7.1.11.7.4 Redelinghuys JG 24071919

Design and application of a bench scale
evaporator for the treatment of coal

power plant waste water
JG Redelinghuys
orcid.org/0000-0002-2328-657X
Dissertation submitted in fulfilment of the requirements for the

degree Master of Engineering in Chemical Engineering at the
North-West University
Supervisor: Prof E Fosso-Kankeu
Graduation Ceremony: May 2019

Student number: 24071919
School of Chemical and Minerals Engineering
Declaration of Authorship
I declare that this report is a presentation of my own original work.
Whenever contributions of others are involved, every effort was made to indicate this clearly, with due
reference to the literature.
No part of this work has been submitted in the past, or is being submitted, for a degree or examination at
any other university or course.
Signed on this, 7th day of November in Potchefstroom.
JG Redelinghuys
_________________________________
INITIALS AND SURNAME
ii
Academic and Technical Outputs

Conference Proceedings
Redelinghuys J, Fosso-Kankeu E, Waanders F, Rogers D, Gericke G. 2017. Scaling conditions inside

improved bench-scale single effect vacuum evaporator. In Advances in Science, Engineering, Technology
& Waste Management (ASETWM-17) conference proceedings organized by EAP (Eminent Association of
Pioneers) and hosted in Parys, North-West Province, SA.
Fosso-Kankeu, E., Brider, C.E., Redelinghuys, J., Gericke, G., Lemmer, N. and Waanders, F., 2017.
Determination of Suitable Anti-Scaling Agent to Inhibit Scale Formation of Water Processed in Coal Power
Plant System. In Advances in Science, Engineering, Technology & Waste Management (ASETWM-17)
conference proceedings organized by EAP (Eminent Association of Pioneers) and hosted in Parys, North-
West Province, SA.
Fosso-Kankeu, E., Redelinghuys, J., Waanders, F., Rogers, D., Bruinsma, D. and Gericke, G., 2016.
Determination of water evaporation rate in an assembled bench scale MED and impact of anti-scaling
agents on the morphology of scale. In International Conference on Advances in Science, Engineering,
Technology and Natural Resources (ICASETNR-16) hosted in Parys, North-West Province, SA.
iii
Acknowledgements
First and foremost, I want to thank my Heavenly Father for granting me the opportunity of completing this
project and his protection throughout.
I would like to convey a special thanks and appreciation to my family, friends, and colleagues in the
engineering department for their love, loyal support, company, assistance and understanding.
I wish to direct a special thanks to the following people and institutions that were involved in this project.
I believe that without them, this project would not have been successful:
Prof Elvis Fosso–Kankeu (Faculty of Engineering, North-West University), for his guidance, advice,
support and patience as my study supervisor.
Mr Jan Kroeze and Mr Adrian Brock (Faculty of Engineering, North-West University – Workshop Senior
Technicians) for their guidance, support, advice, assistance and inputs.
Mr Gerhard Gericke (Eskom, Research and Innovation Centre, Rosherville, Johannesburg), for his
guidance, advice, support and insight into the in-depth functioning of power stations as my co-supervisor.
Prof FB Waanders (Faculty of Engineering, North-West University) for his guidance and support as co-
supervisor.
Mr Nico Lemmer (North-West University) and Mr Christopher Brider (North-West University) for their
support throughout the collection of the data and laboratory analysis.
ESKOM, for sponsoring and funding this investigation.
Mrs René Bekker and Mrs Sanet Botes (Financial Officer/Admin Assistant, Faculty of Engineering, North-
West University) for their patience and trouble.
iv
Abstract
The potential of scale formation during the application of Multi-Effect Distillation (MED) technology for
the treatment of Reverse Osmosis (RO) retentate was investigated in this study. An improved bench scale
single effect evaporation unit was designed using the Solidworks Simulation Package and appropriate
materials. Operating conditions were determined to be 80°C and 0.48 bar with the design constructed to
maximise evaporation. Once assembled, RO-retentate was introduced to the evaporator at an evaporation
rate of 4 L/h (22.3 L/m2/h), and the variation of the concentrations of scale forming elements determined.
The evaporation observed translates to 34.5% water recovery. Speciation using PHREEQ-C modelling
software showed that the concentrations of scale forming ions increased during evaporation. Scale
prediction and prevention were explored using French Creek software, with anti-scaling agents being used
to minimise scaling.
The dosage of the anti-scaling agents inside the MED was determined using French Creek software. The
anti-scaling agents yielded confirmation of their capability of keeping ions in suspension and preventing
them from forming scale. When measuring the performance of the anti-scaling agents on this basis, AMPS
(Acrylic Acid-2-Acrylamide-2-Methylpopane Sulfonic Acid Copolymer) fared better than PBTC (2-
Phosphonobutane-1,2,4-Tricarboxylic Acid). However, HEDP (1-Hydroxy ethylidene-1,1-Diphosphonic
Acid) yielded concentrations of scale forming ions lower than when no anti-scaling agent was dosed. This
may be ascribed to the nature of complexes that HEDP forms with the scale forming ions.
The application of MED technology combined with the use of anti-scaling agents to treat inland waste
waters can see the MED evaporator serving a dual purpose. The primary purpose is to serve as water
producing/recycling technology that can treat saline solutions to recycle as much fresh water as possible.
The MED reject will be suitable for further treatment using other separation technology to ultimately
produce fresh water and salts. The secondary purpose of the MED technology when applied to a power
plant could be to dissipate some of the abundant heat present. The MED technology possesses the potential
of alleviating the amount of heat energy the cooling towers must dissipate to some extent, thus lowering
the amount of cooling water needed, and this should be investigated. Thus, a combination of membrane
technology and thermal technology could theoretically be utilised in order to adhere to the zero liquid
effluent discharge policy Eskom has, whilst lowering the amount of fresh water needed for cooling
purposes.
Additionally, the laboratory evaluation of anti-scaling agents prescribed by scale modelling software was
investigated. The considered water contained high levels of scale forming agents. The selected test method
to determine anti-scaling agent induction times was found to be inefficient and an improvement on the
current method was proposed. The possible causes for the inefficiency of the selected induction time test
method was mainly ascribed to an insufficient heating apparatus setup
Key Words: MED, anti-scaling agents, Reverse Osmosis retentate, French Creek
v
Table of Contents
Declaration of Authorship ........................................................................................................................... ii
Academic and Technical Outputs ............................................................................................................. iii
Acknowledgements ...................................................................................................................................... iv
Abstract ......................................................................................................................................................... v
Table of Contents......................................................................................................................................... vi
List of Figures .............................................................................................................................................. ix
List of Tables ................................................................................................................................................ xi
List of Abbreviations.................................................................................................................................. xii
Background and Motivation.............................................................................................. 1
1.1 Introduction .......................................................................................................................................... 1
1.2 Problem Statement ............................................................................................................................... 1
1.3 Aim and Objectives ............................................................................................................................. 1
1.3.1 Aim ............................................................................................................................................ 1
1.3.2 Objectives .................................................................................................................................. 2
1.4 Research Questions.............................................................................................................................. 2
1.5 Dissertation Structure .......................................................................................................................... 2
References...................................................................................................................................................... 4
Literature Review ............................................................................................................... 5
2.1 Introduction .......................................................................................................................................... 5
2.2 Mineral Scale Formation ..................................................................................................................... 5
2.3 Scale Formation Mechanisms ............................................................................................................. 5
2.3.1 Supersaturation/Clustering........................................................................................................ 6
2.3.2 Nucleation ................................................................................................................................. 7
2.3.3 Crystal growth ........................................................................................................................... 8
2.3.4 Deposition/agglomeration......................................................................................................... 8
2.3.5 Kinetics of Sulphate Removal through Gypsum Crystallization............................................. 8
2.4 Scale Inhibition Mechanisms .............................................................................................................. 8
2.4.1 Sequestration ........................................................................................................................... 11
2.4.2 Crystal distortion ..................................................................................................................... 11
2.4.3 Dispersion and surface conditioning ...................................................................................... 11
2.5 Characteristics of Anti-Scaling Agents ............................................................................................. 11
2.6 Principles of the Different Treatment Methods: ............................................................................... 14
2.6.1 Distillation ............................................................................................................................... 14
2.6.2 Eutectic freeze distillation....................................................................................................... 16
2.6.3 Membrane technology ............................................................................................................ 17
2.6.4 Comparison of different technologies .................................................................................... 19
2.7 Conclusion ......................................................................................................................................... 20
vi
References.................................................................................................................................................... 22
Design, Configuration and Construction of Bench Scale MED .................................. 26
3.1 Introduction ........................................................................................................................................ 27
3.2 Operating Conditions......................................................................................................................... 29
3.2.1 Temperature ............................................................................................................................ 29
3.2.2 Pressure ................................................................................................................................... 29
3.3 Materials............................................................................................................................................. 30
3.4 Design and Configuration.................................................................................................................. 31
3.4.1 Evaporation chamber .............................................................................................................. 31
3.4.2 Vacuum pumps ....................................................................................................................... 33
3.4.3 Drip tray .................................................................................................................................. 33
3.4.4 Operation ................................................................................................................................. 37
3.4.5 Method of evaluating MED performance .............................................................................. 37
3.5 Methodology ...................................................................................................................................... 37
3.5.1 Assembly ................................................................................................................................. 37
3.5.2 Water sampling and analysis .................................................................................................. 39
3.5.3 Operation and control of MED ............................................................................................... 39
3.5.4 Determination of anti-scaling agents by manual calculations ............................................... 39
3.5.5 Determination of anti-scaling agents with modelling software ............................................. 40
3.5.6 Anti-scaling agent dosage ....................................................................................................... 40
3.6 Results ................................................................................................................................................ 42
3.6.1 Performance of MED .............................................................................................................. 42
3.7 Conclusion and Recommendations ................................................................................................... 50
References.................................................................................................................................................... 51
Determination of Anti-Scaling Agent to be used inside Bench Scale MED ............... 53
4.1 Introduction ........................................................................................................................................ 54
4.2 Materials and Methods ...................................................................................................................... 55
4.2.1 Sampling and sample storage ................................................................................................. 55
4.2.2 Synthetic control water preparation........................................................................................ 55
4.2.3 Spiked sample preparation...................................................................................................... 55
4.2.4 ICP-OES analysis.................................................................................................................... 56
4.2.5 EDTA titration ........................................................................................................................ 56
4.2.6 Alkalinity test .......................................................................................................................... 57
4.2.7 Sulphate content in samples.................................................................................................... 57
4.2.8 Scale potential modelling........................................................................................................ 57
4.2.9 Jar heating system ................................................................................................................... 57
4.2.10 Data logging method ............................................................................................................... 58
4.2.11 Static induction test procedure................................................................................................ 59
4.3 Results and Discussion ...................................................................................................................... 60
4.3.1 Sample pre-analysis ................................................................................................................ 60
vii
4.3.2 Bicarbonate and carbonate content from alkalinity test ......................................................... 61

