ethanes + total chlorinated methanes 
Total Chlorinated Benzenes (ClBENZ) = 1,2-dichlorobenzene + 1,3-dichlorobenzene + 1,4-dichlorobenzene + 
chlorobenzene 
Total Chlorinated Ethenes (ClEthenes) = tetrachlorethene + trichloroethene + cis-1,2-dichloroethene + trans-1,2-
dichloroethene + 1,1-dichloroethene + vinyl chloride 
Total Chlorinated Ethanes (ClEthanes) =1,1,1-trichloroethane + 1,1,1,2-tetrachloroethane + 1,1,2,2-tetrachloroethane 
+ 1,1,2-trichloroethane + 1,1-dichloroethane +1,2-dichloroethane  
Total Chlorinated Methanes (ClMethanes) = carbon tetrachloride + chloroform 
Table 3-10: Descriptive statistics for grouped chlorinated hydrocarbon compounds found at 
the Investigation Site 
CIBENZ ClEthenes ClEthanes ClMethanes
Count 460.00 460.00 460.00 460.00
Minimum (µg) 0.01 0.08 0.05 0.02
Maximum (µg) 6.51 705.17 609.78 119.54
Range (µg) 6.50 705.09 609.73 119.52
Mean (µg) 0.03 17.53 2.01 1.51
Standard Deviation (µg) 0.30 73.96 28.75 8.55
Standard Error 0.01 3.45 1.34 0.40
Coefficient of Variation 12.03 4.22 14.28 5.67
Median (µg) 0.01 0.25 0.06 0.05
Mode (µg) 0.01 0.10 0.06 0.02
Confidence Level
(95.0%) 0.03 6.78 2.63 0.78
121 
1000
100
10
1
PCE TCE 11DCE c12DCE t12DCE VC 111TCA 112TCA 11DCA 12DCA 1112TetC 1122TetCA A CC14 CHC13 CIBENZ 12DCB 113DCB 14DCB
n 356 109 34 54 37 20 8 40 36 83 9 6 116 287 9 9 6 9
% 77 24 7 12 8 4 2 9 8 18 2 1 25 62 2 2 1 2
Figure 3-30: Target chlorinated hydrocarbon compounds detected during the passive soil gas survey
Where: 
n = number of detections,  
% = percentage of detections compared to the number of samples
122 
Figure 3-31: Spatial distribution of chlorinated ethene vapours at the Investigation Site (results from ERM, 2010)
123 
Figure 3-32: Spatial distribution of chlorinated ethane vapours at the Investigation Site (results from ERM, 2010) 
124 
Figure 3-33: Spatial distribution of chlorinated methane vapours at the Investigation Site (results from ERM, 2010) 
125 
 
Figure 3-34: Inferred chlorinated hydrocarbon DNAPL release areas (results from ERM, 2010) 
126 
Chlorinated hydrocarbon results obtained from the analysis of soil samples are 
provided in Appendix D. The results obtained indicate large vertical and horizontal 
heterogeneities. Horizontal spatial distributions of total chlorinated hydrocarbons are 
shown in Figure 3-35.  Chlorinated hydrocarbons are spatially located below their 
release area (Figure 3-36). The compositions mimic the types of products that were 
handled at these release areas. It is not possible to distinguish transformation 
products from parent products from the results obtained as a result of the vast array 
of chlorinated hydrocarbon raw materials, intermediates and products handled at the 
site. Maximum total concentrations were detected between 0.5 to 3 m bgl. Figure 3-
37 shows the 3-D architecture of the grouped chlorinated hydrocarbon DNAPLs 
within the weathered zone. 3-D solid models of individual chlorinated hydrocarbon 
DNAPL compounds are provided in Appendix D. The frequency distribution of the 
chlorinated hydrocarbon DNAPLs are skew to the left (Appendix D). The solid models 
maximum concentrations were filtered by excluding outliers. Chlorinated benzenes 
are not very prevalent in the weathered profile. Chlorinated ethanes and ethenes 
dominate the total chlorinated hydrocarbon concentrations. Chlorinated ethenes and 
ethanes are prevalent at the redundant hazardous waste sites as well as at the 
historical production and handling facilities. 
Monte Carlo simulations (Appendix E) were undertaken to determine the soil 
chlorinated hydrocarbon threshold concentrations. These minimum and maximum 
concentrations obtained for the predominant chlorinated hydrocarbon DNAPLs are 
provided in Table 3-11 below. Based on the minimum threshold values, nine 
boreholes are inferred to contain residual phase DNAPL: BH2, BH39, BH187, BH189, 
BH190, BH193, BH196, BH197, BH198. Oil Red O® tests were positive for the 
following sonic holes: BH02 at 0.5 mbgl, BH03 at 0.5 m bgl, BH177 at 0.3 m bgl, 
BH181 at 0.5 m bgl, BH187 at 3.5 m bgl, BH189 at 0.5 m bgl, BH190 at 0.5 m bgl, 
BH196 at 0.5 m bgl and BH197 at 0.5 m bgl.  
The results indicate that the primary source zone for chlorinated hydrocarbons as 
free phase and residual DNAPL in soils is located in the shallow soils beneath the 
redundant production and handling facilities. These high concentrations are related to 
direct releases to surface over a long period of time.  
A sharp decrease in concentrations occurs with distance from the areas of direct 
release. Very low levels of chlorinated hydrocarbon DNAPL is present at the 
127 
redundant waste sites and the southern effluent dams. No free phase DNAPL was 
visible or can be inferred from the results obtained for these areas.   
The DNAPL source zone at the redundant production and handling facilities in the 
weathered zone is inferred based on the multiple lines of evidence (Figure 3-38).  
Table 3-11: Calculated soil chlorinated hydrocarbon concentration threshold minimum and 
maximum values for the Investigation Site 
Chlorinated Hydrocarbon Minimum C Ti Maximum C
T
i
(mg/kg) (mg/kg)
Chlorinated Ethenes
PCE 480 1 600
TCE 1 500 6 000
Cis-1,2-DCE 6 000 19 500
trans-1,2-DCE 4 000 14 000
VC 6 000 16 500
Chlorinated Ethanes
1,1,1-TCA 1 400 4 500
1,1-DCA 4 000 12 500
1,1-DCA 2 000 6 500
1,1,2-TCA 6 000 21 000
1,2-DCA 8 000 25 500
Chlorinated Methanes
CCl4 950 3 000
CHCl3 6 000 19 500
Chlorobenzenes
1,2-DCB 1 200 4 500
1,3-DCB 400 1 400
1,4-DCB 650 2 500
ClBenz 2 500 9 000
128 
Figure 3-35: Soil total chlorinated hydrocarbons concentrations, spatial distribution and composition at the Investigation Site (results from ERM 2012)
129 
i. 1,2-DCA distribution
130 
ii. 1,1,2-TCA distribution
iii. PCE distribution
131 
iv. TCE distribution
v. CHCl3 Distribution
Figure 3-36: Chlorinated hydrocarbons distribution in the soils (in μg/kg) beneath the historical 
production, storage and handling area 1 
132 
N
133 
Figure 3-37: 3-D solid models of grouped chlorinated hydrocarbons soil concentrations 
(μg/kg) 
Where: 
Total chlorinated hydrocarbons = total chlorinated benzenes + total chlorinated ethenes + total chlorinated ethenes + 
total chlorinated ethanes + total chlorinated methanes 
Total Chlorinated Benzenes (ClBENZ) = 1,2-dichlorobenzene + 1,3-dichlorobenzene + 1,4-dichlorobenzene + 
chlorobenzene 
Total Chlorinated Ethenes (ClEthenes) = tetrachlorethene + trichloroethene + cis-1,2-dichloroethene + trans-1,2-
dichloroethene + 1,1-dichloroethene + vinyl chloride 
Total Chlorinated Ethanes (ClEthanes) =1,1,1-trichloroethane + 1,1,1,2-tetrachloroethane + 1,1,2,2-tetrachloroethane 
+ 1,1,2-trichloroethane + 1,1-dichloroethane +1,2-dichloroethane  
Total Chlorinated Methanes (ClMethanes) = carbon tetrachloride + chloroform 
134 
Figure 3-38: Plan view of inferred DNAPL source zones in soil
135 
3.3.4.2 Fractured (bedrock) source zones delineation  
3.3.4.2.1 Fracture network characteristics 
As mentioned in the section on the site’s local geology (Section 3.3.2.), the bedrock 
underlying the site is characterised by fractured networks. This section focuses on 
the characteristics of the fractured network as it is a significant factor in the 
chlorinated hydrocarbon DNAPL migration.  
The following tools were useful in characterising the fractures at the site: 
• Downhole borehole geophysical logging, particularly using the acoustic 
televiewer and the results obtained from the fracture pick analysis.  
• Detailed core borehole logging 
• Ribbon NAPL samplers 
Downhole borehole geophysical logging was restricted to core holes drilled outside 
the source zones. The results obtained were extremely useful in assisting with the 
understanding of the local geology and the extent of fracturing at the site (Figure 3-
39). The probes are however not compatible with use in open holes located within 
source zones, containing free phase DNAPL. The value of information deduced from 
detailed core logging was found to be sufficient for the requirement of this study, 
without the need for further downhole geophysical log profiling in holes with ribbon 
NAPL samplers installed.  
Transmissivity profiling using the ribbon NAPL samplers (FLUTeTM) was useful in 
determining the location of transmissive fractures.  
The dolerite sill is characterised predominantly by steeply dipping, sub-horizontal 
fractures (Figure 3-40). The stereonet and rose diagram shown below were created 
using RockWorks15.  Core holes drilled within the soils source zones were selected 
for the analysis. Examination of the stereonet in Figure 3-40 shows an 
underrepresentation of vertical fractures. This is the result of the drilling orientation as 
all core holes drilled at the site were vertical.  
136 
MLR01
Figure 3-39: Portion of downhole geophysical log for core hole MLR01 showing the large
density of fractures
(a) (b)
Figure 3-40: (a) Schmidt net (equal area, contoured) stereonet and (b) Rose diagram of
fractures within the source zones (uncorrected)
137 
Figure 3-41 provides 3-D strip logs showing fracture orientation of core holes drilled 
into the historical production and handling area footprint (identified in Section 3.3.4.1 
as the primary soil/weathered source zone).  
Figure 3-41: 3-D strip logs showing fracture discs of core holes drilled in the historical
production and handling facility
The variation in the fracture density per meter with elevation within the source zone is 
shown in Figure 3-42. While the density of fractures decreases within the 
unweathered portion of the dolerite sill, it still remains fractured with discrete fracture 
networks. The Karoo Ecca sedimentary rocks are characterised by sub-horizontal 
bedding planes. Sub-vertical fractures extend from the dolerite sill, through the low 
permeability contact zone between the dolerite and the Karoo Ecca sedimentary 
rocks.  
138 
1485 Moderately to 
1480 slightly weathered 
1475 dolerite 
1470
unweathered dolerite CBH42
1465
CBH185
1460
CBH186
1455 Karoo Ecca (bedding CBH187
1450 planes) CBH 196
1445 CBH 197
0 20 40 60 80 100 120
Fracture density/meter 
Figure 3-42: Variation of fracture density with elevation at the historical production and 
handling facility  
Figure 3-43 is a plot of the ribbon NAPL sampler descent velocity into PBH221 with 
depth. A change in the liner velocity occurs when a flow path is sealed. A tranmissive 
fracture/feature is determined through analysis of the descent velocity of the ribbon 
NAPL sampler with depth. Descent velocity decreases rapidly with depth of the 
borehole in the absence of transmissive fractures. An analysis of the data from the 
other boreholes that were installed with ribbon NAPL samplers shows an apparent 
lack of small transmissive fractures. This could be caused either by an absence of 
smaller transmissive fractures or through masking of the presence of other fractures 
due to the larger intake of groundwater by the highly transmissive fracture.  
Descent velocity (m3/m/s) PBH221
0.000 0.005 0.010 0.015 0.020 0.025 0.030
0
5
10 Transmissive 
feature 
15
20
25
30
35
40
Figure 3-43: Location of transmissive fracture in PBH221  
139 
Depth (z, m bgl) Elevation (mamsl) 
The smallest fracture along a flow path will control groundwater flow through a 
fracture in a network. The hydraulic aperture (viz. the value assigned to the smallest 
fracture along a flow path) for CBH42 was calculated using the cubic law (Snow, 
1968 as modified by Bear, 1993):  
 =      Equation 3-2 
Where: 
2b = hydraulic aperture  
μ = dynamic viscosity of water 
ρ = density of water  
g = acceleration due to gravity 
Tfr = fracture transmissivity 
N = number of fractures per interval 
The average groundwater temperature at the site is approximately 200C. Hence 
values for ρ and μ used were 0.998 g/cm3 and 1.003 mPa·s (CRC Handbook of 
Chemistry and Physics, 1986).  The values used for N and Tfr are indicated in Figure 
3-44 below. 
N = 89, Tfr = 1.6 m2/day
N = 5, Tfr = 1.4 m
2/day
Figure 3-44: 3-D strip log of CBH42 showing the intervals and values used for calculating 
hydraulic aperture 
Hydraulic aperture (2b) for CBH42 is calculated to be 63.5 μm at depths of between 9 
– 12 m bgl and 159 μm at depths of between 12 and 15 m bgl.
140 
3.3.4.2.2 DNAPL source delineation  
Historical site data indicated evidence through direct observation of pooled 
chlorinated hydrocarbon DNAPL at two locations at the Investigation Site viz. SRK 14 
S/D and EDC 01 (Figure 3-4).  Neither the transport mechanics nor the transport 
media could however be discerned through the available information.  
Source zone delineation was undertaken through further drilling and sampling, 
utilising both in situ and ex situ tools. Residual and pooled DNAPL was found in 
additional core holes and percussion holes. A summary of the boreholes containing 
pooled or residual DNAPL and the method of detection is provided in Table 3-12. The 
location of boreholes positive for pooled or residual DNAPL is provided in Figure 3-
48.  From this figure it is observed that DNAPL occurrence in the fractured bedrock is 
associated with the redundant production and handling facilities in the north western 
part of the site. The southern effluent dams are not considered release areas for free 
phase DNAPL, but may be associated with the high concentrations of dissolved 
chlorinated hydrocarbon DNAPL in boreholes. 
Three on-site lines of evidence were useful in the detection of pooled DNAPL: 
1. Anomalous PID measurements (> 10 ppm) made on the drill chips, 
2. Odours emanating from drill chips, and 
3. Visual observation of pooled DNAPL through bailer testing of the borehole. 
The depth of visual observation of pooled DNAPL ranged from 20 m bgl – 47 m bgl. 
Pooled DNAPL was intersected at approximately 20 m bgl beneath the historical 1,2-
DCA storage tanks, located west of the historical production facilities (Figure 3-48). 