4.3.3 Sulphate content ...................................................................................................................... 61
4.3.4 Scaling potential model .......................................................................................................... 61
4.3.5 Dosage profile modelling for ATMP and PBTC anti-scaling agents .................................... 64
4.3.6 Static induction test: ................................................................................................................ 65
4.3.7 Static induction test at 25⁰C .................................................................................................... 66
4.3.10 General discussion of possible causes of results .................................................................... 81
4.4 Conclusions ........................................................................................................................................ 81
References.................................................................................................................................................... 82
Conclusions and Recommendations ............................................................................... 84
5.1 Conclusions ........................................................................................................................................ 84
5.2 Recommendations ............................................................................................................................. 85
Appendix A: MED Designs ....................................................................................................................... 87
Appendix B: Academic and Technical Outputs ..................................................................................... 92
viii
List of Figures
Figure 2-1: Schematic of processes involved with the formation of mineral scaling and
scaling control by chemical anti-scaling agents ............................................................6
Figure 2-2: Principle of horizontal tube evaporation .............................................................................15
Figure 2-3: MED schematic ...................................................................................................................15
Figure 2-4: Multi-stage flash distillation (MSF) schematic diagram .....................................................16
Figure 2-5: Nano-filtration schematic ....................................................................................................17
Figure 2-6: Reverse osmosis process diagram .......................................................................................18
Figure 2-7: Schematic illustration of forward osmosis process .............................................................19
Figure 3-1: Schematic of processes involved with the formation of mineral .........................................29
Figure 3-2: Main components in MED design: (a) evaporation vessel; (b) recycle pump; (c)
drip-tray; (d) heat source/exchange surface ..................................................................31
Figure 3-3: Three-dimensional visualization of improved evaporation chamber design .......................31
Figure 3-4: 0.6 MPa (6 bar) pressure difference applied to design in simulation program -
visualization ..................................................................................................................32
Figure 3-5: Solidworks simulation displaying a maximum displacement of 4 mm when 6 bar
pressure difference is applied to evaporation chamber .................................................32
Figure 3-6: Flow contraction after an orifice with area A ......................................................................34
Figure 3-7: Inter tube falling-film modes: (A) droplet mode, (B) jet mode, (C) sheet mode .................35
Figure 3-8: Drip tray prior to installation inside evaporation chamber ..................................................36
Figure 3-9: 3-D view of drip tray ...........................................................................................................36
Figure 3-10: View of aperture sizing and spacing of drip tray...............................................................36
Figure 3-11: Milled polycarbonate .........................................................................................................37
Figure 3-12: Gluing of polycarbonate to form evaporation chamber.....................................................38
Figure 3-13: Evaporation chamber assembly .........................................................................................38
Figure 3-14: AMP dosage profile ...........................................................................................................41
Figure 3-15: HEDP dosage profile .........................................................................................................41
Figure 3-16: PBTC dosage profile .........................................................................................................42
Figure 3-17: Temperature gauge displaying the temperature of steam entering the
evaporation chamber inside the heat exchange piping .................................................42
Figure 3-18: Vacuum gauge displaying vacuum inside evaporation chamber .......................................43
Figure 3-19: Temperature gauge displaying temperature inside evaporation chamber .........................43
Figure 3-20: Assembled, functioning single effect bench scale evaporator ...........................................44
Figure 3-21: Drip tray's laminar water flow distribution onto heat exchange surfaces ..........................45
Figure 3-22: Aragonite saturation level vs. temperature vs. pH .............................................................47
Figure 3-23: Calcite saturaton level vs. temperature vs. pH ..................................................................47
ix
Figure 3-24: Langelier saturation index .................................................................................................48