The pooled DNAPL is associated with the dolerite in this area. Information provided 
by plant personnel indicate a loss of approximately 5 L/day of 1,2-DCA from these 
tanks over a 31 year period. This occurred during routine sample collection to test the 
quality of the product for production purposes. This excludes any additional spills or 
leaks from these tanks during this period.  DNAPL pools in boreholes are found away 
from the original release point. Distances of up to 850 m are recorded for boreholes 
showing pooled DNAPL (Figure 3-48). Boreholes with pooled DNAPL showed 
anomalous PID readings of > 10 ppm.  
Apparent thickness of DNAPL was measured in the boreholes over a 1 month period. 
The change in apparent thickness over time is shown in Figure 3-45. In general, 
DNAPL flow into the boreholes stabilised after 1 week. DNAPL was only measured in 
CBH174 after 1 week, indicating the mobilisation of DNAPL along an interconnected 
141 
horizontal fracture. DNAPL thickness in the boreholes did not decrease during the 
measuring period, indicating a large mass of free phase located  
4
3.5 PBH 210
PBH 2
3 PBH 38
2.5 PBH 57
PBH 58
2
PBH 64
1.5 CBH 174
1 PBH 180
PBH 182
0.5
PBH 209
0 PBH 211
1 10 100 1000 10000 PBH 289
Time (hours) 
 Figure 3-45: Apparent DNAPL thickness measured in boreholes 
Ribbon NAPL samplers containing a hydrophobic dye were installed in the boreholes 
shown in Table 3-13. Residual DNAPL was detected in core hole CBH42 and 
percussion borehole PBH213 through the use of ribbon NAPL samplers.  Both these 
boreholes are located within the footprint of the historical production and storage 
facilities. Residual DNAPL thicknesses ranged from 1 to 50 cm in CBH42 and 
PBH213. From the log of CBH42, it is observed that the location of residual DNAPL is 
related to the fractures in the upper portion of the dolerite sill (Figure 3-46).   
PID measurements taken in CBH42 and PBH213 were < 10 ppm, indicating that the 
effect of volatilization from the retrieved core or drill bits may produce false negatives 
in zones where DNAPL (pooled or residual) may be present and can thus not be used 
as a solitary tool to determine the presence of DNAPL on a site. 
Residual DNAPL was also detected in CBH42, CBH186, CBH196 and CBH197 
through the use of Oil Red O®. Figure 3-47 is a photograph of the staining observed 
in core hole CBH42 at a depth of 22.7 m bgl. The ribbon NAPL sampler was installed 
in this borehole to a depth of 18 m bgl and hence this zone of residual DNAPL was 
not detected. This is a limitation of the technology in fractured bedrock with low 
transmissivities.    
142 
Apparent thickness of DNAPL (m) 
Bedrock cross sections are provided below (Figure 3-49 -Figure 3-53). The cross 
sections show the association of anomalous PID measurements and odours in 
relation to the stratigraphy. Pooling of DNAPL at the site has occurred above the 
following lithologies: siltstone and mudstone (SST10) of the Karoo Ecca, the contact 
zone between the dolerite sill and the Karoo Ecca sandstone (SST10); and the 
carbonaceous shale (SH10).   
Table 3-12: Method of detection of residual and pooled DNAPL within boreholes  
Borehole ID Method(s) of detection
EDC01 Visual observation of pooled DNAPL (bailer test), odours
SRK14 S/D Visual observation of pooled DNAPL (bailer test), odours
Residual DNAPL staining on ribbon NAPL liner,
CBH42 Positive result using hydrophobic dye indicating residual DNAPL 
PBH213 Residual DNAPL staining on ribbon NAPL liner
Visual observation (bailer test) @ 30 m bgl,
PBH2 Anomalous  PID measurements, odours 
Visual observation (bailer test) @ 20 m bgl,
PBH38 Anomalous PID measurements, odours 
Visual observation (bailer test) @ 45 m bgl,
PBH57 Anomalous PID measurements, odours 
CBH174 Visual observations (bail test) @ 30 m bgl
Visual observation (bailer test) @ 37 m bgl,
PBH64 Anomalous PID measurements, odours 
Visual observation (bailer test) @ 20 m bgl,
PBH180 Anomalous PID measurements, odours 
Visual observation (bailer test) @ 25 m bgl,
PBH182 Anomalous PID measurements, odours 
Visual observation (bailer test) @ 20 m bgl,
PBH209 Anomalous PID measurements 
Visual observation (bailer test) @ 20 m bgl,
PBH210 Anomalous PID measurements 
Visual observation (bailer test) @ 20 m bgl,
PBH211 Anomalous PID measurements 
Visual observation (bailer test) @ 37 m bgl,
PBH289 Anomalous PID measurements, odours 
CBH186 Positive result using hydrophobic dye indicating residual DNAPL
CBH196 Positive result using hydrophobic dye indicating residual DNAPL
BH197 Positive result using hydrophobic dye indicating residual DNAPL
143 
Table 3-13: Summary of maximum depth of ribbon NAPL sampler installation 
Borehole ID Maximum Depth (m bgl) of ribbon NAPL sampler installation
CBH42 18.23
PBH213 24.38
PBH228 32
PBH221 36.58
Figure 3-46: Logs of CBH42 and PBH213 showing the location of residual DNAPL (shown in 
blue as determined through the use of ribbon NAPL samplers and in red as observed using 
Oil Red O®) in relation to the litho-stratigraphy 
144 
Figure 3-47: Oil Red O®  staining at 22.7 m bgl in core hole CBH42 indicating the presence of 
residual DNAPL 
In addition to direct observation of DNAPL (described above), chlorinated 
hydrocarbon DNAPL is also inferred at boreholes where the sum of the mole fractions 
of the chlorinated hydrocarbon DNAPL in the groundwater sample exceeds 1% of the 
total effective solubility. The 1% effective solubility threshold is calculated using the 
following equation: 
∑ =  =      Equation 3-3 
Where: 
Ciobs = sampled groundwater concentration in mg/l of component i 
Si = single-component solubility of component i 
α = cumulative mole fraction of sample 
n = number of components in the groundwater sample. 
A summary of the results for boreholes where chlorinated hydrocarbon DNAPL can 
be inferred is provided in Figure 3-48.  
The DNAPL aqueous phase plume migration results are discussed further in Section 
3.3.5.  
DNAPL chemical composition (in mole fraction) for selected samples taken from the 
site is provided in Table 3-14. The spatial distribution of the sampled boreholes and 
145 
their chemical compositions are shown in Figure 3-54. The following observations are 
made: 
• As expected, the DNAPL compositions within the release area mimic the 
chlorinated hydrocarbon that was produced, stored or handled at the facility. 
• The chemical composition of DNAPL found 600 – 800 m down gradient of the 
above ground storage tanks (boreholes PBH64, PBH289, EDC01 and CBH259) is 
primarily 1,2-DCA which is the same as the composition of DNAPL pooled below 
the above ground storage tanks and inferred/pooled at the historical 1,2-DCA 
production facility. Small percentages of other associated chlorinated 
hydrocarbons such as 1,1,2-TCA is also found in these boreholes. 
• DNAPL with chemical compositions of 1,1-DCA, CCl4 and CHCl3 are also found 
downgradient (600 – 800 m) of the original release point.  
Neither free phase nor residual phase has migrated further than the site boundary. 
The length and movement of the dissolved phase plume is discussed in detail in the 
next section. Movement of the DNAPL has been primarily along the contact zone 
between the dolerite sill and the Karoo Ecca sedimentary rocks as well as along 
bedding planes located in the Karoo Ecca sedimentary rocks. However, close 
inspection of the cross sections (such as cross section G-H) also indicate vertical 
upward migration. Additionally, based on the observations made above on the free 
phase DNAPL compositions found in downgradient boreholes it can be inferred that 
two contiguous DNAPL pools (one originating from the historical 1,2-DCA storage 
facility and the other from the historical production facilities) have migrated towards 
the southern boundary of the site. 
Physical property results for EDC01, PBH57 and PBH180 are also provided in Table 
3-14. A plot of the rheological properties of the site samples is shown in Figure 3-55. 
The samples display analogous physical properties. As the samples are composed 
primarily of 1,2-DCA it would have been expected that the rheological properties 
would be the same. However, as is seen from Figure 3-55, the analysed samples 
have higher viscosities compared to pure phase 1,2-DCA. PBH57 also displays a 
higher density value (approximately 5% higher) compared to 1,2-DCA and the other 
samples analysed (PBH 180 and EDC01). This is attributed to the variances in 
composition of PBH 57 compared to the other samples.  
146 

Figure 3-49: Location of bedrock source zone cross sections
148 
Figure 3-50: Bedrock cross section A-B (modified from ERM, 2012)
Figure 3-51: Bedrock cross section C-D (modified from ERM, 2012)
149 
Figure 3-52: Bedrock cross section E-F (modified from ERM, 2012)
 
Figure 3-53: Bedrock cross section G-H (modified from ERM, 2012)
150 
Table 3-14: Summary of mole fractions (%) composition of groundwater samples with inferred chlorinated hydrocarbon DNAPL (results  
from ERM, 2012) 
Borehole Trans 1,1,1- 1,1,2-
ID PCE TCE 1,1-DCE 1,2-DCE VC TCA TCA 1,1-DCA 1,2-DCA CCl4 CHCl3
PBH52 75.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 11.5 13.1
PBH54 15.5 1.6 0.0 0.0 0.0 0.0 0.0 0.0 12.5 8.1 62.3
PBH63 2.7 4.5 0.0 0.0 0.0 0.0 1.8 0.0 41.7 8.9 40.3
PBH223 68.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5.2 8.1 18.2
PBH58 0.0 0.0 0.0 0.0 0.0 0.0 2.9 0.0 97.1 0.0 0.0
PBH66 41.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.8 23.4 32.9
PBH8 16.8 0.0 0.0 0.0 0.0 1.6 0.0 0.0 3.8 0.0 77.8
PBH119 38.4 1.0 0.0 3.6 0.0 0.0 0.0 5.1 0.0 45.1 0.0
PBH54 19.7 1.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 61.9 16.7
CBH63 2.5 0.0 0.0 0.0 0.0 0.0 2.3 1.8 86.2 3.6 3.7
CBH185 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 99.5 0.0 0.0
CBH186 0.0 0.0 2.0 0.0 9.4 0.0 0.0 0.0 87.1 0.6 0.0
CBH187 26.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.7 23.0 48.5
CBH196 27.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.1 70.2
CBH197 36.7 0.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 36.6 25.5
CBH259 0.0 0.0 1.5 1.4 0.0 0.0 0.0 0.0 96.3 0.0 0.0
151 
Table 3-15: Physical and chemical properties of pooled DNAPL samples from the Investigation Site 
Borehole Physical Properties Chemical Composition (mole fraction %)
ID Water
Content Interfacial Viscosity @ Viscosity @
Density (% Tension 150C 220C 1,1- 1,1,2- 1,2- 1,1-
(g/cm3) weight) (mN/m) (mPa∙s) (mPa∙s) PCE DCE VC TCA DCA DCA CCl4 CHCl3
EDC01 1.2511 0.1497 22.31 1.03 0.98 0.0 3.5 0.5 3.0 92.2 0.4 0.0 0.2
PBH57 1.3159 0.0875 23.94 1.04 0.95 5.5 0.6 0.4 8.0 77.3 0.6 7.0 0.2
PBH180 1.2582 0.0985 24.10 1.06 0.96 0.0 0.3 0.3 0.6 98.6 0.1 0.0 0.0
PBH2 13.4 0.1 0.1 0.9 61.1 0.6 20.8 2.8
PBH64 0.4 1.2 0.7 3.3 93.4 0.4 0.0 0.0
PBH182 3.5 0.3 0.2 2.3 89.2 1.7 2.3 0.5
PBH289 Not Analysed 0.7 2.5 0.6 2.4 92.8 0.4 0.0 0.5
152 
Figure 3-54: DNAPL composition in boreholes (Results from ERM, 2012) 
153 
1.7
PCE CCl4 
1.5 CCl3 
1,1,2-TCA TCE 
1,1,1-TCA 1,2-DCB EDC01
1.3 PBH57
1,1-DCE 1,2-DCA 1,4-DCB PBH180
1.1 ChlBenz 
0.9 Water 
0.7
0.5
0 0.2 0.4 0.6 0.8 1 1.2 1.4
viscosity (mPa·s) 
Figure 3-55: Plot of density vs viscosity for samples analysed from the Investigation Site 
3.3.5 Hydrogeochemical characteristics 
Groundwater chemistry data were subdivided to reflect (1) historical data (and 
associated trends) and (2) current data (which includes the results obtained from 
newly drilled boreholes). A comparison is made of the evolving plume with time. 
Production and storage of chlorinated hydrocarbon ceased at the site in 1998. 
Groundwater monitoring results are available from 1998 to present. The historical 
data obtained is sporadic. Groundwater samples were taken from all boreholes 
(overburden as well as bedrock wells) in 2012 and this provides the most recent 
development of the chlorinated hydrocarbons plumes.  
SRK14S/D is located within the source zone at the historical 1,2-DCA above-ground 
storage tanks. Figure 3-56 shows the chlorinated ethenes, ethanes and methanes 
time-series trends.  Both piezometers are located in the dolerite aquifer, with the 
shallow piezometer recording data in the weathered zone and the deep piezometer 
recording data in the fractured bedrock. As expected, similar trends and 
concentrations are observed within the weathered and the fractured zone.  All graphs 
show a general decrease in concentrations followed by an increase in July 2009 and 
a subsequent decrease in concentrations to the present date.  The presence of 
daughter products such as cis-1,2-DCE indicates that biodegradation is occurring at 
the site.  
Figure 3-57 - Figure 3-58 show the changes in the dissolved phase plumes, 
represented as isoconcentration filled contours, in bedrock over time for the sum of 
154 
density (g/cm3) 
the chlorinated ethenes and the sum of the chlorinated ethanes respectively. An 
increased number of boreholes were drilled over the years and this is reflected by the 
shape of the plumes observed, with better refinement in the interpolation with an 
increased number of datasets over the years. The 1998 chlorinated ethanes 
concentration data shows two distinct plumes, one originating from the historical 
chlorinated hydrocarbons production and handling facilities (located in the north 
western part of the site); the other plume originating from the redundant hazardous 
waste site 2 located in eastern part of the site. In 2002, there is a reduction of the 
plumes sizes. Inspection of the 2012 concentration contours show that two 
chlorinated ethanes plumes start emerging from the redundant hazardous waste site 
1, while the chlorinated ethanes plume from the redundant hazardous waste site 2 is 
no longer present.  The isoconcentration filled contour plots for the chlorinated 
ethanes show similarities and differences compared to the chlorinated ethenes plots. 