Figure 4-1: Jar test apparatus with tray ..................................................................................................58
Figure 4-2: Jar heating system................................................................................................................58
Figure 4-3: French Creek MineSat® model, degree of supersaturation for CW....................................62
Figure 4-4: French Creek MineSat® model, degree of supersaturation for RO ....................................63
Figure 4-5: French Creek MineSat® model, dosage profile of ATMP for RO, calculated at
75⁰C ..............................................................................................................................65
Figure 4-6: French Creek MineSat® model, dosage profile of PBTC for RO, calculated at
75⁰C ..............................................................................................................................65
Figure 4-7: Conductivity over time (control solution 25°C) ..................................................................66
Figure 4-8: Conductivity over time (CW sample 25°C) ........................................................................67
Figure 4-9: Conductivity over time (RO sample 25°C) .........................................................................68
Figure 4-10: Conductivity over time (CW spiked sample 25°C) ...........................................................69
Figure 4-11: Conductivity over time (RO spike sample 25°C) ..............................................................69
Figure 4-12: Conductivity over time (control solution 40°C) ................................................................71
Figure 4-13: Conductivity over time (CW sample 40°C) ......................................................................72
Figure 4-14: Conductivity over time (RO sample 40°C) .......................................................................72
Figure 4-15: Conductivity over time (CW spike sample 40°C) .............................................................73
Figure 4-17: Conductivity over time (control solution 75°C) ................................................................76
Figure 4-18: Conductivity over time (CW sample 75°C) ......................................................................77
Figure 4-19: Conductivity over time (RO sample 75°C) .......................................................................78
Figure 4-20: Conductivity over time (CW spike sample 75°C) .............................................................78
x
List of Tables
Table 2-1: Chemical, physical and biological methods and their effectiveness .....................................10
Table 2-2: Examples of the effect that certain compounds have on calcium carbonate scaling ............12
Table 3-1: Operating parameter targets of evaporation chamber ...........................................................29
Table 3-2: Tensile strength of polycarbonate and the effect of temperature ..........................................33
Table 3-3: Ranges for the film patterns and corresponding flow regimes .............................................35
Table 3-4: Anti-scaling agent active ingredient dosages ........................................................................40
Table 3-5: Concentration in mg/L of a) RO-retentate before MED-evaporation and b) after
MED-evaporation .........................................................................................................45
Table 3-6: Speciation of certain elements present in MED reject ..........................................................46
Table 3-7: PPM concentration of certain elements prior and post MED evaporation, with and
without anti-scaling agents ...........................................................................................49
Table 4-1: Table displaying conductivities and pH's of CW-blowdown and RO-retentate ...................55
Table 4-2: Concentration of metals contained in samples ......................................................................60
Table 4-3: Hardness of water samples....................................................................................................61
Table 4-4: Scaling potential and degree of supersaturation for CaCO3 for CW-blowdown
and RO-reject streams ..................................................................................................64
Table 4-5: Initial and final measurements of pH and actual conductivity at 25⁰C .................................70
xi
List of Abbreviations
ATMP Amino Trimethylene Phosphonic Acid
EDTA Ethylene-diamine-tetra-acetic acid
CW Cooling water
PAA Polyacrylic acids
PASP Polyaspartic acid
AMPS Acrylic Acid-2-Acrylamide-2-Methylpopane Sulfonic Acid Copolymer
HEDP 1-Hydroxy ethylidene-1,1-Diphosphonic Acid
PBTC 2-Phosphonobutane-1,2,4-Tricarboxylic Acid
PESA Polyepoxysuccinic acid
PMA Polymaleic acids
RO Reverse osmosis
MED Multi-Effect Distillation
xii
Background and Motivation
1.1 Introduction
More than half of countries worldwide potentially face water shortages in the future. Projections indicate
that, by 2050, up to 75% of the world’s population may be confronted with water scarcity (Afrasiabi &
Shahbazali, 2011). South Africa has already been deemed a water-scarce country. All water consumers are
forced to explore the utilisation of all available water sources, and this water scarcity is further being
affected by climate change. Despite the water shortage the country is currently facing, the demand for water
usage in industry and human consumption remains high, creating an ever-increasing dependence on
seasonal rainfall (Fosso-Kankeu et al., 2009). Waste water effluent is becoming more valuable as advances
in technology makes the recovery, treatment and re-use of waste water more attractive and viable (Fosso-
Kankeu et al., 2016).
1.2 Problem Statement
Eskom, South Africa’s national power utility, is allocated 2-3% of the nationally available fresh water to
generate electricity. Eskom is under pressure to minimise its water usage and has set a target of lowering
its usage to 1.34 L of water per kilowatt hour (kWh) electricity produced by 2020, from the 1.39 L/kWh in
2017 (Roos, Rogers, & Gericke, 2017). Of the total water intake, 98% is lost via evaporation through the
cooling towers at the power plants. The re-usage and recovery of certain streams may potentially lower the
amount of fresh water intake needed to replenish the amount of water lost due to evaporation. Furthermore,
Eskom has a zero liquid effluent discharge (ZLED) clause in its water use licences – leading to the
exploration of lowering the amount of effluent waste water produced and improved disposal methods.
Eskom has thus initiated various investigations towards water management and proper usage of all available
water sources. This investigation will primarily focus on retentate/saline water stream from the primary
reverse osmosis (RO) module. This stream has a salinity above 8000 parts per million (ppm) and
recirculation through the RO system is not economically viable as the balance between optimal membrane
performance, maintenance and cost dictates. Yet, there remains the potential to recover water from this
retentate.
1.3 Aim and Objectives
1.3.1 Aim
The aim of this study was to determine the effectiveness of an evaporator system as a possible water
treatment technology for saline waste water streams in Eskom power plants.
1
1.3.2 Objectives
This aim was achieved by the following objectives:
i. Design, construct and operate a functioning bench scale single effect evaporator.
ii. Determine the most suitable anti-scaling agent required to use inside the evaporator when treating
a saline waste water stream obtained from an Eskom power plant.
1.4 Research Questions
The research was aimed at addressing the following questions:
i. What are the optimal operating conditions when applying evaporation technology to saline waste
water streams produced inside Eskom’s power plants?
ii. What anti-scaling agents should be used to optimise the heat exchange efficiency inside the
constructed bench scale single effect evaporator?
iii. How much water can be reclaimed when using evaporation technology to treat saline waste water
produced inside Eskom’s power plants?
iv. What is the composition of the bench scale single effect evaporator’s reject?
v. How well does anti-scaling agents work on different saline waste water streams obtained from an
Eskom power plant?
1.5 Dissertation Structure
The research presented in this dissertation has been conducted at the North-West University, in the
laboratory and in the workshop. Three articles were written for this dissertation, and they are presented to
conform to the requirements of the North-West University (Potchefstroom Campus) for the degree Master
of Engineering. This dissertation comprises 5 chapters; with a summary of each chapter to follow:
CHAPTER 1: Background and Motivation
This chapter provides a background into the reasoning and motivation for the study. The problem statement,
objectives and research questions were briefly discussed.
CHAPTER 2: Literature Review
This chapter discusses the literature relevant to the study and examines information from other studies under
the same research topic. It includes the in-depth examination of scale forming and inhibition mechanisms,
the discussion and comparison of relevant saline waste water treatment technology and the proposed
2
theoretical optimal combination of water treatment technological solution for inland coal powered, wet-
cooled cooling tower power plants.
CHAPTER 3: Design, configuration and construction of the bench scale MED
This chapter contains the design, configuration and construction of the bench scale single effect evaporator,
the determination of relevant anti-scaling agents, as well as the evaluation of the performance of the
evaporator.
CHAPTER 4: Determination of suitable anti-scaling agent to inhibit scale formation inside the constructed
bench scale MED.
This chapter contains the evaluation of the performance of anti-scaling agents applied to waste water
streams from Eskom’s power plants at different temperatures.
CHAPTER 5: Conclusions and Recommendations
This chapter contains the conclusions and recommendations to benefit the relevant stakeholders.
3
References
Afrasiabi, N. & Shahbazali, E. 2011. RO brine treatment and disposal methods. Desalination and Water
Treatment, 35: 39-53.
Fosso-Kankeu, E., Redelinghuys, J.G., Gericke, G., Waanders, F., Rogers, D. & Bruinsma, D. 2016.
Determination of Water Evaporation Rate in an Assembled Bench scale MED and Impact of Anti-Scaling
Agents on the Morphology of Scale. Parys, International Conference on Advances in Science, Engineering,
Technology and Natural Resources (ICASETNR-16) Nov. 24-25, 2016.
Fosso-Kankeu, E., Mulaba-Bafubiandi, A., Mamba, B. & Barnard, T. 2009. Mitigtion of Ca, Fe, and Mg
loads in surface waters around mining areas using indigenous micro-organism strains. Journal of Physics
and Chemistry of the Earth, 34: 825-829.
Roos, T.H., Rogers, D.E.C. & Gericke, G. 2017. Solar-assisted MED treatment of Eskom power station
waste water. s.l. AIP Publishing.
4
Literature Review
2.1 Introduction
Coal-fired power stations that use wet-cooled cooling towers produce large amounts of mineral rich waste
water streams due to the vast amounts of evaporation during the cooling process. The evaporation of the
cooling water concentrates the salts and minerals found in the fresh water entering the power plant. This
water is treated on site to recover as much water as possible, and various water streams are treated using
membrane technologies (nano-filtration and reverse osmosis). When water is being treated for re-use,
scaling is common when any form of heat exchange or membrane technology is used if precautions are not
taken (MacAdam & Parsons, 2004). This is due to the increase in mineral concentration. The negative
effects of scaling can take various forms, including fouling of membrane systems, reduced heat transfer
efficiencies in heat exchange systems, scheduled and unscheduled shutdowns, as well as increased
maintenance, operation and capital costs (Piyadasa et al., 2017). Saline water can be purified or recycled
by making use of various desalination techniques, namely distillation (thermal and crystallisation
distillation) and non-thermal membrane techniques (nano-filtration and reverse osmosis) (Amjad, 1996).
2.2 Mineral Scale Formation
Scale forms in alkaline water and is caused by the presence of Ca2+ and/or Mg2+ and other ions of the
solution, with calcium carbonate being one of the more common scale components (Zhang et al., 2001).
Sub-micron size crystals are the main contributors to the scale layer on a metal surface during the initial
stages of scale formation. Supersaturation of the water can be described as the main cause of scale formation
(Chen et al., 2005). Process parameters also plays a role in scale formation – with temperature and flow
rate being the most influential (Wildebrand et al., 2007). Other factors include the deposit (scale) structures,
shape, size and composition. It is important to note that, when mineral constituents remain soluble, they
have no negative effect on heat transfer efficiency (Kavitha et al., 2011).
2.3 Scale Formation Mechanisms
When assessing the fundamental processes involved in the formation of scale and mineral precipitation, it
becomes apparent that there are multiple steps involved, and these will be briefly explained with the
schematic representation in Figure 2-1.
5
Figure 2-1: Schematic of processes involved with the formation of mineral scaling and scaling control by
chemical anti-scaling agents
Reproduced from Antony et al. (2011)
When mineral ions are introduced to water, either artificially or through the continual concentrating of the
ions present by decreasing the water volume, the ions will eventually become sufficiently concentrated to
deem the water supersaturated. Mineral ions in solution under increasingly supersaturated conditions will,
due to their polarity and random thermal motion, form clusters (Al-Roomi & Hussain, 2016). Once these
clusters pass a critical size, nucleation will take place. As soon as nuclei are formed, spontaneous
crystallisation will occur, forming larger particles. Subject to the surface properties and composition of the
particles, they can either aggregate in order to create larger particles or remain suspended in the water for
a period of time, after which these particles and present mineral ions will diffuse through the hydraulic
boundary layer (Li, 2010) and form scale after depositing onto a solid surface.
2.3.1 Supersaturation/Clustering
Water containing ions of sparingly soluble salts, such as CaCO3, CaSO4 or Mg(OH)2, has the potential to
become supersaturated and/or allow mineral clustering to form under certain process conditions.
Supersaturated conditions are achieved when the solution containing these salts is concentrated beyond the
solubility limits of one or more of the constituents. This can occur by either separation of pure water
(membrane processes such as RO) or water evaporation (usually present in thermal desalination systems or
water cooling towers) (Hasson et al., 2011).
6
Supersaturated conditions can also be achieved by a change in temperature. The majority of generally
encountered scaling salts exhibit inverse solubility characteristics with regard to temperature. When a
solution containing an inverted-solubility salt comes into contact with a hot surface, supersaturation by this
inverse-solubility effect can cause scale to deposit on the hot surface, even when the bulk of the solution is
undersaturated (Hasson et al., 2011).
2.3.2 Nucleation
Once supersaturation conditions are achieved, the mineral ions present continuously adopt different
configurations for solid mineral formation (Yang et al., 2002). The clustering that occurs can be deemed
random, as it is caused by ionic thermal motion that is easily reversible. Mineral clusters consist of a broad
size range, starting at dimers, trimers and continuing to i-mers (i>100, where i is the number of ions) and
stopping at the critical number, which can be denoted as number j. These clusters constantly form, break
up and disappear. Once a cluster reaches the critical j-number/size, the addition of a monomer leads to
nucleation. The formation of clustering centres from which spontaneous crystal growth can proceed is also
referred to as nucleation (Li, 2010).
To summarise classic nucleation theory, the free energy of the formation of a nucleus consists mainly of
the energy gained from creating bonds paired with the work required to create the nucleus surface (Stumn
et al., 1992).
𝛥𝐺𝑗 = 𝛥𝐺𝑏𝑢𝑙𝑘 + 𝛥𝐺𝑠𝑢𝑟𝑓𝑎𝑐𝑒 Equation 2-1
Equation 2-1
Bulk energy gain for a nucleus can be expressed as follows:
𝛥𝐺𝑏𝑢𝑙𝑘 = −𝑗𝑘𝑇ln𝑆 Equation 2-2
With 𝑗 being the number of monomers (ionic units) in the nucleus. “ 𝑗 ” can be expressed using volume
where 𝑉𝑖𝑜𝑛 is the ionic volume (m3) and 𝑟 being the radius of a spherical nucleus:
4𝜋𝑟3
( )
3
𝑗= 𝑉𝑖𝑜𝑛
Equation 2-3
𝑆 which is the saturation ratio can be expressed as:
1
𝐼𝐴𝑃 𝑛
𝑆= ( ) Equation 2-4
𝐾𝑠𝑝
Where 𝐼𝐴𝑃 represents the ionic activity product, 𝐾𝑠𝑝 the solubility product and 𝑛 the number of ions. The
number of ions is obtained from the formula unit of the mineral (i.e., 𝑛 = 𝑣 + 𝑤 where mineral formula
unit is 𝐴𝑣𝐵𝑤).
7
The energy cost/work required during surface formation can be expressed using interfacial energy (surface
tension), which is assumed to be independent of nucleus size, has a spherical area, is measured in mJ/m2
and is represented by the symbol γ in the following equation:
𝛥𝐺𝑠𝑢𝑟𝑓𝑎𝑐𝑒 = 4𝜋𝑟 2 γ Equation 2-5
Combining the relevant equation results in the nucleation energy is written as follows:
4𝜋𝑟3 1
( ) 𝐼𝐴𝑃 𝑛
3
𝛥𝐺𝑗 = − 𝑘𝑇ln ( ) + 4𝜋𝑟 2 γ Equation 2-6
𝑉𝑖𝑜𝑛 𝐾𝑠𝑝
Nucleation energy can be expressed as a function of temperature, saturation ratio and nucleus size. It is
apparent that spontaneous nucleation will take place as soon as the solution is supersaturated and the size
of the ion clusters present in the solution is sufficient. Additionally, temperature increases result in
conditions favourable for nucleation.
2.3.3 Crystal growth
Upon the commencement of nucleation, microcrystals are present in the solution, and they can agglomerate
and/or absorb onto surfaces exposed to the solution where the micro-crystals will grow into larger micro-
crystals to eventually form macro-crystals by fusing together (Al-Roomi & Hussain, 2016). Dandekar and
Doherty (2014) reported that crystal growth will occur under non-equilibrium conditions when the chemical
potential of the growth medium (𝜇𝑚𝑒𝑑𝑖𝑢𝑚 ) is more than the chemical potential of the present bulk crystal
(𝜇𝑐𝑟𝑦𝑠𝑡𝑎𝑙 ). The difference between these two chemical potentials is the impelling cause for crystal growth
(Dandekar & Doherty, 2014).
∆𝜇 = 𝜇𝑚𝑒𝑑𝑖𝑢𝑚 − 𝜇𝑐𝑟𝑦𝑠𝑡𝑎𝑙 Equation 2-7
2.3.4 Deposition/agglomeration
According to Al-Roomi and Hussain (2016), the adsorption of additional ions from the solution will in turn
cause the macro-crystals to continue growing until their size deems them colloidal particles, and ultimately
form a scale film onto any susceptible surface, growing into a deposit. It is important to note, however, that
should there be an availability of active sites, scale forming species can bypass the crystal forming stage
and directly proceed to crystal growth onto these sites (Ferguson et al., 2013).
2.3.5 Kinetics of Sulphate Removal through Gypsum Crystallization
A solution will be deemed over saturated pertaining to gypsum in two cases (Maree, et al., 2004):
8
i) When an effluent rich in MgSO4 is treated with lime in order to precipitate Mg2+ as Mg(OH)2, an over
saturated CaSO4 solution will form. (This typically happens if [SO42+] is greater than 1500 mg/L).
ii) When an acid solution with a high enough sulphate concentration is treated by dosing lime so much so
that the solution is over saturated with respect to gypsum after neutralization.
The sulphate removal rate by gypsum crystallization can be predicted by equation 2 – 8 (Geldenhuys, et
al., 2003):
𝑑[𝐶𝑎𝑆𝑂4 .2𝐻2 𝑂]
𝑑𝑡
= 𝑘[𝐶𝑎𝑆𝑂4 . 2𝐻2 𝑂](𝑆)[𝐶 − 𝐶𝑂]2 Equation 2-8
𝑑[𝐶𝑎𝑆𝑂4 .2𝐻2 𝑂]
Where: 𝑑𝑡
is the crystallization rate; 𝑘 represents the reaction rate constant, [𝐶𝑎𝑆𝑂4 . 2𝐻2 𝑂](𝑆)
represents the surface area of the seed crystals, 𝐶 represents the initial calcium sulphate concentration and
𝐶𝑂 represents the saturated calcium sulphate in solution concentration.
From this equation 2-8 it becomes apparent that the gypsum concentration and gypsum crystallization is
proportional, as an increased gypsum concentration will result in an increased rate of gypsum
crystallization(Geldenhuys, et al., 2003). This serves as motivation for sludge recirculation during lime
dosing stages when sulphate removal is required, or if sulphate is required to stay in solution – no
recirculation of sludge (Maree, et al., 2004).
When 𝐶𝑂2 is used for pH adjustment, the occurrence of 𝐶𝑎𝐶𝑂3 crystallization can be indicated by a
decrease in the concentration of calcuim (Maree, et al., 2004). The under saturation, super saturation and
saturation states of 𝐶𝑎𝐶𝑂3 can theoretically be identified by the activity product of 𝐶𝑎2+ and 𝐶𝑂32− species
that compares with the solubility product constant 𝑘𝑠𝑝 for 𝐶𝑎𝐶𝑂3 (Loewenthal & Marais, 1976). If above
the solubility product, 𝐶𝑎𝐶𝑂3 will precipitate from the solution. If below, 𝐶𝑎𝐶𝑂3 will dissolve. Equation
2-9 describes the precipitation or dissolution rate (Geldenhuys, et al., 2003):
1 2
1
𝑑[𝐶𝑎 2+ ] 2+ )2 2
− 𝑑𝑡
= 𝑘𝑆 {(𝐶𝑎 − 𝑘𝑠𝑝 } Equation 2-9
Where 𝑘 represents the precipitation rate constant and 𝑆 represents the surface area of the 𝐶𝑎𝐶𝑂3
growth/dissolution as well as the activity in mol/L
The precipitation rate depends on 𝑘, 𝑆 and the term inside the curled brackets. Although 𝑘 is affected by
crystal structure to an unknown extent, 𝑆 depends on the mass, size and structure of the crystals. By
controlling the term in curled brackets (the degree of supersaturation) as well as 𝑆 (the surface area of the
𝐶𝑎𝐶𝑂3 ) – the rate of dissolution or precipitation of 𝐶𝑎𝐶𝑂3 can be determined (Geldenhuys, et al., 2003).
9
2.4 Scale Inhibition Mechanisms
Physical, chemical and biological methods can be used to inhibit the formation and treat formed scale with
varying levels of effectiveness. Table 2-1 displays the effectiveness of various methods.
Table 2-1: Chemical, physical and biological methods and their effectiveness
Method Reported Effectiveness Mechanism
Ion Exchange Softening Complete scale control Removal of scale forming calcium
species
Acid Dosing Complete scale control (possible Alters calcium carbonate solubility
corrosion occurrence)
Chemical Inhibitors Complete scale control possible, Block growth sites, stops crystal
dosage type and water quality formation
dependent
Metal Ions 80% reduction Blockage of growth sites/reacts with
carbonate
Magnetic Conditioners 80% reduction None agreed on
Electronic Conditioners 40% reduction Preferential formation of particles in
bulk
Electrolytic 30% reduction Releases inhibitory metal ions into
solution
Ultrasound 65% reduction Induction of precipitation by changes
to local pH
Surface Modification 90% achievable Production of lower fouling surfaces
Reproduced from MacAdam and Parsons (2004)
The methods of control listed in Table 2-1 can be arranged into three categories. The first category includes
the methods that change and influence the growth mechanisms of the crystals, the second those that
influence the solubility, with the last being the methods that influence the potential of a surface to foul
(MacAdam & Parsons, 2004).
The addition of chemical inhibitors to potentially scaling waters is deemed one of the most effective
methods of controlling and inhibiting the formation of scale (MacAdam & Parsons, 2004). Anti-scaling
chemicals are capable of preventing the formation of scale in numerous ways. Three of the foremost
mechanisms used by the majority of common scale inhibitors, according to the Nalco Water Handbook, are
threshold inhibition, crystal modification and sequestration (Flynn, 2009).
With referral to the schematic in Figure 2-1, certain anti-scaling agents can sequester (or “complex with”)
cationic species with scale forming potential, thus raising the solubility of the potentially problematic
mineral ions (Li, 2010), thus preventing clustering in order to inhibit scale formation. Anti-scaling agents
can additionally act as crystal modifiers, and by modifying the crystallisation pathways, inhibit the growth
of mineral particles (Zhang et al., 2001). Furthermore, certain anti-scaling agents can act as dispersants by
10
introducing a steric and/or electronic repellence between present mineral particles, therefore inhibiting
aggregation and flocculation (Li, 2010). Lastly, certain anti-scaling agents can be used as a surface
conditioner to alter the deposition surface, resulting in the surface being unfavourable for scale formation.
A desirable anti-scaling agent should preferably make use of multiple of the above-named mechanisms to
effectively inhibit scaling (Amjad, 2002).
2.4.1 Sequestration
Various organic molecules have been reported to exhibit an ability to inhibit or retard mineral precipitation
from a solution when present at very low (ppm) dosages (Anklam & Firoozabadi, 2005). The addition of
these anti-scaling agents is equivalent to a supersaturation level reduction of the mineral solution (Oner et
al., 1998).
2.4.2 Crystal distortion
Growth inhibition can be accomplished by particle stabilisation (Amjad, 2002). Mineral colloids’ stability
can be achieved by increasing inter-particle repulsive forces (e.g. electrostatic and steric forces) or by
preserving them.
2.4.3 Dispersion and surface conditioning
Apart from the complexation action used in the sequestration and crystal distortion mechanisms, anti-
scaling agents can also interact with nuclei, unsettling the crystallisation process (Shakkthivel &
Vasudevan, 2006). This in turn implies the dispersion of crystal particles back into the aqueous solution,
making them less prone to adhesion onto available surfaces. Certain ions present in anti-scaling agents have
the ability to adsorb onto energetically favourable growth positions on the available surface to completely
inhibit further precipitation (Meyer, 1984).
2.5 Characteristics of Anti-Scaling Agents
When referring to anti-scaling agents, it must be noted that these include only the active ingredients that
form part of the anti-scaling agents. The manufacturers of anti-scaling agents conduct business in a secretive
and “patented” nature. Anti-scaling agents do not entirely eliminate the possibility of scaling to occur; they
do however (when correctly dosed) control and delay the formation of scale.
The most widely, commercially used anti-scaling agents are derived from three “chemical families,”
namely organophospates, condensed polyphosphates, and poly-electrolytes (or carboxylic polymers)
(MacAdam & Parsons, 2004). Polyphosphate anti-scaling agents have been recently replaced with polymers
11
such as poly-acrylic and -maelic acid, as these have been proven stable at high operating temperatures as
well as having an increased resistance to chemical and biological breakdown (Hasson et al., 2011).
The most notable advantage of using anti-scaling agents is that, when dosed correctly, very low doses can
subdue scale formation. These dosages are far below the stoichiometric concentration of the species that
will scale, thus making it a cost-effective solution (Hasson et al., 2011). The agents are generally composed
of organic compounds, and the techniques are adaptable as they offer more possibilities for efficient water
treatment over a wider range of parameters and technical specifications (Drela et al., 1998).
Table 2-2 details examples of the effects that specific compounds (organic and inorganic) can have on
calcium carbonate scaling (i.e. scale formation, precipitation and crystal growth):
Table 2-2: Examples of the effect that certain compounds have on calcium carbonate scaling
Inhibitor Effect Concentration Reference
Organic agents
Citric/Maleic Acid Effected crystal Morphology 0.1 – 1mg/l Meldrum and Hyde (2001)
Polymaelic Acid 56% precipitation inhibition 10mg/l Harris and Marshall (1981)
EDTA 20% precipitation inhibition 10 mg/l Harris and Marshall (1981)
Dissolution of existing calcite 73,1 x 103 mg/l Fredd and Fogler (1998)
crystals
Chrondroitin Sulphate 65% crystal growth rate 100mg/l Manoli and Dalas (2000)
reduction
Parvaalbumin Chelated calcium solution 1 mg/l Chen and Chan (2000)
from 100 ppm to below 1 ppm
Polyacrylate (PAA) 60% scale formation 5 mg/l Ravichandran and Sankara
reduction Narayanan (2002)
93% precipitation inhibition 10 mg/l Harris and Marshall (1981)
Scale coverage reduction 25 mg/l Neville and Morizot (2000)
from 20,7% to 6,8%
Scale formation reduction by 25 mg/l Yang et al. (2001)
76%
Polycarboxylic Acids 80% growth rate reduction 1 mg/l Reddy and Hoch (2001)
Inhibition of nucleation and 0,1-200 mg/l Verraest et al. (1996)
reduction of crystal growth
2-phosphonobutane-1,2,3- Scale formation reduction by 25 mg/l Yang et al. (2001)