Four plumes are visible in the 1998 plot, one originating from the historical 
chlorinated hydrocarbons production and handling facilities/storage tanks (located in 
the western part of the site); the second plume originating approximately 500 m 
downgradient of historical production facilities/storage tanks, a third originating from 
the redundant hazardous waste site 2 located in eastern part of the site; and a fourth 
originating from the redundant hazardous waste site 1. The chlorinated ethanes plot 
for 2002 is significantly different. In this case only one plume originating from the 
historical production facilities/storage tanks is visible. The 2012 chlorinated ethanes 
plot shows a single continuous plot from the historical production facility/storage tank. 
The anomalous data obtained for the 2002 chlorinated ethanes plot may be a result 
of sampling and/or analytical errors.  
The total chlorinated hydrocarbons isoconcentration plots are provided in Figure 3-
59. As discussed in the previous section, no free-phase DNAPL was evident at the 
redundant hazardous waste site 2. The groundwater plume originating from this 
source has likely resulted from residual DNAPL sources. The total chlorinated 
groundwater plume lengths are shorter than what would be expected for the volume 
of pooled and residual DNAPL at the site. A number of reasons (or combinations 
thereof) could be the cause of this. Fracture flow dominates movement of 
groundwater and contaminant at the site. It is possible that low interconnectivity of 
fractures together with low transmissivity has led to smaller than expected 
groundwater plumes. 
155 
Figure 3-57 to Figure 3-59 indicate a persistent dissolved phase plume emanating 
from the source zones. The persistence of the plumes resulting from the redundant 
hazardous waste sites may be attributed to current releases through the leakages of 
drums. However, this is not the case in the source zones associated with the 
historical handling and production facilities. The time required for dissolved and 
adsorbed contaminants to be flushed from the source zone was calculated assuming 
unidirectional and steady-state flow conditions which are subject to advection and 
sorption (dispersion was not considered in this calculation). The following equation 
was used (modified from Kueper and Davies, 2009) in the calculations: 
 =     Equation 3-4 
Where: 
t = the time required for contaminants to migrate through the source zone of length L 
in the direction of groundwater flow 
v = average linear groundwater velocity  
L = length of source zone in the direction of the flow 
Rf = retardation factor for the contaminant of interest 
The average retardation factor for each compound was calculated using Monte Carlo 
simulations (Appendix F). Results obtained are provided in Table 3-16 and indicate 
the presence of persistent plumes in the weathered zone. 
Table 3-16: Calculated length of time required for the dissolved phase plume to detach from 
the weathered source zone 
Chlorinated Hydrocarbon Rf (Avg) L (m) v (m/yr) t(yr)
PCE 79 90 0.43 16534.88
TCE 48 90 0.43 10046.51
Cis-1,2-DCE 34 90 0.43 7116.279
Trans-1,2-DCE 33 90 0.43 6906.977
1,1-DCE 26 90 0.43 5441.86
VC 18 90 0.43 3767.442
1,1,1-TCA 35 90 0.43 7325.581
1,1-DCA 25 90 0.43 5232.558
1,1,2,2-TetCA 77 90 0.43 16116.28
1,1,2-TCA 48 90 0.43 10046.51
1,2-DCA 33 90 0.43 6906.977
CCl4 38 90 0.43 7953.488
CHCl3 27 90 0.43 5651.163
Table 3-17 provides the summary of pore water analysis of samples taken from 
stream piezometers. Only four (4) piezometers showed detectable trace values of 
dissolved chlorinated hydrocarbon. Analysis of chlorinated ethenes and methanes 
were below detection limit in all samples. Trace concentrations of 1,2-DCA was 
156 
present in   the piezometer samples shown below. The presence of 1,2-DCA in pore 
water beneath the streams indicates that dissolved phase 1,2-DCA discharges to the 
southern intermittent stream. This is consistent with the measurement of upward 
vertical gradients obtained from the piezometer data. 
Table 3-17: Pore water results for stream piezometer samples 
1,2-DCA
Piezometer ID (μg/L)
SMPA14 2.1
SMPA15 1.5
SMPA16 2.5
SMPA18 0.74
In addition to a hydrophobic liner installed in CBH42, a felt activated carbon 
technology (FACT) was concurrently installed in the hole in order to test the 
technology as a method of obtaining depth discrete dissolved phase chlorinated 
hydrocarbon pore water concentrations. The complete set of results obtained is 
provided in Appendix G. The results are semi-quantitative and are measured in μg/g 
of activated carbon. Figure 3-60 shows the comparison of the results obtained from 
the FACT liner to the locations of DNAPL found using the ribbon NAPL sampler. Two 
(2) peaks in concentrations are seen in the graph of concentration vs depth, which 
correspond to the locations of the transmissive fractures identified through installation 
of the ribbon NAPL sampler.   
The biogeochemical characteristics of the free phase DNAPL and the aqueous phase 
DNAPL are provided in the next chapter. Plots of time-series graphs of the 
compounds in the PCE and 1,1,2,-TetCa sequential biodegradation sequence along 
transacts was inconclusive. Generally, all compounds have high concentrations in the 
source zone with no distinct patterns. Concentrations of daughter products outside 
the source zone are negligible. An additional complication is that a vast number of 
compounds where used or produced (either as final product or as intermediates) at 
the site. This makes it extremely difficult to determine any sequential patterns using 
concentration data alone. 
157 
100000 10000000 10000000
Tetrachloroethene 1,1-Dichloroethane
1,1-Dichloroethene 1000000 1,2-Dichloroethane 1000000
10000 trans-1,2-Dichloroethene 1,1,2-Trichloroethane
100000 100000
cis-1,2-Dichloroethene
1000 Vinyl chloride 10000 10000 Dichloromethane
Bromochloromethane
1000 1000
100 Chloroform
100 100 Carbon tetrachloride
10 10 10
SRK14S Chlorinated Ethenes SRK14S Chlorinated Methenes
SRK14S Chlorinated Ethenes 11
1 Oct-95 Jul-98 Apr-01 Jan-04 Oct-06 Jul-09 Apr-12Oct-95 Jul-98 Apr-01 Jan-04 Oct-06 Jul-09 Apr-12
Oct-95 Jul-98 Apr-01 Jan-04 Oct-06 Jul-09 Apr-12 TimeTime
Time
10000000 10000000 1000000
Tetrachloroethene 1,1-Dichloroethane Dichloromethane
1000000 1000000 1,2-Dichloroethane1,1-Dichloroethene 100000 Bromochloromethane
1,1,2-Trichloroethane Chloroform
100000 trans-1,2- 100000
Dichloroethene 10000 Carbon tetrachloride
cis-1,2-
10000 10000Dichloroethene 1000
Vinyl chloride
1000 1000
100
100 100
10
10
10
SRK14S Chlorinated Ethanes SRK14D Chlorinated Methanes
SRK14D Chlorinated Ethenes 1 1
1 Oct-95 Jul-98 Apr-01 Jan-04 Oct-06 Jul-09 Apr-12 Oct-95 Jul-98 Apr-01 Jan-04 Oct-06 Jul-09 Apr-12
Oct-95 Jul-98 Apr-01 Jan-04 Oct-06 Jul-09 Apr-12
Time Time
Time
Figure 3-56: Variation in chlorinated ethenes, ethanes and methanes concentrations in SRK14S/D 
158 
CCoonncceennttrraattiioonn ((μμgg//LL)) CCoonncceennttrraattiioonn ((μμgg//LL))
CCoonncceennttrraattiioonn ((μμgg//LL)) CCoonncceennttrraattiioonn ((μμgg//LL))
CCoonncceennttrraattiioonn ((μμgg//LL)) CCoonncceennttrraattiioonn ((μμgg//LL))
a
b
c
Figure 3-57: Total chlorinated ethenes plume (a) 1998, (b) 2002 and (c) 2012 
159 
a
b
c
Figure 3-58: Total chlorinated ethanes plume (a) 1998, (b) 2002 and (c) 2012 
160 
Figure 3-59: Total chlorinated hydrocarbon DNAPLs isoconcentrations filled contours plot (boreholes located in unweathered dolerite) 
 
161 
Total Chlorinates Hydrocarbon
(ug/g Activated Carbon)
1 1000 1000000
0
2
4
6
8
10
12
14
16
18
Figure 3-60: Semi-quantitative depth discrete total dissolved hydrocarbon concentrations 
3.3.6 Biogeochemical characteristics 
3.3.6.1 Biodegradability of chlorinated solvent DNAPL 
The modelled results for the individual components of the DNAPL samples (Table 3-
15) using EpiWeb v4.0.BioWIN are provided in Appendix H. The results indicate that 
the DNAPL at the site is not readily biodegradable.  
3.3.6.2 Toxicity profile of the groundwater downgradient of source zones 
The BOD, COD and DO results for groundwater samples from selected boreholes at 
the site are provided in Table 3-18. The biodegradability indices of the selected 
boreholes (located downgradient of potential source zones) indicate that in general 
biodegradation is not likely to occur at the site. The BOD:COD ratio  for CAP01S 
indicate that there is potential for biodegradation at this sample point. This borehole is 
located outside of any DNAPL source zones or plumes. 
162 
DDeepptthh ((mm bbggll))
Table 3-18: Chemical oxygen demand, biological oxygen demand and dissolved oxygen 
results 
Borehole COD BOD BOD:COD Biodegradability DO (mg/L)
ID (mg/L) (mg/L) Index Descriptor
SRK02S 159 15 0.09 Not biodegradable 0.1
SRK02D 15 5 0.33 May be 0.1
biodegradable 
MLR01S 68 10 0.14 Not biodegradable 0.1
SRK34S 192 20 0.1 Not biodegradable 0.1
SRK12D 140 10 0.07 biodegradable 0.1
SRK03S 184 10 0.05 biodegradable <0.1
SRK03D 184 20 0.1 Not biodegradable 0.1
SRK21D 121 20 0.16 Not biodegradable 0.1
CAP01S 13 13 1 biodegradable 0.1
SRK15S 179 30 0.17 Not biodegradable 0.1
SRK15D 181 10 0.05 Not biodegradable 0.1
The percentage inhibition can be interpreted as the toxicity profile for the site. In 
general, all the results (Table 3-19) indicate that the groundwater located 
downgradient of source zones are fairly toxic, a factor that must be considered in risk 
assessments. While this is a rapid, on-site test for toxicity, it does not provide any 
information on the cause of toxicity. This is a limitation for sites with a large variety of 
contaminants that can contribute to the groundwater toxicity. 
Table 3-19: Toxicity screening of groundwater 
Borehole ID % Inhibition Location Descriptor
SRK02S 54.3 Downgradient of redundant waste site 2
SRK02D 49.2 Downgradient of redundant waste site 2
MLR01S 41.4 Located east of historical production and handling 
facilities 
SRK34S 74 Downgradient of redundant waste site 1
SRK12D 27.3 North of southern effluent dams
SRK03S 31 Southern boundary of site
SRK03D 42.1 Southern boundary of site
SRK21D 65 Downgradient of historical production and handling 
facilities 
CAP01S 19.4 Downgradient of southern effluent dams
SRK15S 71.2 Downgradient of redundant waste site 1
SRK15D 84.2 Downgradient of redundant waste site 1
163 
3.3.6.3 Preliminary screening for PCE degradation using DNA-based techniques 
Field measurements collected from the boreholes that were screened for 
biodegradation are provided in Table 3-20. Redox measurements indicate reducing 
conditions. Samples were taken during winter, hence the moderate groundwater 
temperatures recorded. 
Table 3-20: Field physical measurements 
Borehole Temperature (oC) pH Conductivity ORP (MV)
ID (mS/cm)
MLR01S 18.3 7.59 6.72 4
SRK12D 15.5 7.27 1.20 31
SRK21D 18.3 7.12 9.43 12
SRK34S 19.4 7.37 2.62 20
SRK02S 18.4 6.51 13.8 -65
SRK02D 20.0 6.71 10.55 -29
Visible DNA was extracted from samples obtained from MLR01S, SRK34S and 
SRK02S (Figure 3-61a). All other samples were below detection limits. Subsequent 
addition of enzyme stabilisation additives allowed for further amplification of the DNA 
(shown in Figure 3-61b) shows visible DNA in all boreholes apart from SRK21D 
(located within the primary source zone). DNA concentration results and the purity of 
the genomic DNA extracts are provided in Table 3-21. The results indicate a 
correlation between DNA amplification and DNA concentration, with higher DNA 
extraction yields resulting in higher amplifications.    
As this was only a screening method, no targeted PCE degrading bacteria were 
isolated from the samples. However, the presence of bacteria is evident in the 
samples, apart from the sample taken from SRK21D. The absence of bacteria from 
SRK21D is likely related to the high levels of contaminants in this sample, indicating 
that the technique might not be viable within source zones.  
164 
a b
Figure 3-61: Genomic DNA extracted from selected borehole samples (a) pre-optimisation
and (b) post-optimisation. Lane M, MassRulerTM DNA Ladder (Fermentas); lane 1, MLR01S,
lane 2, SRK02D; lane 3, SRK12D; lane 4, SRK34S; lane 5, SRK02S, lane 6, SRK21D
Table 3-21: Summary of DNA concentration and purity of genomic DNA extracts
Borehole ID Concentration (ng/μL) A260/280
MLR01S 28.53 1.63 
SRK12D 15.91 1.44 
SRK21D 21.83 1.55 
SRK34S 27.77 1.74 
SRK02S 23.21 1.55 
SRK02D 17.43 1.8 
3.3.6.4 Compound specific isotope analytical results 
CSIA (δ13C and δ37Cl) was undertaken in 2008 and 2009 from existing boreholes at 
the site. The results obtained are summarised in Table 3-22. Samples were 
submitted from boreholes along downgradient transects of potential source zones (as 
was thought in 2008/2009) and included the historical production and handling 
facilities and the redundant waste sites.  
SRK 21D, SRK 12 D and CAP01S are located on a transact with increasing distance 
from the primary source area. PCE and its daughter products were present in all 
boreholes in measurable concentrations. However, apart from TCE, the other 
daughter products (1,2-DCE and vinyl chloride) concentrations were too low to detect 
using CSIA. A further complication in interpreting the data is that there may have 
165 
been convergence of degradation pathways as parent products such as 1,1,2,2-TCA 
and PCE could have produced daughter products.  
Figure 3-62 and Figure 3-63 show the plot of δ13C versus the natural logarithm of the 
groundwater concentrations for PCE and TCE respectively measured during the 
same time period (2009). Also shown in the plot are the ranges for pure phase 
products, based on literature values provided in Section 2.1.2.2. SRK02S and SRK02 
D display the same range in δ13C as that expected for pure product PCE or TCE. This 
indicates very little transformation (biotic or abiotic) of the groundwater at this point. 