tricarboxylic acid (PBTCA) 96%
Strongly Inhibited 270 mg/l Compton and Brown (1995)
Sodium alginate 95% crystal growth rate 17,9 x 103 mg/l Manoli and Dalas (2002)
reduction
Phosphonic Acids Inhibits aragonite & calcite 20,6 x 103 mg/l Nancollas and Sawada
growth (1982)
12
Scale growth rate reduction 0,7 mg/l Garcia et al. (2001)
Inhibition of precipitation 10 mg/l Harris and Marshall (1981)

(Up to 94%)
Organophosphorus Increases induction time from 18,2 x 103 mg/l Xyla et al. (1992)
Compounds 38 to 288 minutes
Metallocene dichlorides Growth rate reductioin of 9,96 mg/l Malkaj and Dalas (2002)
66%
Inorganic agents
Phosphate Calcite precipitation 1,9 mg/l Plant and House (2002)
inhibition
Polyphosphate Complete precepitation 10 mg/l Harris and Marshall (1981)
inhibition
Iron Complete calcite growth 55,8 x 103 mg/l Takasaki et al. (1994)
inhibition
Magnesium Precipitation induction time 13 mg/l Tai and Chien (2002)
increases from 120 minutes to
200 minutes
Calcite formation inhibition 24 mg/l Compton and Brown (1994)
Preferential aragonite 37 mg/l Meldrum and Hyde (2001)
precipitation
Induction time increase of 1 240 mg/l Wada et al. (1995)
hour
Precipitation inhibition 295 mg/l Meldrum and Hyde (2001)
Manganese No effect 2 mg/l Pernot et al. (1998)
Copper Complete inhibition 5 mg/l Ledion et al. (2002)
3
Induction time increase of 30 31,8 x 10 mg/l Wada et al. (1995)
minutes
Chromium Induction time increase of 90 26 x 103 mg/l Wada et al. (1995)
minutes
Silver Nucleation promotion of 54 x 103 mg/l Wada et al. (1995)
calcium carbonate
Sodium No effect Tai and Chen
Aluminium Induction time increase of 48 130 mg/l Wada et al. (1998)
minutes
Zinc Induction time increase of 30 Wada et al. (1998)
minutes
80% calcite growth rate 13,1 x 103 mg/l Meyer (1984)
reduction
13
77% scale formation 0,2 mg/l Coetzee et al. (1996)
reduction
Increase in number of 0,5 mg/l Pernot et al. (1999)
particles in bulk solution
Calcite precipitate decrease 0,98 mg/l Glasner and Weiss (1980)
and Zn/Ca ratio increased
Due to water having such a unique nature in terms of its exact composition, various factors, including
supersaturation level, pH, temperature, presence of other metal ions and flow velocity, all influence the
effectiveness of the anti-scaling agents as well as the dosage required for the same results (Wildebrand et
al., 2007). Thus, the water that needs treatment must be analysed with an inductively coupled plasma optical
emission spectrometer (ICP-OES) scan before treatment to determine the concentration of the minerals
present in the water, pH and alkalinity.
2.6 Principles of the different treatment methods:
2.6.1 Distillation
Distillation is the oldest type of desalination and is described as a process whereby the induction of a phase
change to a feed stream separates the pure water fraction by physical means (Amjad, 1996).
Multi-effect distillation
The majority of multi-effect distillation evaporators makes use of a horizontal tube evaporation principle.
This principle makes use of a distribution system that evenly distributes feed water on top of a bundle of
heated horizontal tubing. The water then flows from one tube to another, partially evaporating on the tubes.
A typical horizontal tubing network generally evaporates approximately 33.3% of the feed water (Greffrath,
2002), as illustrated in Figure 2-2.
14
Figure 2-2: Principle of horizontal tube evaporation
(Greffrath, 2002).
This method of evaporation has been succesfully applied to desalinate seawater and when making use of
multiple chambers that are in sequence; it ensures the effective application of the principle. Each following
chamber operates at a lower pressure than the preceding one, as illustrated in Figure 2-3. This can be
repeated for eight to sixteen chambers to ensure that a high percentage of fresh water is removed from the
feed and that a brine is produced that has a high percentage of total dissolved solids (Kennedy & Bremere,
2009). Depending on the application of this technology and the scaling properties of the feed water, the
parameters and operating conditions should be adjusted accordingly to result in optimal evaporation by
evaluating characteristics of the scaling properties of the scale forming minerals present (Wildebrand et al.,
2007).
Figure 2-3: MED schematic

(Kennedy & Bremere, 2009)
15
Multi-stage flash distillation
Process water is heated to a temperature between 105°C and 115°C as the water flows through
stages/chambers of decreasing pressure (Khoshrou et al., 2017). Water vapour is produced as the water
flashes off from the heat exchange surfaces. In an attempt to decrease the energy usage of the plant, the
vapour is used to heat the incoming feed water; more recently solar power as a source of heat energy has
been incorporated to lower the energy requirements even further. Figure 2-4 demonstrates the process
schematically. It is important to take into account the nature of the feed water (e.g. scaling potential at
certain temperatures) (Khoshrou et al., 2017) when assessing the feasibility of using this technology at
power generation plants.
Figure 2-4: Multi-stage flash distillation (MSF) schematic diagram

(Khoshrou et al., 2017)
2.6.2 Eutectic freeze distillation
Sea-ice has a much lower salt content than seawater (Ahmad et al., 2018). Eutectic freeze distillation has
been proven to be a successful method by which liquids can be concentrated and purified on a pilot scale.
This process can be described in three simplified stages: Firstly, the water/liquid is frozen, secondly the ice
crystals generated are separated from the remaining liquids, and thirdly the ice crystals are melted
(producing clean water), whilst the brine is disposed of or dried to produce salts (Ahmad et al., 2018).
Theoretically, the production of pure salts is possible due to the difference in salt crystallisation
temperatures (Randall et al., 2011).
More recently, this technology has been used for desalination purposes, yielding a few important
advantages, as described by Williams et al. (2015) and Randall and Nathoo (2015), including removing the
need for chemical pre-treatment, therefore eliminating the discharge of toxic chemicals, a high separation
factor, and high energy efficiency due to the latent heat of freezing (33.5 kJ/kg) being lower than the latent
heat of evaporation (2256.7 kJ/kg), leading to a lower energy requirement in comparison to other processes
(Rogers and Mayhew, 1994). Further advantages include fouling, scaling and corrosion resistance due to
the lower operating temperature (lowering maintenance costs), and finally, less expensive construction
materials (Randall & Nathoo, 2015 and Williams et al., 2015). Due to the desalination ability of this
16
technology, the possibility of incorporating it as a final stage in a water treatment process is feasible and
attractive to various industries as a result of the following advantages (Ahmad et al., 2018):
• No need for chemical pre-treatment, eliminating the discharge of toxic chemicals