SRK02S/D is located immediately downgradient of the redundant hazardous waste 
site 2.  The variation of the other samples from the expected ranges in δ13C for pure 
phase product indicates transformation from the original source. The more positive 
δ13C indicates enrichment of the samples compared to the source, indicating that 
degradation is occurring at the site. The variations of the samples in Figure 3-62 and 
Figure 3-63  from the Rayleigh correlation (shown as a straight line) indicate mixing 
with different contaminant sources. As different batches of pure product were 
manufactured at the site, this is not an unexpected result as different batches of pure 
product may have different isotopic signatures.  Due to the small data set it is not 
possible to definitively state whether the field data show a Rayleigh correlation or not 
and hence whether Equation 2-8 can be applied.   
The rapid decrease in chlorinated solvents concentrations from the source area, 
along the groundwater pathways to the southern intermittent stream (Figure 3-59) 
indicate that some biotic degradation may be a factor within source area (evidenced 
from the presence of daughter products that are known not to have been used or 
produced on site such as cis- and trans-1,2-DCE) while abiotic degradation becomes 
the significant contaminant reduction process within the dissolved phase plume.  
166 
3.5 4.5 5.5 6.5 7.5 8.5 9.5
0
-5
-10
y = -2.0352x - 9.3465 
-15 R² = 0.2036 
SRK34S SRK21D
-20 SRK12D
-25
-30 Expected range for pure SRK02D SRK02S
product PCE 
-35
-40
-45
Ln (PCE) ug/L 
Figure 3-62: δ13C vs natural logarithm of PCE concentration in boreholes
3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 11.5
0
-5
-10 y = 3.1818x - 45.087 
-15 R² = 0.5912 SRK34S SRK21D
-20
-25
Expected range of pure product TCE
-30
SRK02D
-35 SRK02S
-40
-45
Ln (TCE) ug/L 
Figure 3-63: δ13C vs natural logarithm of PCE concentration in boreholes
167 
δ13C (0/00) δ13C (0/00) 
Table 3-22: Summary of compound specific isotope results 
δ13CVPDB (0/00) δ37ClSMOC
c is -1,2-
1,1-DCE cis-1,2-DCE PCE T CE 1,2-DCA 1,1,2-T CA CHCl3 CCl4 PCE T CE DCE
Bo re ho le
ID Aug -08 Ap r-09 Aug -08 Ap r-09 Aug -08 Ap r-09 Aug -08 Ap r-09 Aug -08 Ap r-09 Aug -08 Ap r-09 Aug -08 Ap r-09 Aug -08 Ap r-09 Aug -08 Aug -08 Aug -08
SRK12D BDL BDL BDL BDL -18.84 -21.11 BDL BDL BDL BDL BDL BDL -36.56 -25.52 -26.63 -32.54 -0.81 BDL BDL
SRK21D -22.44 -15.49 BDL BDL -19.02 -18 -20.17 -14.42 0.26 -5.31 -12.81 -8.56 -28.56 -14.52 -25.52 BDL 4.45 -0.41 10.91
SRK34S BDL BDL BDL -24.2 -17.74 -18.51 -21.23 -16.22 11.41 13.05 BDL BDL -18.67 -21.45 BDL BDL 4.00 8.33 9.05
SRK02S NA BDL NA -29.49 NA -29.77 NA -33.7 NA 7.63 NA BDL NA -30.55 NA BDL NA NA NA
SRK02D NA BDL NA -30.57 NA -30.5 NA -33.33 NA 8.47 NA BDL NA -31.12 NA BDL NA NA NA
CAP01S BDL BDL BDL BDL BDL BDL BDL BDL -5.92 8.88 BDL BDL BDL BDL BDL BDL BDL BDL BDL
SRK03D BDL BDL BDL BDL BDL BDL BDL BDL -8.47 -7.15 BDL BDL BDL BDL BDL BDL BDL BDL BDL
MLR01S BDL BDL BDL BDL -22.06 -27.73 BDL BDL BDL BDL BDL -17.84 -52.25 BDL -42.10 -40.72 BDL BDL BDL
SRK15S NA BDL NA BDL NA -26.7 NA BDL NA BDL NA BDL NA BDL NA BDL NA NA NA
SRK15D NA BDL NA BDL NA -28.5 NA BDL NA BDL NA BDL NA BDL NA BDL NA NA NA
168 
3.4 Conceptual Site Model 
The conceptual site models of the site are provided as simplified 2-dimensional 
interpretations of the DNAPL sources, the affected and influencing geological media, 
fate and transport mechanisms and potential exposure pathways. These are shown 
in Figure 3-64, which is the plan view conceptual site model; and Figure 3-65, which 
is the schematic cross sectional conceptual site model for the western perimeter of 
the site. 
A mixture of chlorinated hydrocarbon DNAPLs are found at the Investigation Site in 
vapour, adsorbed, DNAPL and aqueous phases in the source zones and as vapour, 
adsorbed and aqueous phases in the dissolved plume. The distribution and 
concentrations of the chlorinated hydrocarbons at the site varies from one locality to 
the next and is dependent on the following: 
1. DNAPL source 
o Location of release area 
o Volume of release 
o Nature of release 
o Physical properties of DNAPL(s) 
2. Geological heterogeneities 
o Bulk density 
o Primary porosity 
o Permeability 
o Depth and extent of weathering 
o Joint/fracture intensity 
o Fracture orientation 
o Fracture interconnectivity 
3. Hydrogeology 
o Fracture transmissivity 
o Groundwater flow direction 
o Hydraulic gradient 
4. Biotic and abiotic degradation processes 
This research has allowed an updated understanding of the release areas at the site. 
This is summarised in Table 3-23.  Distinct vapour phase plumes are associated with 
the adsorbed DNAPL in the overburden. Smaller concentrations of vapour are 
169 
however found throughout the site indicating the partitioning of dissolved phase 
chlorinated hydrocarbon DNAPL in shallow groundwater into the unsaturated zone.  
Table 3-23: Release areas and nature of release for the Investigation Site 
Release Area Nature of Release
Historical 1,2-DCA above Regular release over a long time-frame
ground storage tanks area 
Historical production and Intermittent spills over a large area for a long time-frame
handling facilities 2 
Redundant hazardous Once-off release of DNAPL (likely as a result of a drum 
waste sites (1 and 2) leak or spills during off-loading) in a localised area 
Sampling of the weathered profile indicated that the primary source zone in the 
weathered profile is located at the historical production and handling facilities (2) and 
sporadically around the redundant hazardous waste sites. Adsorbed and free phase 
DNAPL is found in the weathered profile.   
Residual and free phase pooled DNAPL is found in the fractured bedrock in the 
following areas:  
1. Below and continuously for a distance of approximately 850 m downgradient of 
the historical 1,2-DCA above ground storage tanks and 1,2-DCA production 
facilities. The signature of the DNAPL is predominantly 1,2-DCA and 1,1,2-TCA.  
2. Below and continuously for a distance of approximately 250 m downgradient of 
the CCl4 and PCE production facilities, with representative signatures.  
3. At 2 localised positions associated with the redundant hazardous waste site 1. 
The locations of residual DNAPL were identified using ribbon NAPL samplers and 
hydrophobic dyes on the cores. Pooled DNAPL was found primarily associated with 
stratigraphic traps associated with a decrease in porosity or above less permeable 
geological layers. Pooling of DNAPL occurs generally occurs at approximately 20 m 
bgl and 37 m bgl.  
2-D surface resistivity anomalies in shallow and deep bedrock are mostly associated 
with the location of DNAPL source zones at the Investigation Site. 
The dissolved phase chlorinated hydrocarbon plumes at the site were significantly 
less horizontally extensive than expected, given the extensive source zones. This is 
likely attributed to low transmissivity, limited interconnectivity of fractures and biotic 
and abiotic degradation.  
170 

2
1
3 6
5
4
Figure 3-65: Schematic cross sectional view of the conceptual site model for the Investigation Site
1 Residual and adsorbed DNAPL in Vapour plume Residual DNAPL in saturated zone2 3
unsaturated zone    
Pooled DNAPL Dissolved phase plume Dissolved phase discharge to stream
4 5 6
172 
The chlorinated hydrocarbon DNAPL found at the site do not readily degrade. 
Additionally, microbial activity within source zones seems to no longer be viable. This 
is likely related to the high volume and toxicity of the source zones. However, the 
presence of daughter products of chlorinated ethenes and ethanes within source 
zones indicate that biotic degradation did take place initially. The predominantly 
reducing conditions as well as the presence of microbial activity in the dissolved 
plume indicate a potential for biodegradation of contaminants.    
The permeability of the weathered profile is very low, with groundwater migration 
being controlled by the location, orientation and interconnectivity of joints and 
hydraulic gradients. In general, zones of deeper weathering are consistent with sub-
vertical joint patterns and coincidental with synclinal and anticlinal flexure axes. The 
bedrock profile also has very low permeability. The flow of groundwater is controlled 
by the location, orientation and interconnectivity of joints/fractures and hydraulic 
gradients. Locally developed sub-vertical joint sets act as conduits for vertical 
groundwater (and DNAPL) transport through the low permeability dolerite sill to the 
more intensely jointed basal contact zone between the dolerite sill and the Karoo 
Ecca sandstone.  
Based on the location and extent of the chlorinated hydrocarbon contaminated on the 
site, the following are identified as potential exposure pathways and receptors: direct 
dermal contact with shallow contaminated soils during construction, vapour intrusion 
into on-site buildings, and discharge into the intermittent stream south west of the site 
border.  Human health and ecological risk assessments are beyond the scope of this 
study. A Tier 2 risk assessment is being undertaken by Consultants.  
173 
CHAPTER 4 
4 DEFINING THE VALUE AND LEVEL OF SOURCE 
ZONE CHARACTERISATION REQUIRED IN A 
FRACTURED ROCK ENVIRONMENT 
4.1 Source Zone Characterisation in a Complex Geo-contaminant 
Setting 
DNAPL source zone characterisation is important to help define source zone 
remediation design. Site sub-surface heterogeneities make the detection of source 
zones (particularly adsorbed or residual DNAPL) difficult. Chapter 3 of this research 
explored a case study of an industrial site in South Africa where various approaches 
were used to delineate the DNAPL source zones in a complex geological setting. The 
site makes an interesting case study to explore the evolution of the value of efficient 
characterisation through different approaches over time of evolving legislative 
requirements and increasing knowledge of the risks and challenges posed by 
DNAPLs. Figure 4-1 is a schematic illustration of the effort made into understanding 
the site characteristics and risks associated with contaminants through 
reconnaissance and detailed assessments over time. The drilling of monitoring wells 
in the mid- to late- 1990’s corresponds with the increased environmental awareness 
internationally of non-aqueous phase liquids as well as the promulgation of legislation 
related to the protection of natural resources and the duty of preventing and/or 
rectifying environmental damage.  
A toolbox approach was adopted for characterisation of the site in the mid-2000’s. 
This was followed by combination with the Triad approach in order to obtain better 
resolution data for the site. The outcome of studies at the site were to produce 
various numerical or conceptual models in order to assist with determining the risks 
associated with the contaminants at the site and to assist with further data gathering 
or  remediation planning.  
174 
Increased understanding on 
the role of heterogeneities on 
DNAPL transport 
Increasing environmental awareness 
and new South African legislation 
(National Water Act and National 
Environmental Management Act)
1968 Time 2012
Figure 4-1: Schematic illustration of the extent of site assessments at the site  
over time 
Traditional approaches can lead to either an under or over estimation of the source 
zones. Figure 4-2 is a schematic illustration of incorrect estimation of source zones 
based on a traditional approach. In the schematic, the borehole located in “actual 
source 1” is the only borehole that identifies a “true positive” source, while the extent 
of the “Actual Source 1” and the “Actual Source 2” go undetected. Real life geological 
setting are typically heterogeneous and anisotropic which further complicates the 
scenario of achieving cost optimisation of source zones through characterisation 
using traditional approaches.  
The approaches used in this study for source zone characterisation are novel 
approaches that have been adapted to the South African scenario. In this case the 
use of rapid measurement tools and in situ bedrock characterisation tools has 
allowed for an improved understanding of site conditions. The decision on what 
remediation alternatives should be considered would be dependant on a site-specific 
risk assessment. This is being considered through other studies. 
175 
Understanding of site and
 associated risks
 Borehole  True positive
Actual Source 1
False negatives
Potential source  
Actual Source 2
Figure 4-2: Schematic representation of incorrect estimation of source zones based on
traditional site characterisation approach (modified from Ramsey et al., 2002)
For the weathered zone investigation, the on-site laboratory equipment was imported 
from the USA at a cost of USD290 000/month (actual 2010 cost) or 
R2 552 000/month (using a R8.8 to a dollar exchange rate). A total of 1392 samples 
were collected and analysed over a 2 month period. The total cost for acquiring near 
real-time data that influenced decision making and allowed for delineation of the 
weathered source zone was hence approximately R5 million. This equates to a cost 
of approximately R3600/sample. If the same number of samples were collected using 
traditional approaches the cost of the analysis alone would be R1.7 million (assuming 
a cost per sample of R1200). This cost excludes the cost of mobilisation and 
demobilisation for every sample batch and the wait time for acquiring the analysis nor 
the time to analyse data periodically. Based on this comparison alone, the use of an 
on-site laboratory seems less cost-effective if one uses the quotient of efficiency of 
cost, where 
    =     Equation 4-1
(modified from Miansney and McBratney, 2002). 
In order to ensure that the quality of information obtained using an on-site laboratory 
was comparable to that which would have been obtained from an off-site laboratory 
strict QA/QC protocols were adhered to as described previously. If quality of 
information is assumed to be 1, efficiency of cost in this case then becomes a 
function of the reciprocal of the cost of information. A higher cost will hence produce a 
lower cost efficiency and vice versa. This calculation of efficiency does not take into 
consideration whether the data collected has a value or not. This is considered 
176 
further in the next chapter. Economies of scale however produce a very different 
scenario. Commercial laboratories analyse samples in batches as they are received 
from different clients. An on-site laboratory on the other hand is dedicated to the 
sampling program. An on-site laboratory can within a 2 month period analyse more 
samples than a commercial laboratory if allowed to run 16 hours a day4. Hence, if it is 
known (or planned) that sampling is likely to produce large sample sizes; use of an 
on-site laboratory can have a higher Efficiency of cost. Figure 4-3 is a plot of the cost 
of analysis for varying sample batch size using a commercial laboratory. At a critical 
mass of approximately 4000 samples the cost of analysis using a commercial off-site 
laboratory is equivalent to having a dedicated laboratory on-site for a two month 
period.  
10000
1000
100
10
1
0 5 000 000 10 000 000 15 000 000
Cost of Analysis (ZAR) 
 Figure 4-3: Plot of cost of analysis for batches of samples using an off-site commercial 
laboratory 
The use of ribbon NAPL samplers in fractured rock at the Investigation Site allowed 
for qualitative, semi-quantitative as well as quantitative determinations of bedrock 
source zone characteristics. The information gathered from the use of the ribbon 
4 The on-site laboratory at the Investigation Site only operated during day shift hours (8 am-
5pm) during a normal work week. Additionally, a limiting factor was the speed at which 
samples are supplied to the laboratory. Drilling also only took place during day shift/normal 
work week hours. This limitation can be overcome by having additional drill crews.  
177 
Number of Samples 
NAPL samplers included determination of the location of residual DNAPL, depth 
discrete fracture transmissivity calculations and semi-quantitative determination of 
depth discrete diffuse plume concentrations. In this case we have a single technology 
that can provide information that would traditionally require several other techniques. 
Efficiency of effort can be expressed as: 
    =      Equation 4-2 
 
The installation of a ribbon NAPL sampler at the Investigation Site and the sampling 
took place on two separate days. The time taken to install the ribbon NAPL sampler 
was on average 3 hours and 23 minutes per borehole. Several other methodologies 
would have been employed (such as packer testing, single well tracer testing, 
installation and sampling of depth discrete piezometers) to gather the same 
information gathered through a single technology. Each of the alternative 
methodology would require a series of site mobilisation/demobilisation and site 
assessments of the drilled hole. A comparison of the tranmissivity measurements 
obtained through the use of the ribbon NAPL sampler vs that obtained by Packer 
testing at the Investigation Site indicates that higher sensitivities in terms of depth 
discrete fracture transmissivities are obtained through the use of the ribbon NAPL 
sampler vs Packer testing. This is however not always the case. Hence we can 
assume that the quality of information for the use of several technologies vs the 
ribbon NAPL sampler is the same. The Efficiency of effort therefore becomes the 
reciprocal of the effort to gather information. The ribbon NAPL sampler has a higher 
Efficiency of effort factor than using several other methodologies. As a result of less 
mobilisation the Efficiency of cost for the use of ribbon NAPL samplers are also 
higher than that of traditional approaches. 
Table 4-1 provides a summary of the efficiency of cost and the efficiency for novel 
approaches used at the site compared to traditional approaches.  
Table 4-1: Summary on the cost efficiency and the effort efficiency of novel approaches used 
at the Investigation Site 
Methodology Efficiencycost Efficiencyeffort
High resolution soil sampling Yes Yes
On-site chemical analysis Maybe Yes
Ribbon NAPL sampling Yes Yes
178 
4.2 Source Zone Characterisation Efficiency  
While the gathering of further site information is useful in obtaining a clearer 
conceptual site model that can allow for the application of appropriate remediation 
technologies to reduce or mitigate risks at a site, a point is eventually reached when 
the knowledge obtained through the additional information does not justify the costs 
associated to obtain the additional information.  
In the context of this study, the concept of the source zone characterisation efficiency 
is defined as the assessment of whether the methodology for assessing source 
zones in fractured rock environments are fit-for-purpose compared to the value that 
the method has in meeting an objective at minimum cost incurred. This is considered 
as an adaptation of the Level 2-type analysis in the Value of Information Analysis 
(VOIA framework proposed by Back, 2006) or data worth analysis  (Freeze et al., 
1992) and is based on Bayesian decision cost-benefit analysis  (Davis et al., 1972; 
Korving and Clemens, 2002).  Pre-posterior analysis is recommended by Baird 
(1989) and Freeze et al. (1992) prior to sample collection in order to determine its 
value.  
Source zone characterisation efficiency can be estimated through analysing the value 
of information utilising the methodology outlined by Freeze et al.  (1992). DNAPL 
source zones in fracture rock environments are typically heterogeneous and 
anisotropic. Efficiency can be a function of value of information using mathematical 
methods and/or the use of (subjective) intuition. The following steps are 
recommended which combines mathematical cost comparisons and intuitive 
methodologies.  
Step 1: Undertake prior analysis 
Let’s assume a hypothetical scenario such as in Figure 4-4. A potential DNAPL 
source zone is identified at time t0 through prior information analysis (reconnaissance 
site visits, interview of personnel, considering site history etc.). The potential DNAPL 
source zone is 100 m2 (10m x 10 m) in this hypothetical case. The site is located on 
an unconsolidated sandstone unit. Fractured bedrock underlies the unconsolidated 
sandstone. This is underlain by compactly bedded unit of mudstone and siltstone. 
The water table level is at 5 m bgl. Release on site occurred 20 years ago over a 5 
year period and has since stopped. Potential receptors are identified as the 
downgradient river and residents living downgradient of site 
179 
Potential DNAPL source zone
Downgradient river
Residential 
development 
Depth (m bgl) Resident 
borehole 
5
20
30
Figure 4-4: Hypothetical knowledge scenario at time t0
Prior analysis is the analysis of available information available at time t0. In this 
scenario very little information is available. A site owner is faced with the question as 
to whether the risk to potential receptors justifies the cost of further investigation.  The 
probability that there is a risk to potential receptors can be designated as P(Risk), 
where P(Risk)at t0 is a function of empirical indicators using the source-transportation 
pathways-potential receptors scenario. Hence, if P(Risk) is >0 (i.e. there is potential 
for harm to receptors), the cost of further investigation (Cinv) is justified.  
Step 2: Undertake a pre-posterior analysis   
Pre-posterior analysis considers the information that may be obtained through 
sampling at a particular point or interval. As various alternatives in sampling 
technologies are available in source zone characterisation, each alternative should 
be considered in terms of cost vs benefit. Another factor to consider is time-scale for 
completing an investigation. 
For example, in the hypothetical scenario (Figure 4-4) the potential risks to receptors 
might not be considered imminent and the potential source zone may be further 
delineation through traditional approaches (i.e. batch sampling and analysis at an 
external laboratory over different time periods).  
Pre-posterior analysis can be undertaken by following the data quality objective 
process (United States Environmental Protection Agency, 1994) combined with a 
cost-benefit analysis of each alternative. 
180 
Step 3: Posterior Analysis 
Evaluate whether the information obtained through the data calculation has allowed 
for a decision. The question can be asked at this stage about whether there is 
sufficient information regarding the DNAPL source zone to evaluate appropriate 
remedial options to reduce risks to receptors.  
Figure 4-5 is a proposed qualitative source zone efficiency estimation methodology 
that can be followed by site owners in evaluating how much more data collection is 
required. Should no data collection be warranted at that time, monitoring of the 
source zone should continue, with regular reviews of the situation to determine if the 
risk profile of the site has changed. This process is applicable to the South African 
context. A methodology for source zone characterisation in proposed in Figure 4-6, 
which integrates traditional and novel technologies. This methodology is based on 
the evaluations undertaken at the Investigation Site. 
181 
Prior Analysis
review  to determine if there are any changes
Risk to Continue
potential NO monitoring
receptor?
YES
Consider data
collection 
alternatives
Undertake cost 
benefit analysis of 
each alternative
Pre-posterior 
analysis
Will an 
action result 
from the NO
data 
collection?
YES
Obtain data Use proposed
process for 
source zone 
characterisation
Evaluate data 
worth
Decision 
Analysis
Posterior 
Analysis
Sufficient NO
Data?
YES
Is there 
Risk to NO
receptors?
YES
Evaluate remediation 
alternatives
Figure 4-5: Proposed integrated appraoch to determine the level of source zone 
characterisation required 
182 
In Situ 
Measurements:
Use existing information Bai ler drawdown
to inform detailed Drilling and detailed Use high resolution Ribbon NAPL Samplers,
upfront planning and geological examination methods to delineate 
development of a of cores outside of vapour plume and/or 
Downhole geophsyics
preliminary CSM potential source areas release areas
Use the rock core for 
measurements of geo-structural 
Use 2-D geophysical source zone drilling and physical properties
measurements to 
identify Detailed planning of 
phase corresponding 
with geophysical 
geological/contaminant source zone drilling anomalies and core 
anomolies examination information UMseea tshuere r opchky sciocrael  afonrd  chemical 
pmreoapseurrteiems ednutrsin ogf  dgreiolli-nsgtructural 
and physical properties Other measurements:
Groundwater gauging,
Groundwater analysis
Figure 4-6: Proposed source zone characterisation methodology for the South African scenario
183 
CHAPTER 5 
5 CONCLUSIONS AND RECOMMENDATIONS 
5.1 Novel Approaches to DNAPL Source Zone Characterisation in 
a Fractured Rock Environment 
This research presents pioneering work in terms of source zone characterisation and 
the understanding on the architecture of chlorinated hydrocarbon DNAPLs in a 
fractured rock environment.  
• This study is the largest known and most detailed mixed chlorinated hydrocarbon 
DNAPLs source zone characterisation in a fractured undertaken internationally.  
• This is the first record of successfully locating and characterising in 3 dimensions 
the architecture of mixed chlorinated hydrocarbon DNAPLs in a fractured rock 
environment on a field scale. 
• The success of locating and successfully characterising chlorinated hydrocarbon 
DNAPL sites has lagged in South Africa. This is largely due to the lack of 
available technologies. This study marks the first use of FLUTeTM to characterise 
DNAPL in the South African fractured environment. The study also evaluates the 
cost efficiency and effort efficiency of this in situ technology and concludes that 
there is merit in importing this technology to South Africa due to its high success 
rate in locating residual DNAPL as well as in depth discrete fracture and 
dissolved phase plume characterisation. 
• While previous South African authors (for e.g. Usher et al., 2008) have proposed 
DNAPL site characterisation methodologies adapted to the South African 
scenario, the author in this research uses a systematic approach of combining 
traditional technologies together with  emerging/novel technologies (such as high 
resolution sampling) to characterise chlorinated hydrocarbon DNAPL source 
zones. Various technologies are systematically evaluated in this study and the 
author can hence propose improved integrated approaches to undertake 
fractured environment source characterisation.  
5.2 Source Zone Characteristics in a Fractured Rock Environment 
The following conclusions are based on the evaluation of multiple lines of evidence to 
characterise mixed chlorinated hydrocarbon DNAPLs in a fractured rock environment: 
184 
• Properties that affect the architecture of DNAPL in fractured rock 
environments include: 
o The nature, location, size and extent of the release 
o Geological properties such as the depth of weathering, fracture 
intensity, fracture orientation and interconnectivity 
o Hydrogeological properties such as the fracture aperture transmissivity 
or permeability, vertical and horizontal hydraulic gradients and 
groundwater flow direction; and 
o Biotic and abiotic degradation processes 
• Successful source characterisation in a fractured rock environment 
requires a systematic approach integrating traditional and novel 
approaches and utilising multiple lines of evidence.  
5.3 Source Zone Architectures of the Investigation Site 
The Investigation Site has had a long history of releases of chlorinated hydrocarbon 
DNAPLs. The following are concluded with respect to the architecture of source 
zones at the site: 
• Release areas at the site can be inferred based on the soil-gas profile of the 
chlorinated hydrocarbon DNAPLs. Release areas are inferred at the historical 
production and handling facilities as well as the hazardous waste sites. 
• Drip-releases of chlorinated hydrocarbon DNAPLs has led to variable 3-
dimensional architectures in the weathered profile underlying the inferred 
release areas. The source zones in the weathered profile have an “onion” 
architecture with the highest concentrations located at the core of the 
contaminated mass. The most significant mass of chlorinated hydrocarbon 
DNAPLs at the Investigation Site is located at the historical production and 
handling facilities. Significantly less mass of chlorinated hydrocarbon 
DNAPLs is located at the redundant waste sites. This could be as a result of 
storage of the chemicals in drums prior to disposal on the redundant waste 
sites vs regular spills and leaks at the historical production facilities. While 
there has been vertical migration of chlorinated hydrocarbon DNAPLs in the 
weathered profile, there is very little evidence of lateral migration, largely 
attributed to the media properties. 
• Continuous leaks of chlorinated hydrocarbon DNAPLs into the subsurface 
from the historical production and handling facilities has led to two long 
continuous DNAPL plumes (approximately 650 and 850 m in length) in the 
185 
fractured bedrock. The bedrock DNAPL plume emanating from the historical 
above-ground storage tanks is predominantly 1,2-DCA and VC while the 
plume emanating from the historical organics plant consists of PCE and CCl4. 
The contiguous DNAPL plumes in the bedrock indicate movement along 
fractures. 
• There are no significant biotic degradation reactions occurring in the DNAPL 
source zones. This can be attributed to the large volume of chlorinated 
hydrocarbon DNAPL present. 
5.4 Recommendations for Future Work 
• Sampling of boreholes along the source zone plumes and analysing for 
compound specific isotopes to determine the site-specific rate of 
biodegradation at the site. This will assist in the development of 
representative numerical models. 
• Multiphase numerical models of the site. 
• Simulating the effect of partial source removal on the dissolved phase 
plume. 
• Evaluating the FLUTeTM technologies in other DNAPL contaminated sites, 
with diverse geo-contaminant settings, in South Africa. 
• Undertaking cost modelling to optimise sampling scales for high resolution 
investigations. 
186 
6 REFERENCES 
 Abe, Y.; Aravena, R.; Zopfi, J.; Shouakar-Stash, O.; Cox, E.; Roberts, J.D.; Hunkeler, 
D. (2009). Carbon and chlorine isotope fractionation during aerobic odixadion 
and reductive decholorinatiom of vinyl chloride and cis-1,2-dichloroethene. 
Environmental Science and Technology, 43(1), 101-107. 
Adamson, A.W.; Gast, A.P. (1997). Physical Chemistry of Surfaces (6th ed.). New 
York: Wiley. 
Adamson, D.T.; Parkin, G.F. (2001). Product distribution during transformation of 
multiple contaminants by a high-rate, tetrachlorethene-dechlorinating 
enrichment culture. Biodegradation, 12(5), 337-348. 
Air Force Center for Engineering and the Environment. (2007, May). AFCEE source 
zone initiative. Retrieved May 10, 2009, from www.clu-in.org: http://www.clu-
in.org/download/contaminantfocus/dnapl/Chemistry_and_Behavior/AFCEE-
szi-2007a.pdf 
Anderson, M.R. (1988). The dissolution and transport of dense non-aqueous phase 
liquids in saturdated porous media. PhD dissertation (unpublished), Oregon 
Graduate Center, Department of Environmental Science and Engineering, 
Beaverton. 
Back, P.-E. (2006). Value of information analysis for site investigations in remediation 
projects. PhD Thesis (unpublished), Chalmers University of Technology, 
Göteborg. 
Baird, B.F. (1989). Managerial decisions under uncertainty: an introduction to the 
analysis of decision making. New York: John Wiley and Sons. 
Ballapragada, B.S.; Stensel, H.D.; Puhakka, J.A.; Ferguson, J.F. (1997). Effect of 
hydrogen on reductive dechlorination of chlorinated ethenes. Environmental 
Science and Technology, 31(6), 1728-1734. 