• A very high and desirable separation factor
• A high and desirable energy efficiency due to the latent heat of freezing (333.5 kJ/kg) being lower
than the latent heat of evaporation (2256.7 kJ/kg), leading to a lower energy requirement in
comparison to other processes (Rogers & Mayhew, 1994)
• Fouling, scaling and corrosion resistance due to low operating temperature (lower maintenance
costs)
• Less expensive construction materials can be used at low temperatures (lowering capital cost)
2.6.3 Membrane technology
Membrane technology and techniques are a reliable means by which high quality water can be produced
(Greenlee et al., 2009). Typically, membrane technology applications include, amongst others, ultra-
filtration, forward osmosis, nano-filtration and reverse osmosis (Jiang et al., 2017).
Nano-filtration
Nano-filtration is usually used as a pre-treatment to RO, and is a method that rejects specific multivalent
ions from feed water through the use of a membrane or filter. This filtration produces a retentate,
concentrate or brine that can be treated elsewhere, which in turn lowers the load of the feed to the RO and
increases the performance of RO modules (Pramanik et al., 2016). Figure 2-5 depicts a schematic of the
basic principle on which this process functions.
Figure 2-5: Nano-filtration schematic

(Pramanik et al., 2016)
17
Reverse osmosis
Semi-permeable membranes are the principle component in an reverse osmosis plant. This membrane is
permeable to certain smaller particles, such as water molecules, but prevents solutes from passing through
the membrane. In order to reverse the osmotic process by forcing pure water through the semi-permeable
membranes, away from the low water potential, the dynamic pressure difference needs to exceed the
osmotic pressure. The pressure difference and permeability coefficient of these membranes has an effect
on the water passage through the membrane. Furthermore, the pre-treatment prior to the RO process, such
ultra-filtration, nano-filtration and the use of anti-scaling agents, can increase the efficiency of the process
(Jiang et al., 2017). This process is illustrated in Figure 2-6.
Figure 2-6: Reverse osmosis process diagram