Barnes, B. (2009, February). Framework for the use of rapid measurement 
techniques (RMT) in the risk management of land contamination. Bristol, 
United Kingdom: Environmental Agency. Retrieved November 18, 2012, from 
187 
Environmental Agency web site: http://publications.environment-
agency.gov.uk/pdf/SCHO0209BPIA-e-e.pdf 
Barranco, F.T.; Dawson, H.E. (1999). Influence of aqueous pH on the interfacial 
properties of coal tar. Environmental Science and Technology, 33(10), 1598-
1603. 
Barranco, F.T.; Dawson, H.E.; Christener, J.M.; Honeyman, B.D. (1997). Influence of 
aqueous pH and ionic strength on the wettability of quartz in the presence of 
dense non-aqueous-phase liquids. Environmental Science and Technology, 
31(3), 676-681. 
Bear, J. (1993). Modeling flow and contaminant transport in fractured rocks. In J. 
Bear, C. Tsang, & G. D. Marsily (Eds.), Flow and Contaminant Transport in 
Fractured Rock (p. 560). San Diego: Academic Press. 
Belay, N.; Daniels, L. (1987). Production of ethane, ethylene and acetylene from 
halogenated hydrocarbons by methanogenic bacteria. Applied Microbiology 
Biotechnology, 53, 1604-1610. 
Beneteau, K.M.; Aravena, R.; Frape, S.K. (1999). Isotopic chracterization of 
chlorinated solvents - laboratory and field results. Organic Geochemistry, 30, 
739-753. 
Bingham, E.C. (1916). An investigation of the laws of plastic flow. U.S. Bureau of 
Standards Bulletin, 13, 309-353. 
Birkholzer, J.; Tsang, C. F. (1997). Solute channeling in unsaturated heterogeneous 
porous media. Water Resources Research, 33, 2221-2238. 
Bottcher, J.; Strebel, O.; Voerkelius, S.; Schmidt, H.L. (1990). Using isotope 
fractionation of nitrate-nitrogen and nitrate-oxygen for evaluation of microbial 
denitrification in a sandy aquifer. Journal of Hydrology, 114, 419-424. 
Boyles, W. (1997). The science of chemical oxygen demand technical information.
Retrieved November 8, 2010, from Hach Company Web Site: 
http://www.hach.com/download-resources 
188 
Broholm, K.; Feenstra, S.; Cherry, J.A. (1999). Solvent release into a sandy aquifer. 
1. Overview od source distribution and dissolution behavior. Environmental 
Science and Technology, 33(5), 681-690. 
Burton, J.C.; Walker, J.L.; Jennings, T.V.; Aggarwal, P.K.; Hastings, B. (1993). 
Expedited site characterisation: a rapid, cost-effective process for remedial 
site characterisation. Proceedings of the Superfund XIV Conference . II, pp. 
809-826. Greenbelt, MD: Hazardous Materials and Control Institute. 
Carey, M.A.; Finnamore, J.R.; Morrey, M.J.; Marsland, P.A. (2000). Guidance on the 
Assessment and Monitoring of Natural Attentuation of Contaminants in 
Groundwater. National Groundwater and Contaminated Land Centre, UK 
Environment Agency. 
Cary, J.W.; Simmons, C.S.; McBride, J.F. (1989). Predicting oil infiltration and 
redistribution in unsaturated soils. Soil Science Society of America Journal, 
53(2), 335-342. 
Chang, L.C.; Chen, C.H.; Shan, H.; Tsai, J. (2009). Effect of connectivity and 
wettability on the relative permeability of NAPLs. Environmental Geology, 56, 
1437-1447. 
Chapman, S.W.; Parker, B.L. (2005). Plume persistence due to aquitard back 
diffusion following dense nonaqueous phase liquid removal or isolation. Water 
Resource Research, 41(12), W12411. 
Chatzis, I.; Morrow, N.R.; Lim, H.T. (1983). Magnitude and detailed structure of 
residual oil saturation. Society of Petroleum Engineers Journal, 311-326. 
Chen, C.; Puhakka, J. A.; Ferguson, J. F. (1996). Transformation of 1,1,2,2-
tetrachloroethane under methanogenic conditions. Environmental Science 
and Technology, 30, 542-547. 
Cheremisinoff, P.N. (1990). Calculating and reporting toxic chemical releases for 
pollution control. New Jersey: SciTech Publishers. 
189 
Cho, H.J.; Fiacco, R.J.; Daly, M.H.; McTigue, J.W. (2004). Hydrochemical facies 
analysis of 1,1,1-trichloroethane and its degradation products in fractured 
bedrock. USEPA/National Groundwater Association Fractured Rock 
Conference. Portland. 
Clark, I.D.; Fritz, P. (1997). Environmental isotopes in Hydrogeology (1st ed.). Boca 
Raton Florida: Lewis Publishers. 
Cohen, R.M.; Mercer, J.W. (1993). DNAPL site evaluation. Florida: CRC Press Inc. 
Cook, P.F. (Ed.). (1991). Enzyme mechanism from isotope effects. Boca Raton, 
Florida: CRC Press, Inc. 
Cope, N.; Hughes, J. B. (2001). Biologically-enhanced removal of PCE from NAPL 
source zones. Environmental Science and Technology, 35(10), 2014-2021. 
Davis, D. R.; Kisiel, C.C.; Duckstein, L. (1972). Bayesian decision theory applied to 
design in hydrology. Water Resources Research, 8(1), 33-41. 
Dawson, H.; Illangasekasre, T.H. (1999). Influence of geologic heterogeneity and 
chemical complexity on the transport and distribution of nonaqueous phase 
liquid waste . USEPA Region 8: United States Environmental Protection 
Agency. 
de Best, J.H.; Hunneman, P.; Doddema, H.J.; Janssen, D.B.; Harder, W. (1999). 
Transformation of carbon tetrachloride in an anaerobic packed-bed reactor 
without addition of another electron donor. Biodegradation, 10(4), 287-295. 
de Best, J.H.; Salminen, E.; Doddema, H.J.; Janssen, D.B; Harder, W. (1997). 
Transformation of carbon tetrachloride under sulfate reducing conditions. 
Biodegradation, 8(6), 429-436. 
Dennis, I.; Pretorius, J.; Steyl, G. (2010). Effect of fracture zone on DNAPL transport 
and dispersion: a numerical approach. Environmental Earth Sciences, 61(7), 
1531-1540. 
190 
Diomampo, G.P.; Chen, C.Y.; Li, K.; Horne, R.N. (2002). Relative permeability 
through fractures. Proceedings of the 27th Stanford Workshop on Geothermal 
Reservoir Engineering. Stanford: Stanford University. 
Doherty, R.E. (2000). A history of production and use of carbon tetrachloride, 
tetrachloroethylene, trichloroethylene, and 1,1,1-trichloroethane in the United 
States; Part 1, Historical background; carbon tetrachloride and 
tetrachloroethylene. Journal of Environmental Forensics, 1(2), 69-81. 
Durnford, D.S.; McWhorter, D.B.; Miller, C.D.; Swanson, A.; Marinelli, F.; Trantham, 
H.L. (1997). DNAPL and LNAPL distributions in soils: experimental and 
modeling studies. Colorado State University, Chemical and Bioresources 
Engineering, Colorado. 
Elsner, M.; Zwank, L.; Hunkeler, D.; Schwarzenbach, R. P. (2005). A new concept 
linking observable stable isotope fractionation to transformation pathways of 
organic pollutants. Environmental Science and Technology, 39, 6896–6916. 
Ennis, E.; Reed, R.; Dolan, M; Semprini, L'; Istok, J.D.; Field, J.A. (2005). Reductive 
dechlorination of the vinyl chloride surrogate chlorofluoroethene in TCE 
contaminated groundwater. Environmental Science and Technology, 39, 
6777-6785. 
Feenstra, S.; Cherry, J. A.; Parker, B. L. (1996). Conceptual models for the behavior 
of dense non-aqueous phase liquids (DNAPLs) in the subsurface. In J. F. 
Pankow, & J. A. Cherry (Eds.), Dense Chlorinated Solvents and Other 
DNAPLs in Groundwater (pp. 53-88). Portland: Waterloo. 
Fetter, C.W. (1999). Contaminant Hydrogeology (2nd ed.). New Jersey: Prentice Hall. 
Fiacco, R.J.; Daly, M.H.; Drobinski, J.C. (2005). Application of the toolbox approach 
to characterize fractured bedrock sites. Proceedings of the The Annual 
Conference on Soils, Sediments, Water and Energy. Amherst, MA: Univesity 
of Massachusetts. 
191 
Field, J.A.; Istok, J.D.; Semprini, L.; Bennett, P.; Buscheck, T. (2005). 
Trichlorofluorethene: A reactive tracer for evaluating in situ trichloroethene 
remediation. Ground Water Monitoring and Remediation, 25, 1-10. 
Fischer, A.; Herklotz, I.; Herrmann, S.; Thullner, M.; Weelink, S. A. B.; Stams, A. J. 
M.; Schlomann, M.; Richnow, H. H.; Vogt, C. (2008). Combined carbon and 
hydrogen isotope fractionation investigations for elucidating benzene 
biodegradation pathways. Environmental Science and Technology, 42, 4356–
4363. 
Freeze, R. A.; Bruce, J.; Massman, J.; Sperling, T.; Smith, L. (1992). Hydrogeological 
Decision Analysis: 4. The Concept of Data Worth and Its Use in the 
Development of Site Investigation Strategies. Ground Water, 30(4), 574-588. 
Freeze, R.A.; Cherry, J.A. (1979). Groundwater. New Jersey: Prentice Hall, Inc. 
Ge, S. M. (1997). A governing equation for fluid in rough fractures. Water Resources 
Reseach, 33, 53-61. 
Gebrekristos, R.A. (2007). Site characterisation methodologies for DNAPLs in 
fractured South African aquifers. PhD thesis (unpublished), University of the 
Free State, Institute for Groundwater Studies, Bloemfontein. 
Gebrekristos, R.A.; Pretorius, J.A.; Usher, B.H. (2008). Manual for site sssessment at 
DNAPL contaminated sites in South Africa. Pretoria: Water Research 
Commission. 
Gooddy, D.C.; Bloomfield, J.P.; Harold, G.; Lehame, S.A. (2002). Towards a better 
understanding of tetrachloroethene entry pressure in the matrix of Permo-
Triassic sandstones. Journal of Contaminant Hydrology, 59, 247-265. 
Gossett, J. M. (1987). The measurement of Henry’s Law constants for C1 and C2 
chlorinated compounds.  Environmental Science and Technology. 21, 202–
208. 
Griebler, C.; Safinowski, M.; Vieth, A.; Richnow, H.H.; Meckenstock, R.U. (2004). 
Combined application of stable carbon isotope analysis and specific 
192 
metabolites determination for assessing in situ degradation of aromatic 
hydrocarbons in a tar oil-contaminated aquifer. Environmental Science and 
Technology, 38, 617 – 631. 
Griffin, T.W.; Watson, K.W.. (2002). A comparison of field techniques for confirming 
dense nonaqueous phase liquids. Ground Water Monitoring and Remediation, 
22, 48-59. 
Hageman, K.J.; Istok, J.D.; Buscheck, T.E.; Semprini, L. (2001). In situ anaerobic 
transformation of trichlorofluroroethene in trichloroethene-contaminated 
groundwater. Environmental Science and Technology, 35, 1729-1735. 
Halogenated Solvents Industry Alliance. (2004). Solvents Applications. Retrieved 
May 8, 2009, from http://www.hsia.org/applications.htm 
Harrison, B.; Sudicky, E.A.; Cherry, J.A. (1992). Numerical analysis of solute 
migration through fractured clayey deposits into underlying aquifers. Water 
Resources Research, 28, 515-526. 
Harrold, G.; Gooddy, D.C.; Reid, S.; Lerner, D.N.; Leharne, S.A. (2003). Changes in 
interfacial tension of chlorinated solvents following flow through U.K. soils and 
shallow aquifer material. Environmnetal Science and Technology, 37(9), 
1919-1925. 
Hassanizadeh, S.M.; Celia, M.A.; Dahle, H.K. (2002). Dynamic Effect in the Capillary 
Pressure-Saturation Relationship and its Impacts on Unsaturated Flow. 
Vadose Zone Journal, 1, 38-57. 
He, J.; Ritalahti, K. M.; Aiello, M. R.; Löffler, F.E. (2003). Complete detoxification of 
vinyl chloride by an anaerobic enrichment culture and identification of the 
reductively dechlorinating population as a Dehalococcoides Species. Applied 
Environmental Microbiology, 69, 996-1003. 
Hiemenz, P.C.; Rajagopalan, R. (1997). Principles of colloid and surface chemistry
(3rd ed.). New York: Marcel Dekker, Inc. 
193 
Horvath, A.L. (1982). Halogenated Hydrocarbons: Solubility-Miscibility with Water.
New York, USA: Marcel Dekker.  
Horvath, A.L.; Getzen, F.W.; Maczynska, Z. (1999). IUPAC-NIST solubility data 
series 67.Halogenated ethanes and ethenes with water. Journal of Physical 
Chem Reference Data, 28, 395-628. 
Hughes, B.M.; Gillham, R.W.; Mendoza, C.A. (1990). Transport of trichloroethylene 
vapors in the unsaturated zone: a field experiment. International Association 
of Hydrogeologist Conference on Subsurface Contamination by Immiscible 
Fluids. Calgary. 
Huling, S.G.; Weaver, J.W. (1991). Ground Water Issue: Dense Nonaqueous Phase 
Liquids, EPA/540/4-91-002. Washington, D.C.: United States Environmental 
Protection Agency Office of Solid Waste and Emergency Response. 
Hulley, V. (2009, December). Application of GIS in contaminated site investigations 
and remediation planning: transition from data generation to decision-making. 
PositionIT, p. 4. 
Hulley, V.; Smit, J.; Kotze, J. (2008). Case Study of Mixed-Contaminant Plume 
Characterization in a Fractured Rock Aquifer Environment. Monterey, 
California: Battelle. 
Hunkeler, D; Aravena, R.; Butler, B. J. (1999). Monitoring microbial dechlorination of 
tetrachloroethene (PCE) in groundwater using compound-specific stable 
carbon isotope ratios: Microcosm and field studies. Environmental Science 
and Technology, 33, 2733-2738. 
Hunkeler, D.; Andersen, N.; Aravena, R.; Bernasconi, S. M.; Butler, B.J. (2001). 
Hydrogen and carbon isotope fractionation during aerobic biodegradation of 
benzene. Environmental Science and Technology, 35, 3462-3467. 
Hunkeler, D.; Aravena, R.; Cox, E. (2002). Carbon isotopes as a tool to evaluate the 
origin and fate of vinyl chloride: laboratory experiments and modelling of 
isotope evolution. Environmental Science and Technology, 36, 3378 – 3384. 
194 
Hunkeler, D.; Aravena, R.; Berry-Spark, K.; Cox, E. (2005). Assessment of 
degradation pathways in an aquifer with mixed chlorinated hydrocarbon 
contamination using stable isotope analysis. Environmental Science and 
Technology, 39(16), 5975-5981. 
Hunkeler, D.; Meckenstock, R.U.; Sherwood Lollar, B.; Schmidt, T.C.; Wilson, J.T. 
(2008). A Guide for Assessing Biodegradation and Source Identification of 
Organic Ground Water Contaminants using Compound Specific Isotope 
Analysis (CSIA). Office of Research and Development. Ada, Oklahoma: 
United States Environmental Protection Agency. 
Illangasekare, T.H.; Ramsey, J. L.; Jensen, K.H.; Butts, M. (1995). Experimental 
study of movement and distribution of dense organic contaminants in 
heterogeneous aquifers. Journal of Contaminant Hydrology, 20, 1-25. 
International Atomic Energy Agency. (2002). Study of environmental change using 
isotope techniques. Proceedings: Study of environmental change using 
isotope techniques (p. 531). Vienna: International Atomic Energy Agency. 
Interstate Technology and Regulatory Council. (2000, June). Dense non-aqueous 
phase liquids (DNAPLs): review of emerging characterization and remediation 
technologies. Retrieved May 5, 2009, from www.http://itrcweb.org/: 
www.itrcweb.org/Guidance/GetDocument?documentID=18 
Interstate Technology and Regulatory Council. (2003, December). Technical and 
regulatory guidance for the Triad Approach: a new paradigm for 
environmental project management. Retrieved August 29, 2009, from 
http://www.itrcweb.org/GuidanceDocuments/SCM-1.pdf 
Istok, J.D; Field, J.A.; Raes, E.; Millings, M.R.; Peacock, A.D. (2007). Detecting and 
quantifying reductive dechlorination during monitored natural attenuation at 
the Savannah River CBRP site. Washington: Savannah River Company. 
Jackson, R.E.; Dwarakanath, V. (1999). Chlorinated degreasing solvents: physical-
chemical properties affecting aquifer contamination and remediation. Ground 
Water Monitoring and Remediation, 19(4), 102-110. 
195 
Jeffers, P. M.; Ward, L. M.; Woytowitch, L. M.; Wolfe, N. L. (1989). Homogeneous 
hydrolysis rate constants for selected chlorinated methanes, ethanes, 
ethenes, and propane. Environmental Science and Technology, 23, 965-969. 
Jellali, S.; Muntzer, P.; Razakarisoa, O.; Schäefer, G. (2001). Large scale experiment 
on transport of trichloroethylene in a controlled aquifer. Transport in Porous 
Media, 145-163. 
Kerndorff, H.; Schleyer, R.; Milde, G.; Plumb, R.H. Jr. (1992). Geochemistry of 
groundwater pollutants at German waste disposal sites. In S. Lesage, & R. 
Jackson (Eds.), Groundwater contamination and analysis at hazardous waste 
sites (pp. 245-271). New York: Marcel Dekker Inc. 
Kiersch, G.A. (1958). Geological causes for failure of Lone Pine Reservoir, Eastern 
Central Arizona. Economic Geology, 55, 854. 
Klecka, G. M.; Carpenter, C. L.; Gonsior, S. J. (1998). Biological transformations of 
1,2-dichloroethane in subsurface soils and groundwater. Journal of 
Contaminant Hydrology, 34, 139-154. 
Kopinke, F.D.; Georgi, A.; Voskamp, M.; Richnow, H.H. (2005). Carbon isotope 
fractionation of organic contaminants due to retardation on humic substances: 
implications for natural attenuation studies in aquifers. Environmental Science 
and Technology, 39(16), 6052-6062. 
Korving, H.; Clemens, F. (2002). Bayesian decision analysis as a tool for defining 
monitoring needs in the field of effects of CSOs on receiving waters. Water 
Science and Technology, 45(3), 175-184. 
Kram, M.L.; Keller, A.A.; Rossabi, J.; Everett, L.G. (2001). DNAPL characterization 
methods and approaches, Part 1: Performance comparisons. Ground Water 
Monitoring and Remediation, 21, 19-123. 
Kram, M.L.; Keller, A.A.; Rossabi, J.; Everett, L.G.. (2002). DNAPL Characterization 
Methods and Approaches, Part 2: Cost Comparisons. Ground Water 
Monitoring and Remediation, 46-61. 
196 
Kueper, B.H.; Davies, K.L (2009, September). Assessment and delineation of DNAPL 
source zones at hazardous waste sites, EPA/600/R-09/119. Retrieved 
February 17, 2010, from www.epa.gov/nrmrl/pubs/600r09119.htm 
Kueper, B.H.; McWhorter, D.B. (1991). The behaviour of dense, nonaqueous phase 
liquids in fractured clay and rock. Ground Water, 29(5), 716-728. 
Kueper, B.H.; Abbot, W.; Farguhar, G. (1989). Experimental observations of 
muliphase flow in heterogeneous porous media. Journal of Contaminant 
Hydrology, 5, 83-95. 
Kueper, B.H.; Redman, D.; Star, R.B.; Reitsma, S.; Mah, M. (1993). A field 
experiment to study the behavior of tetrachloroethylene below the water table: 
spatial distribution of residual and pooled DNAPL. Ground Water, 31(5), 756-
766. 
Kueper, B.H.; Walthall, G.P.; Smith, J.W.N.; Leharne, S.A.; Lerner, D.N. (2003). An 
illustated handbook of DNAPL transport and fate in the subsurface. Bristol: 
Environmental Agency. 
Lambe, W.T.; Whitman, R.V. (1969). Soil mechanics. John Wiley and Sons Inc. 
Leighton, D.T.J.R.; Calo, J.M.. (1981). Distribution coe cients of chlorinated 
hydrocarbons in dilute air–water systems for groundwater contamination 
applications. Journal of Chemical Engineering, 26, 382-385. 
Lenhard, R.J.; Oostrom, M.; Dane, J.H. (2004). A constituitive model for air-NAPL-
water flow in the vadose zone accounting for immobile, non-occluded 
(residual) NAPL in strongly water-wet porous media. Journal of Contaminant 
Hydrology, 71, 261-282. 
Liang, X.; Dong, Y.; Kuder, T.; Krumholz, L.R.; Philp, P.; Butler, E.C. (2007). 
Distinguishing abiotic and biotic transformation of tetrachloroehtylene and 
trichloroethylene by stable carbon isotope fractionation. Environmental 
Science and Technology, 41(20), 7094-7100. 
197 
Liu, C.; Ball, W. P. (2002). Back diffusion of chlorinated solvent contaminants from a 
natural aquitard to a remediated aquifer under well-controlled field conditions: 
predictions and measurements. Ground Water, 40(2), 175-184. 
MacDonald, J.A.; Kavanaugh, M.C. (1994). Restoring contaminated groundwater: an 
achievable goal? Environmental Science and Technology, 28(8), 362-368. 
MacDougall, K.A.; Fenning, P.; Cooke, D.A.; Preston, H.; Brown, A.; Hazzard, J.; 
Smith, T. (2002). Guidance on the selection of non-intrusive techniques for 
groundwater pollution studies, R&D Technical Report P2-178/TR/2.
Almondsbury, Bristol: Environmental Agency . 
Mackay, D.; Bobra, A.; Shiu, W.Y.; Yalkowsky, S.H. (1980). Relationships between 
aqueous solubility and octanol-water partition coefficient. Chemosphere,
9, 701-711. 
Mackay, D.; Shiu, W.Y.; Maijanen, A.; Feenstra, S. (1991). Dissolution of non-
aqueous phase liquids in groundwater. Journal of Contaminant Hydrology, 8, 
23-42. 
Mackay, D.M.; Cherry, J.A. (1989, June). Groundwater Contamination: pump-and-
treat remediation. Environmental Science and Technology, 23(6), 630-636. 
Major, D.W.; McMaster, M.L.; Cox, E.E.; Edwards, E.A.; Dworatzek, S.M.; 
Hendrickson, E.R.; Starr, M.G.; Payne, J.A.; Buonanici, L.W. (2002). Field 
demonstration of successful bioaugmentation to achieve dechlorination of 
tetrachloroethene to ethene. Environmental Science and Technology, 36, 
5106-5116. 
Mancini, S.A.; Ulrich, A.C.; Lacrampe-Couloume, G.; Sleep, B.; Edwards, E.A.; 
Sherwood Lollar, B. (2003). Carbon and hydrogen isotopic fractionation during 
anaerobic biodegradation of benzene. Applied Environmental Microbiolgy, 69, 
191-198. 
Martinez, M. J.; Dykhuizen, R. C.; Eaton, R. R. (1992). The apparent conductivity for 
steady unsaturated flow in periodically fractured porous media. Water 
Resources Research, 28, 2879-2887. 
198 
McCarty, P.L. (1997). Biotic and abiotic transformations of chlorinated solvents in 
groundwater. Proceedings of the symposium on natural attenuation of 
chlorinated organics in groundwater (EPA/540/R-97/504) (p. 191). 
Washington, D.C.: United States Environmental Protection Agency Office of 
Research and Development. 
Meckenstock, R.U.; Morasch, B.; Griebler, C.; Richnow, H.H. (2004). Stable isotope 
fractionation analysis as a tool to monitor biodegradation in contaminated 
aquifers. Journal of Contaminant Hydrology, 75, 215-255. 
Melander, L.; Saunders, W.H. (1980). Reaction rates of isotopic molecules (1st ed.). 
New York: John Wiley and Sons. 
Melber, C.; Kielhorn, J.; Mangelsdorf, I. (2004). Coal tar creosote. Geneva: World 
Health Organization. 
Mercer, J.W.; Cohen, R.M. (1990). A review of immiscible fluids in the subsurface: 
Properies, models, characterisation and remediation. Journal of Contaminant 
Hydrology, 6, 107-163. 
Meylan, W.M.; Howard, P.H.; Boethling, R.S. (1992).  “Molecular topology/fragment 
contribution method for predicting soil sorption coefficients.” Environmental 
Science and Technology, 26, 1560-1567. 
Miansney, B; McBratney, A.B. (2002). The efficiency of various approaches to 
obtaining estimates of soil hydraulic properties. Geoderma, 107(1), 55-70. 
doi:10.1016/S0016-7061(01)00138-0 
Middeldorp, P.J.M.; Luijten, M.L.G.C.; van de Pas, B.A.; van Eekert, M.H.A.; Kengen, 
S.W.M.; Schraa, G.; Stams, A.J.M. (1999). Anaerobic microbial reductive 
dehalogenation of chlorinated ethenes. Bioremediation, 3, 151-169. 
Mohammad, O.I.; Kibbey, T.C.G. (2005). Dissolution-induced contact angle 
modification in dense nonaqueous phase liquid/water systems. Environmental 
Science and Technology, 39(6), 1698-1706. 
199 
Moran, M.J. (2006). Occurence and implications of selected chlorinated solvents in 
ground water and source water in the United States and in the drinking water 
in 12 Northeast and Mid-Atlantic States, 1993-2002. U.S. Geological Survey 
Scientific Investigations Report. 
Murphy, J. R.; Thomson, N. R. (1993). Two-phase flow in variable aperture fracture. 
Water Resources Research, 24, 2033-2048. 
National Research Council . (2000). Natural attenuation for groundwater remediation.
Washington D.C.: National Academy Press. 
National Research Council. (2004). Contaminants in the subsurface: Source zone 
assessment and remediation. Washington D.C.: National Academic Press.  
Newell, C.J.; Aziz, C.E. (2004). Long-term sustainability of reductive dechlorination 
reactions at chlorianted solvents sites. Biodegradation, 15, 387-394. 
O'Hara, S.K.; Parker, B.L.; Jorgensen, P.R.; Cherry, J.A. (2000). Trichloroethene 
DNAPL flow and mass distribution in naturally fractured clay: Evidence of 
aperture variability. Water Resources Research, 36(1), 135–147. 
Oron, A. P.; Berkowitz, B. (1998). Flow in rock fractures: the local cubic law 
assumption reexamined. Water Resources Research, 34, 2811-2825. 
Pagan, M.; Cooper, W. J.; Joens, J. A. (1998). Kinetic studies of the homogeneous 
abiotic reactions of several chlorinated aliphatic compounds in a aqueous 
solution. Applied Geochemistry, 13, 779-785. 
Pankow, J.F.; Cherry, J.A. (1996). Dense chlorinated solvents and other DNAPLs in 
groundwater. Ontario: Waterloo Press. 
Parker, B.L. (2007). Investigating contaminated sites on fractured rock using the DFN 
Approach. U.S. EPA/NGWA Fractured Rock Conference: State of the Science 
and Measuring Success in Remediation. Portland, Maine: U.S. EPA/NGWA. 
200 
Parker, B.L.; Cherry, J.A.; Chapman, S.W. (2004). Field study of TCE diffusion 
profiles below DNAPL to assess aquitard integrity. Journal of Contaminant 
Hydrology, 74, 197-230. 
Parker, B.L.; Gillham, R.W.; Cherry, J.A. (1994). Diffusive disappearance of 
immiscible phase organic liquids in fractured geologic media. Ground Water, 
32(5), 805-820. 
Parker, B.L.; McWhorter, D.B.; Cherry, J.A. (1997). Diffusive loss of non-aqueous 
phase organic solvents from idealized fracture networks in geologic media. 
Ground Water, 35(6), 1077-1088. 
Poulsen, M.M.; Kueper, B.H. (1992). A field experiment to study the behavior of 
tetrachloroethylene in unsaturated porous media. Environmental Science and 
Technology, 26(5), 889-895. 
Powers, S.E.; Abriola, L.M.; Weber, W.J. Jnr. (1992). An experimental investigation of 
nonaqueous phase liquid dissolution in saturdated subsurface systems: 
steady state mass transfer rates. Water Resources Research, 28(10), 2691-
2705. 
Ramsey, M.H.; Taylor, P.D.; Lee, J. (2002). Optimized contaminated land 
investigation at minimum overall cost to achieve tness-for-purpose. Journal 
of Environmental Monitoring, 4, 809-814. 
Republic of South Africa. (1998, August 26). National Water Act (Act No 36 of 1998). 
398. Cape Town, Republic of South Africa: Republic of South Africa 
Government Gazette. 
Rivett, M.O.; Lerner, D.N.; Lloyd, J.W. (1990). Chlorinated solvents in UK aquifers. 
Water and Environment Journal, 4(3), 242-250. 
Robertson, B. K.; Alexander, M. (1996). Mitigating toxicity to permit bioremediation of 
constituents of nonaqueous-phase liquids. Environmental Science and 
Technology, 30, 2066-2070. 
201 
Sabljic, A.; Gusten, H.; Verhaar, H.; Hermens, J. (1995). QSAR modelling of soil   
sorption. Improvements and systematics of logKoc vs. logKow correlations. 
Chemosphere, 31, 4489–4514. 