(reproduced from Jiang et al., 2017)
Reverse osmosis has been extensively investigated and deemed the most energy efficient desalination
technology with an energy consumption of approximately 1.8 kWh/m3 (Jiang et al., 2017). Fouling is the
main challenge opposing reliable membrane performance and is influenced by the condition of the water
being treated as well as its pre-treatment (Perez-Gonzalez et al., 2012). Reverse osmosis produces a
retentate, concentrate or brine that is mineral rich and requires disposal similar to the retentate produced by
nano-filtration.
Forward osmosis
The movement of a substance from a solution with a low osmotic pressure to a solution with a higher
osmotic pressure through a semi-permeable membrane is known as forward osmosis (Boo et al., 2012). In
this water treatment process, a draw solution needs to be created, which can be either an organic or inorganic
substance with a high osmotic pressure that will draw the water from the saline stream with the lower
osmotic pressure (Boo et al., 2012). Colloidal fouling can occur should the draw solution or cross flow
velocities be inadequate, adding emphasis to the intricate cleaning methods required by this technology
(Boo et al., 2012). This process is illustrated in Figure 2-7.
18
.
Figure 2-7: Schematic illustration of forward osmosis process
(Boo et al., 2012)
2.6.4 Comparison of different technologies
When comparing different desalination technologies, energy consumption, scaling and maximum brine
salinity achievable are the main and most relevant terms by which to compare the technologies and
specifically keeping in mind the abundant heat/energy available on power plants in the form of steam to be
cooled. The combination of various technologies has the potential to provide the most efficient water
treatment and waste disposal (Ahmad et al., 2018).
Scaling
When examining these technologies, fouling and scaling are the common problems that desalination
industries encounter, as indicated by Chen et al. (2005). This is the result of the water characteristics being
altered as the water being treated becomes increasingly concentrated. Multi-stage flash (MSF) and multi-
effect distillation (MED) encounter scaling on heat exchange surfaces, negatively affecting heat exchange
efficiencies. Reverse osmosis (RO) and nano-filtration (NF) encounter scaling and fouling with the main
contributing compounds usually being, but not limited to, calcium carbonate, calcium sulphate and silica
(El-Dessouky & Ettouney, 2002). Using NF as pre-treatment to remove bivalent ions prior to the primary
desalination process has been a successful method by which scaling risk can be decreased as the process
progresses. This allows RO to proceed with less risk of scale forming and also allowing MSF the ability to
operate at more effective temperatures when operated in series (Hassan et al., 1998); and similar
conclusions can be drawn with the use of MED post-RO. The retentate and saline waste stream produced
from the desalination processes, or a similar sequence of water treating technologies can be theoretically
treated using Eutectic freeze crystallisation (EFC) to increase water recovery even more whilst producing
19
salts as a by-product. A 99.9% conversion of a waste stream generated using an RO plant to viable products
has been reported (Randall et al., 2011).
Salinity limitations
The largest pressure difference that an RO membrane can support before rupture has been reported to be
55-75 bar (Fichtner, 2011), with a maximum salinity of 70 000 ppm in order to keep the osmotic pressure
beneath this limiting salinity value. Additionally, the scale forming species’ concentration must be kept
below their respective supersaturation concentrations. MED and MSF salinity limitations are driven purely
by the scale forming ions’ concentrations and not by salinity itself (Roos et al., 2017), as the presence of
non-scale forming ions in suspension will not negatively influence the efficiency of these technologies.
This theory, having been tested as a pilot scale 3-effect MED system in California, reportedly achieved an
evaporated water salinity of 80 000 ppm, demonstrating that when the scaling risk is addressed, using MED
does not conform to the salinity limitations experienced by membrane processes. The use of anti-scaling
agents at the correct dosage can increase the performance of all of the above-named technologies by
preventing fouling on membranes, in pipes and on surfaces. When used in MED and MSF technologies, it
could additionally lead to the optimisation of heat transfer efficiencies by preventing scale formation on
heat exchange surfaces. Eutectic freeze crystallisation processes can operate right up to the salt saturation
limits to produce precipitated salt crystals as a product (Ahmad et al., 2018).
Energy consumption
Reverse osmosis electric energy consumption is dependent on the osmotic pressure (i.e. salinity) of the feed
water as well as the retentate/brine, as this influences the amount of pressure needed for the RO process.
Reverse osmosis specific consumption of electricity for Eskom mine water with a salinity of 2000 ppm and
a retentate of 8000 ppm has been recorded at 1.9 kWh/m3 (Roos et al., 2017). Power consumption for
desalinating seawater with salinities of 39 000 ppm and 49 000 ppm has been reported to be 3.5 kWh/m3
and 4.25 kWh/m3 respectively (Fichtner, 2011). The main energy requirement for thermal desalination
processes such as MED and MSF is the thermal heat required to evaporate the feed water, with the electrical
requirements (such as pumping and to overcome pressure drops) contributing to a lesser extent. Reported
electrical consumption of MED technology is between 1.5 kWh/m3 and 2.5 kWh/m3 (Fichtner, 2011), and
for MSF processes between 3 kWh/m3 and 5 kWh/m3. When using EFC for seawater desalination, power
consumption of 12.5 kWh/m3 has been reported (Lewis et al., 2010); however, varying values have been
reported by different publications.
2.7 Conclusion
Eskom currently has nano-filtration as well as reverse osmosis as water reclamation and purification
technologies present at some of its plants, such as Grootvlei and Lethabo, producing saline RO-retentate
that needs to be disposed of to adhere to the zero liquid effluent discharge stipulations in its water usage
20
licenses. Due to the abundant heat present at Eskom’s power plants and the versatility of multi-effect
distillation, the possibility exists to implement a sequence of membrane and thermal technologies to
approach a zero waste water reclamation process. An example of such a sequence is as follows: The use of
nano-filtration to remove bivalent scaling ions from feed water before treating nano-filtration permeate
using RO. The treatment of nano-filtration retentate to precipitate scaling ions. The combination of treated
nano-filtration retentate with RO-retentate and treatment using multi-effect distillation –a method of heating
that harnesses some of the abundant heat that needs to be dissipated at the power plant, as well as the correct
selection and dosage of anti-scaling agents. The last step in such a sequence could be making use of Eutectic
freeze crystallisation to separate the reject from the multi-effect distillation into fresh water and dry salt.
In this study, a bench scale single effect evaporator (using the same principles and conditions produced in
multi-effect distillation) is proposed to examine the possibility of water recovery from waste water streams
produced at Eskom’s power plants while using correctly selected anti-scaling agents. Reverse osmosis
retentate from the Lethabo Power Plant will be treated in the bench scale single effect evaporator with and
without the use of scale-inhibiting agents.
21
References
Ahmad, M., Oatlry-Radcliff, D. & Williams, P. 2018. Can a hybrid RO-Freeze process lead to sustainable
water supplies? Desalination, 431: 140-150.
Al-Roomi, Y. & Hussain, H. 2016. Potential kinetic model for scaling and scale inhibition mechanism.
Desalination, 393: 186-195.
Amjad, Z. 1996. Scale Inhibition is Desalination Applications: An Overview. Houston, Texas (United
States), NACE International.
Amjad, Z. 2002. Advances in crystal growth inhibition technologies. New York: Kluwer Academic
Publishers.
Anklam, M. & Firoozabadi, A., 2005. An Interfacial energy mechanism for the complete inhibition of
crystal growth by inhibitor adsorption. The Journal of Chemical Physics, 123.
Antony, A., How Low, J., Gray, S., Childress, A.E. Le-Clech, P & Leslie, G. 2011. Scale formation and
control in high pressure membrane water treatment systems: A review. Journal of Membrane Science: 1-
16.
Boo, C., Lee, S., Elimelech, M., Meng, Z. & Hong, S. 2012. Colloidal fouling in forward osmosis: Role of
reverse salt diffusion. Membrane Science, Volume 390-391: 277-284.
Chen, T., Neville, A. & Yuan, M. 2005. Calcium carbonate scale formation—assessing the initial stages
of precipitation and deposition, s.l. Elsevier.
Dandekar, P. & Doherty, M. 2014. A Mechanistic Growth Model for Inorganic Crystals: Growth
Mechanism. AIChE Journal, 60(11): 3720-3731.
Drela, I., Falewicz, P. & Kuczkowska, S. 1998. New rapid test for evaluation of scale inhibitors, s.l.
Elsevier.
El-Dessouky, H.T. & Ettouney, H. 2002. Fundamentals of Salt Water Desalination. Amsterdam: Elsevier.
Ferguson, R., Ferguson, B. & Stancavage, R. 2013. Scale formation and optimizing scale-inhibitor
dosages. IDA Journal of Desalination and Water Reuse, 2(3): 30-39.
Fichtner, M. 2011. Regional Water Outlook Part 2: Desalination USing Renewable Energy Task 1 -
Desalination Potential Stuttgart: s.n.
Flynn, D. 2009. Nalco Water Handbook. 3rd ed. New York, Chicago, San Francisco, Lisbon, London,
Madrid, Mexico City, Milan, New Delhi, San Juan, Seoul, Singapore, Sydney, Toronto: McGraw-Hill.
22
Geldenhuys, A. J., Maree, J. P., de Beer, M. & Hlabela, P., 2003. An integrated limestone/lime process for
partial sul;hate removal. The Journal of The South African Institute of Mining and Metallurgy, 103(6): 345-
354.
Greenlee, L., Lawler, D.F., Freeman, B.D., Marrot, B. & Moulin, P. 2009. Reverse Osmosis Desalination:
Water sources, technology and today’s challenges. Water Research, 43(9): 2317-2348.
Greffrath, D.R. 2002. Effect of Scaling and Operation of Thermal Seawater Desalination Plants.
Germany: SERCK COMO GmbH.
Hassan, A., Al-Sofi, M.A.K., Al-Amoudi, A.S., Jamaluddin, A.T.M., Farooque, A.M., Rowaili, A., Dalvi,
A.G.I., Kither, N.M., Mustafa, G.M. & Al-Tisan, I.A.R. 1998. A new approach to membrane and thermal
seawater desalination processes using nanofiltration membranes (Part 1). Desalination, 118: 35-51.
Hasson, D., Shemer, H. & Sher, A. 2011. State of the Art of Friendly “Green” Scale Control Inhibitors:
A Review Article. Texas, ACS Publications: 7601-7607.
Jiang, S., Li, Y. & Ladewig, B. 2017. A review of reverse osmosis membrane fouling and control
strategies. Science of the Total Environment, 595: 567-583.
Kavitha, A., Vasudevan, T. & Prabu, H.G. 2011. Evaluation of synthesized antiscaling agents for cooling
water system application. Desalination, 268: 38-45.
Kennedy, M.D. & Bremere, I. & S.J.C. 2009. Desalination technology. In: Small community water
supplies: technology, people & partnership. Hague, Netherlands: s.n.
Khoshrou, I., Jafari Nasr, M. & Bakhtari, K. 2017. New opportunities in mass and energy consumption of
the Multi-Stage Flash Distillation type of brackish water desalination process. Solar Energy, 153: 115-125.
Lewis, A.E., Nathoo, J., Thomsen, K., Kramer, H.J., Witkamp, G.J., Reddy, S.T. & Randall, D.G. 2010.
Design of a Eutectic Freeze Crystallization process for multicomponent waste water stream. Chemical
Engineering Research and Design, 88: 1290-1296.
Li, H. 2010. Mineral precipitation and deposition in cooling systems using impaired waters: mechanisms,
kinetics, and inhibition Pittsburgh: s.n.
Loewenthal, R. & Marais, G., 1976. Carbonate chemistry of aquatic systems: theory and application.
United States: Ann Arbor Science Publishers.
MacAdam, J. & Parsons, S. 2004. Calcium carbonate scale formation and control. Reviews in
Environmental Science and Bio/Technology, 3: 159–169.
Maree, J. et al., 2004. Treatment of Mine Water for Sulphate and Metal Removal using Barium Sulphide.
Mine Water and the Environment, 23: 195-203.
23
Meyer, H. 1984. The influence of impurities on the growth rate of calcite. Journal of Crystal Growth, 66:
639-646.
Oner, M., Dogan, O. & Oner, G. 1998. The influence of polyelectrolytes architecture on calcium sulfate
dihydrate growth retardation. Journal of Crystal Growth, 186(3): 427-437.
Perez-Gonzalez, A., Urtiaga, A., Ibanez, R. & Ortiz, I. 2012. State of the art review on the treatment
technologies of water reverse osmosis concentrates. Water Research, 46:. 267-283.
Piyadasa, C., Ridgway, H.F., Yeager, T.R., Stewart, M.B., Pelekani, C., Gray, S.R. & Orbell, J.D. 2017.
The application of electromagnetic fields to the control of the scaling and biofouling of reverse osmosis
membranes - A review. Desalination, 418: 19-34.
Pramanik, B., Thangavadivel, K., Li, S. & Jegatheesan, V. 2016. A critical review of membrane
crystallization for the purification of water and recovery of minerals. Environmental Science and
Biotechnology, 15: 411-439.
Randall, D. & Nathoo, J. 2015. A succinct review of the treatment of Reverse Osmosis brines using Freeze
Crystallization. Journal of Water Process Engineering, 8: 186-194.
Randall, D., Nathoo, J. & Lewis, A. 2011. A case study for treating reverse osmosis brine using Eutectic
Freeze Crystallization - Appoaching a zero waste process. Desalination, 266: 256-262.
Rogers, G. & Mayhew, Y. 1994. Thermodynamic and Transport Properties of Fluids. 5th ed. Oxford,
United Kingdom: Blackwell Publishing.
Roos, T.H., Rogers, D.E.C. & Gericke, G. 2017. Solar-assisted MED treatment of Eskom power station
Shakkthivel, P. & Vasudevan, T. 2006. Systems, Acrylic acid-diphenylamine sulphonic acid copolymer
threshold inhibitor for sulphate and carbonate scales in cooling water. Desalination, 197(1-3): 179-189.
Stumn, W., Sigg, L. & Sulzberger, B. 1992. Chemistry of the solid water interface: Processes at mineral-
water and particle water interface in natural systems. New York: Wiley.
Wildebrand, C., Glade, H., Will, S., Essig, M., Rieger, J., Buchner K.H. & Brodt, G. 2007. Effects of
process parameters and anti-scaling agents on scale formation in horizontal tube falling film evaporators,
s.l. Elsevier.
Williams, P., Ahmad, M., Connolly, B. & Oatley-Radcliff, D. 2015. Technology for freeze concentration
in the desalination industry. Desalination, 356: 314-327.
Yang, Q., Liu, Y., Gu, A., Ding, J. & Shen, Z. 2002. Investigation of induction period and morphology of
CaCO3 fouling on a heated surface. Chemical Engineering Science, 6(57): 921-931.
24
Zhang, Y., Shaw, H., Farquhara, R. & Dawe, R. 2001. The kinetics of carbonate scaling—application for
the prediction of downhole carbonate scaling, s.l. Elsvier.
25
Design, Configuration and Construction of Bench
Scale MED
Abstract— An investigation into the potential scale formation during the application of multi-effect
distillation (MED) technology for the treatment of reverse osmosis (RO) retentate was performed. An
improved bench scale single effect evaporation unit was designed using the Solidworks Simulation Package
and appropriate materials. Operating conditions were determined to be 80°C and 0.48 bar with the design
constructed to maximise evaporation. Once assembled, RO-retentate was introduced into the evaporator,
with an evaporation rate of 4 L/h (22.3 L/m2/h), and the variation of the concentrations of scale forming
elements was determined – the evaporation translates to 34.5% water recovery. Speciation using PHREEQ-
C modelling software showed that the concentrations of scale forming ions increased during evaporation.
Scale prediction and prevention were explored using French Creek software, with anti-scaling agents used
to minimise scaling.
The dosage and type of the anti-scalant, as predicted by the French Creek software, yielded the desired
results by preventing scaling on the heat exchanger surface and keeping ions in suspension. When
measuring the performance of the anti-scaling agents on this basis, AMPS fared better than PBTC, whilst
HEDP yielded concentrations of scale forming ions lower than when no anti-scaling agent was dosed. This
may be attributed to complexation of the scale forming ions with the HEDP.
Keywords—MED, scale formation, RO-retentate, PHREEQ-C, French Creek.
26
3.1 Introduction
South Africa is a water-scarce country. Therefore, water consumers are forced to explore the utilisation of
all available water sources, and the water scarcity is being worsened by climate change. Despite this water
shortage, the demand for water usage in industry and for human consumption remains high, creating an
ever-increasing dependence on seasonal rainfall (Fosso-Kankeu et al., 2016).
Waste water effluent is becoming more valuable as advances in technology make the recovery, treatment
and re-use of waste water more attractive (Fosso-Kankeu et al., 2011). Scale formation is a common
occurrence when water is being treated and re-used, as treatment methods such as heat exchange and
membrane processes increase the concentration of the initial minerals present in the feed water (Roos et al.,
2017).
Eskom, South Africa’s national power utility, is allocated 2-3% of the national available fresh water to
generate electricity. Eskom, however, is under pressure to minimise its water usage and has set a target of
lowering its usage to 1.34 L of water per kilowatt hour electricity produced by 2020, from the 1.39 L/kWh
presently (Fosso-Kankeu et al., 2016). Of the total water intake, 98% is lost via evaporation in the cooling
towers of wet cooled power plants (Surenda, 2014). Therefore, the re-usage of certain water streams may
potentially decrease the freshwater intake needed to replenish the amount of water lost due to evaporation.
Eskom has thus initiated an investigation towards water management and proper usage of all available
water sources. Reject from the primary RO module is high in salinity (depending on the feed water), and
has been identified as a possible source from which to abstract usable water.
Sea water has been treated using multi-effect distillation (MED) systems, encouraging the usage of this
technology for this application (Wildebrand et al., 2007). The majority of multi-effect distillation
evaporators makes use of a horizontal tube evaporation principle. This principle makes use of a distribution
system that evenly distributes feed water on top of a bundle of heated horizontal tubing. The water then
flows from one tube to another, partially evaporating on the tubes. A typical horizontal tubing network
generally evaporates approximately 33.3% of the feed water (Greffrath, 2002). A limitation of such a
system is the scale formed on the heat exchange surfaces due to the quality of the water, therefore negatively
affecting the efficiency of the MED (Wildebrand et al., 2007).
This problem is amplified when treating concentrated and saline waters that are prone to scaling on heat
exchange surfaces, especially at elevated temperatures. Reducing the temperature of the heat exchange
surfaces to minimise scaling causes the evaporation rate of the clean water to decrease. The lower
evaporation rate can be countered by reducing the pressure of the entire chamber for the water to evaporate
at an adequate or even a higher rate. The reduction of the heat exchange surface temperature requires the
steam supply pressure to be decreased as well. This can be accomplished by introducing vacuum pumps to
the system, while maintaining the steam in the vapour phase to ensure efficient heat exchange.
27
These evaporation favouring, scale avoiding conditions indicate that MED technology could potentially be
incorporated into the existing water treatment system at Eskom. This technology will aid Eskom in adhering
to its zero liquid effluent discharge (ZLED) policy, stipulated in the utility’s water use licences (Roos et
al., 2017), by treating relatively saline solutions for reuse. The primary purpose of MED is to serve as a
water producing/recycling technology that can treat saline/brine solutions. The potential secondary purpose
of the technology when applied to a power plant could be to dissipate some of the abundant heat that would
otherwise have to be dissipated in the cooling towers, thus decreasing the amount of cooling water needed
(Drela et al., 1998).
Scaling is caused by concentrated minerals in water when they become insoluble. When the mineral
constituents remain soluble, they have no negative effect on heat transfer efficiency (Kavitha et al., 2011),
and this should be considered when exposing mineral-rich waters to heat exchange surfaces. Scale
formation occurs in alkaline water and is attributed to the presence of scale forming ions such as Ca 2+ and
or Mg2+ and calcium carbonate (Li, 2010). Process parameters also plays a role in scale formation – with
temperature and flow rate being the most influential (Wildebrand et al., 2007), as these parameters can
influence deposit (scale) structures, shape, size and composition.
Scaling potential can be determined by using PHREEQ-C, an open-source modelling package developed
by the US Geological Survey (USGS, 2016). French Creek software can additionally be used, as it combines
multiple saturation and stability indices to accurately predict the scaling potentials of the minerals present
in the water. This is similar to the capabilities of PHREEQ-C, with the additional capability of predicting
which anti-scaling agent will yield the best results and the simulation of addition (French Creek Software,
2017).
When mineral ions are present in water, they may become more concentrated over time and eventually
cause the water to become supersaturated. Mineral ions under increasingly supersaturated conditions will,
due to their polarity and random thermal motion, form clusters. Once these clusters pass a critical size,
nucleation will take place. As soon as nuclei are formed, spontaneous crystallisation will take place,
forming larger particles. Subject to the surface properties and composition of these particles, the particles
can either aggregate to create larger particles, or remain suspended in the water for a period. After sufficient
time, these articles and mineral ions will diffuse through the hydraulic boundary layer (Li, 2010) and form
scale after depositing onto a solid surface. The scaling process is illustrated in Figure 3-1.
28
Figure 3-1: Schematic of processes involved with the formation of mineral
Reproduced from MacAdam and Parsons (2004)
In the present study, reverse osmosis (RO) retentate from the Lethabo Power Plant was evaporated in an
improved bench-top single effect evaporator and the variation in the concentration of scale forming
elements was evaluated.
3.2 Operating Conditions
3.2.1 Temperature
A steam generator was used as the evaporator’s heat source; it supplied steam through the heat exchange
surface. The temperature of the steam was a concern when using the steam generator, as the steam had a
higher temperature than the critical temperature for scale formation. It has been reported that, at a
temperature above 80°C, the rate of scale formation is substantially increased (Wildebrand et al., 2007).
The temperature set point target is consequently 80 °C.
3.2.2 Pressure
Saturation pressure and operating parameter targets were determined using Thermosolver software
(Koretsky, 2014), specifically the Saturation Pressure Calculator and Antoine Equation of state. The
ambient pressure of Potchefstroom (0.875 bar) (Everson et al., 2013) was considered during vacuum
calculations at different temperatures. These results are detailed in Table 3-1.
Table 3-1: Operating parameter targets of evaporation chamber