Sale, T.; Newell, C.; Stroo, H.; Hinchee, R.; Johnson, P. (2008). Frequently asked 
questions regarding management of chlorinated solvents in soils and 
groundwater. Environmental Security Technology Certification Program 
(ESTCP). 
Sara, M. (2003). Site assessment and remediation handbook. Florida: CRC Press. 
Schwarzenbach, R.P.; Gashwend, P.M.; Imboden, D.M. (1993). Environmental 
organic chemistry. New York: John Wiley and Sons Inc. 
Schwille, F. (1988). Dense chlorinated Solvents in porous and fractured media - 
model experiments. (J. Pankow, Trans.) Florida: Lewis Publishers. 
SERDP and ESTCP. (2001, August). Expert panel workshop on research and 
development needs for cleanup of chlorinated solvent sites. Retrieved March 
3, 2008, from http://www.clu-in.org/products/tins/tinsone.cfm?num=29432415 
SERDP and ESTCP. (2006, September). Expert panel workshop on reducing the 
uncertainty of DNAPL source zone remediation. Retrieved March 3, 2008, 
from http://www.serdp.org/Featured-Initiatives/Cleanup-Initiatives/DNAPL-
Source-Zones 
Shouakar-Stash, O.; Frape, S.K.; Drimmie, R.J. (2003). Stable hydrogen, carbon and 
chlorine isotope measurements of selected chlorinated organic solvents. 
Journal of Contaminant Hydrology, 60, 211-228. 
Slater, G.F.; Dempster, H.D.; Sherwood Lollar, B.; Ahad, J. (1999). Headspace 
analysis: a new application for isotopic characterization of dissolved organic 
contaminants. Environmental Science and Technology, 33(1), 190-194. 
Slater, G.F.; Dempster, H.D.; Sherwood Lollar, B.; Spivack, J.; Brennan, M.; 
Mackenzie, P. (1998). Isotope tracers of degradation of dissolved chlorinated 
solvents. First International Battelle Conference on Remediation of 
202 
Chlorinated and Recalcitrant Compounds. Monterey, California: Battelle 
Press. 
Snow, D.T. (1968). Rock fracture spacings, openings, and porosities. Journal of the 
Soil Mechanics and Foundations Division, 94(1), 73–91. 
Soesilo, J.A.; Wilson, S.R. (1997). Site remediation - planning and management.
Florida: CRC Press. 
Steele, A.; Lerner, D. N. (2001). Predictive modelling of NAPL injection tests in 
variable aperture spatially correlated fractures. Journal of Contaminant 
Hydrology, 49(3-4), 287-310. 
Stiber, N.A.; Pantazidou, M.; Small, M.J. (1999). Expert system methodology for 
evaluating reductive dechlorination at TCE sites. Environmental Science and 
Technology, 33(17), 3012-3020. 
Stroo, H.F.; Leeson, A.; Shepard, A.J.; Koenigsberg, S.S.; Casey, C.C. (2006). 
Monitored natural attenuation forum: environmental remediation applications 
of molecular biological tools. Remediation Journal, 16(2), 125-137. 
Sturchio, N. C.; Clausen, J. C.; Heraty, L. J.; Huan, g L., Holt, B. D.; Abrajano, T. 
(1998). Stable chlorine isotope investigation of natural attenuation of 
trichloroethene in an aerobic aquifer. Environmental Science and Technology, 
32, 3037-3042. 
Sturchio, N.C.; Bohlke, J.K.; Beloso, A.D.; Streger, S.H.; Heraty, L.J.; Hatzinger, P.B. 
(2007). Oxygen and chlorine isotope fractionation during perchlorate 
biodegrtadation: laboratory results and implications for forensic and natural 
attenuation studies. Environmental Science and Technology, 41(18), 2796-
2802. 
Sudicky, E.A.; Gillham, R.W.; Frind, E.O. (1985). Experimental investigations of 
solute transport in stratified porous media 1) The non reactive case. Water 
Resource Research, 21(7), 1035-1041. 
203 
Taylor, A.E.; Dolan, M.E.; Bottomley, P.J.; Semprini, L. (2007). Utilization of 
fluoroethene as a surrogate for aerobic vinyl chloride transformation. 
Environmental Science and Technology, 41(18), 6378-6383. 
Thomas, G.W. (1982). Principles of hydrocarbon reservoir simulation. Boston: 
International Human Resources Development Corporation. 
Tsang, C.F.; Tsang, Y.W.; Hale, F.V. (1991). Tracer transport in fractures: analysis of 
field data based on a variable-aperture channel model. Water Resources 
Research, 27, 3095-3106. 
United States Environmental Protection Agency. (1980). Sources of toxic compounds 
in household wastewater, EPA 600/2-80-123. Cincinnatti: Office of Research 
and Development. Retrieved from http://nepis.epa.gov 
United States Environmental Protection Agency. (1990). Handbook: Groundwater 
and Contamination Volume 1, EPA 625/6-90/016a. Cincinnati: Center for 
Environmental Research Information. 
United States Environmental Protection Agency. (1992, January). Estimating 
potential for occurrence of DNAPL at Superfund Sites, 9355.4-07FS. Office of 
Solid Waste and Emergency Response. Washington D.C.: Office of Solid 
Waste and Emergency Response. Retrieved April 18, 2008, from 
http://www.epa.gov/superfund/health/conmedia/gwdocs/pdfs/estdnapl.pdf 
United States Environmental Protection Agency. (1993). Evaluation of the likelihood 
of DNAPL presence at NPL Sites: national results, EPA 540R-93-073. Office 
of Solid Waste and Emergency Response. Washington D.C.: Office of Solid 
Waste and Emergency Response. 
United States Environmental Protection Agency. (1994). Guidance for the data quality 
objectives process, EPA/600/R-96/055. Washington D.C.: Office of Research 
and Development. 
United States Environmental Protection Agency. (1998). Technical protocol for 
evaluating natural attenuation of chlorinated solvents in ground water, 
EPA/600/R-98/128. Washinton D.C.: Office of Research and Development.  
204 
United States Environmental Protection Agency. (1999, April). Use of monitored 
natural attentuation at Superfund, RCRA corrective action and underground 
storage tank sites, Directive Number: 9200.4-17P. Office of Solid Waste and 
Emergency Response (OSWER). Washington D.C.: Office of Solid Waste and 
Emergency Response. Retrieved October 5, 2009, from 
http://www.epa.gov/oust/directiv/d9200417.pdf 
United States Environmental Protection Agency. (2001, July). The state-of-the-
practice of characterization and remediation of contaminated ground water at 
fractured rock sites, EPA542-R-01-010. Washington D.C.: Office of Solid 
Waste and Emergency Response. Retrieved October 12, 2009, from 
www.epa.gov: http://epa.gov/tio/download/misc/fracrock_state.pdf 
United States Environmental Protection Agency. (2003, December). The DNAPL 
remediation challenge: Is there a case for source depletion?, EPA/600/R-
03/143. Cincinnati: National Risk Management Research Laboratory. 
Retrieved May 26, 2011, from www.epa.gov: 
http://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=300061GP.PDF 
United States Environmental Protection Agency. (2004). Site Characterization 
Technologies for DNAPL Investigations, EPA 542-R-04-017. United States 
Environmental Protection Agency. Cincinnati: National Service Center for 
Environmental Publications.   
United States Environmental Protection Agency. (2011). Estimation Programs 
Interface Suite™ for Microsoft® Windows, v 4.10. Washington, D.C.: United 
States Environmental Protection Agency. 
Usher, B.H.; Pretorius, J.A.; Gebrekristos, R.A. (2008). Guidelines for the acceptance 
of monitored natural attentuation processes in South Africa. Pretoria: Water 
Research Commission. 
Usher, B.H.; Pretorius, J.A.; Dennis, I.; Jovanovic, N.; Clarke, S.; Cave, L.; Titus, R. 
(2004). Identification and prioritisation of groundwater contaminants in South 
Africa's urban catchments, Report Number: 1326/1/04. Pretoria: Water 
research Commission. 
205 
van Wamerdam, E.M.; Frape, S.K.; Aravena, R.; Drimmie, R.J.; Flatt, H.; Cherry, J.A. 
(1995). Stable chlorine and carbon isotope measurement of selected 
chlorianted organic solvents. Applied Geochemistry, 10, 547-552. 
Vogel, T.M.; McCarty, P.L. (1987). Abiotic and biotic transformations of 1,1,1-
trichloroethane under methanogenic conditions. Environmental Science and 
Technology, 21, 1208-1213. 
Vogel, T.M.; Criddle, C.S.; McCarty, P.L.. (1987). Transformations of halogenated 
aliphatic compounds. Environmental Science and Technology, 21(8), 722-
736. 
Wang, J.S.Y.; Narashiman, T.N. (1985). Hydrologic mechanisms governing fluid flow 
in partially saturated, fractured, porous medium. Water Resources Research, 
21, 1861-1874. 
Wang, Y.; Huang, Y. (2003). Hydrogen isotopic fractionation of petroleum 
hydrocarbons during vaporization: implications for assessing artificial and 
natural remediation of petroleum contamination. Applied Geochmistry, 18(10), 
1641-1651. 
Wiedemeier, T.H.; Rifai, H.S.; Newell, C.J.; Wilson, J.T. (1999). Natural attenuation of 
fuel hydrocarbons and chlorinated solvents. New York: John Wiley and Sons. 
Wilson, J.L.; Conrad, S. H.; Mason, W.R.; Peplinski, W.; Hafgan, E. (1990). 
Laboratory investigations of residual liquid organics from spills, leaks, and the 
disposal of hazardous wastes in groundwater. Oklahoma: United Stated 
Environmental Protection Agency. 
Woodford, A.C.; Chevallier, L. (2002). Hydrogeology of the Main Karoo Basin: current 
knowledge and future reserach needs. Pretoria: Water Research 
Commission. 
Yang, Y.; McCarty, P.L. (2000). Biologically enhanced dissolution of 
tetrachloroethene DNAPL. Environmental Science and Technology, 34(14), 
2979-2984. 
206 
Yang, Y.; McCarty, P.L. (2002). Comparison between donor sustrates for biologically 
enhanced tetrachloroethene DNAPL dissolution. Environmental Science and 
Technology, 36, 3400-3404. 
Zhang, Z.F.; Smith, J.E. (2002). Visualization of DNAPL fingering processes and 
mechanisms in water-saturated porous media. Transport in Porous Media, 48, 
41-59. 
Zogorski, J.S.; Carter, J.M.; Ivahnenko, T.; Lapham, W.W.; Moran, M.J.; Rowe, B.L.; 
Squillace, P.J.; Toccalino, P.L. (2006). The quality of our Nation’s waters—
Volatile organic compounds in the Nation’s ground water and drinking-water 
supply wells. United States Geological Survey. 
207 
ABSTRACT 
The remediation of sites contaminated by dense non-aqueous phase liquids 
(DNAPLs) continues to present a significant environmental challenge globally. 
Contributing to this challenge is the difficulty in locating source zones due to local 
heterogeneities in the sub-surface. Heterogeneities are significant in fracture rock 
environments, such as those found in South Africa, which together with the fluid 
properties determine the fate and transport of DNAPLs.  
This research is based on evaluating the effectiveness of combining traditional and 
novel source zone characterisation methodologies in order to delineate chlorinated 
hydrocarbon DNAPLs in a fractured rock environment. The research documents and 
evaluates the characterisation process followed in the application of various 
methodologies to an Investigation Site in South Africa. A site-specific conceptual site 
model is presented indicating the delineation of the multiple chlorinated hydrocarbon 
DNAPL source zones at the site. Additionally, a DNAPL source characterisation 
approach is proposed for application in fractured rock environments. This approach 
allows for the convergence of traditional approaches (such as drilling within a fixed 
grid) with more novel approaches (such as high resolution sampling and analysis).  
The pioneering use of ribbon NAPL samplers (FLUTeTM activated carbon technology 
membranes) in South Africa is documented in this research. In situ source zone 
characterisation using this technology in a fractured rock environment is shown to be 
successful in determining depth discrete fracture transmissivities and residual 
DNAPL zones that would have gone unobserved through methods such as direct 
observation and testing rock cores with hydrophobic dyes.  The efficiency of this 
technology renders it ideal for future continued use in South Africa.  
208 
OPSOMMING 
Die remediëring van terreine wat besoedel is met digte vloeistowwe in ’n nie-
akwatiese fase (DNAPL’s), bied steeds wêreldwyd ’n wesentlike 
omgewingsuitdaging. Die problematiese bepaling van bronsones weens die lokale 
heterogeniteite ondergronds dra verder by tot hierdie uitdaging. Heterogeniteite is 
betekenisvol in breukrotsomgewings soos wat in Suid-Afrika aangetref word, wat 
tesame met die vloeistofeienskappe, die lot en vervoer van DNAPL’s bepaal.  
Hierdie navorsing is gegrond op die evaluering van die doeltreffendheid van 
tradisionele en ongewone bronsone-karakteriseringsmetodologieë in kombinasie om 
chloorkoolwaterstof-DNAPL’s in ’n breukrotsomgewing te delinieer. Die navorsing 
dokumenteer die karakteriseringsproses wat in die toepassing van verskeie 
metodologieë met betrekking tot ’n Ondersoekterrein in Suid-Afrika gevolg is. ’n 
Terreinspesifieke konseptuele terreinmodel word aangebied wat die deliniasie van 
die veelvuldige chloorwaterstof-DNAPL-bronsones op die terrein toon.   
Daarbenewens word ’n DNAPL-bronkarakteriseringsbenadering voorgestel wat in 
breukrotsomgewings toegepas kan word. Hierdie benadering maak daarvoor 
voorsiening dat tradisionele benaderings (soos om in ’n vaste rooster te boor) en 
ongewone benaderings (soos hoëresolusie-monsterneming en analisering) 
samelopend gevolg kan word.  
Die baanbrekerswerk in die gebruik van NAPL-lintmonsters (FLUTeTM-geaktiveerde 
koolstoftegnologiemembrane) in Suid-Afrika word in hierdie navorsing 
gedokumenteer. Waar hierdie tegnologie in ’n breukrotsomgewing vir in situ
bronsone-karakterisering gebruik is, was dit suksesvol in die bepaling van 
dieptediskrete breuktransmissiwiteite en residuele DNAPL-sones wat nie deur 
metodes soos direkte observasie en die toets van rotskerne met hidrofobiese 
kleurstowwe waargeneem sou word nie.  Die doeltreffendheid van hierdie tegnologie 
maak dit ideaal vir voortgesette toekomstige gebruik in Suid-Afrika.  
209 
KEY WORDS 
Dense non-aqueous phase liquid (DNAPL) 
Fractured rock 
Source characterisation 
Chlorinated hydrocarbons 
Site characterisation 
Fracture network characteristics 
High resolution characterisation 
Source zones 
Conceptual site model 
Characterisation efficiency 
210