Temperature (° C) Solver Saturation Pressure (bar) Size of Vacuum Required (bar)
60 0,1992 0,6758
65 0,2502 0,6249
70 0,3117 0.5633
29
75 0,3856 0.4894
80 0,4737 0.4013
According to the thermodynamic principle, when a temperature target of 80°C is achieved, the pressure
inside the evaporation chamber of the MED needs to be as close to 0.47 bar as possible for optimal
evaporation to take place, as this is the saturation pressure of water at 80°C.
3.3 Materials
Before the evaporator could be built, the scale forming properties as well as the strength and temperature
effects of the materials had to be evaluated. These materials were to be exposed to pressure reduction, high
temperatures and mineral rich water with increasing scaling potential. The design of the evaporator can be
divided into three main components: The evaporation chamber, the dual heat exchange surface running
through the evaporation chamber, and drip tray that dispenses the water onto the heat exchange surface
inside the evaporation chamber.
Polymers and certain polyvinyl chlorides were deemed appropriate materials for the construction of the
evaporation chamber of the evaporator as well as the drip tray. These materials are scaling resistant and
would therefore not influence the concentrated water (RO-retentate) that was used in the evaporator (The
International Nickel Company, INC, 1963). Stainless steel was used inside the evaporation chamber to
control the height of the drip tray above the heat exchange surface due to its unreactive nature, largely due
to the unavailability of nucleation sites suitable for the scale to attach to (Copper Development Association,
2016).
The dual heat exchange surface inside the evaporator is metallic to provide the necessary heat conductivity
needed at the exchange surface. Cu-Ni and copper alloys have been used as heat exchanging surfaces
successfully, particularly in sea-water cooled heat exchangers (Kirk & Tuthill, 1992), with a mild corrosion
resistance (Copper Development Association, 2016). When comparing the high salinity of sea water to RO-
retentate, copper will perform as a sufficient heat exchange surface due to its mild corrosion resistance.
Without proper pre-treatment, the copper heat exchange surfaces are mildly susceptible to scaling.
The material of the pump that recycles the brine from the evaporation chamber to the drip tray is exposed
to the high scaling potential water and must thus be constructed with appropriate materials. A magnetically
driven, single stage, centrifugal pump constructed using thermoplastics will suffice due to its chemical
resistance.
These three key components are depicted in Figure 3-2:
30
Figure 3-2: Main components in MED design: (a) evaporation vessel; (b) recycle pump; (c) drip-tray; (d) heat
source/exchange surface
3.4 Design and Configuration
3.4.1 Evaporation chamber
The initially commissioned bench scale MED (Fosso-Kankeu et al., 2016) had to be improved to conduct
more realistic and extensive experiments. Firstly, the structural integrity of the evaporation chamber was
significantly improved to increase the vacuum applied to it without the structure failing (Fosso-Kankeu et
al., 2016). Secondly, the initial bench scale MED could function with a pressure reduction of 0.09 bar before
the risk of structural decomposition occurs.
A simulation on Solidworks (3D design and simulation software) on a range of polycarbonate

thermoplastics resulted in a material selection and design consisting of multiple and extensive internal rib
structures. The Solidworks Simulation Packet recommended a material thickness of 12 mm. Trial and error
with the types, sizes and placement of the rib reinforcing structures resulted in the design displayed in
Figure 3-3.
Figure 3-3: Three-dimensional visualization of improved evaporation chamber design

31
This design was simulated and is represented in Figures 3-4 and 3-5.
Figure 3-4: 0.6 MPa (6 bar) pressure difference applied to design in simulation program - visualization
Figure 3-5: Solidworks simulation displaying a maximum displacement of 4 mm when 6 bar pressure
difference is applied to evaporation chamber
32
Stratasys (2013) performed an investigation to evaluate the effects of time, temperature and environment
on a polycarbonate (PC) thermoplastic. They evaluated the PC thermoplastic at temperatures ranging from
-40°C to 140°C. The tensile strength of this PC thermoplastic at a temperature of 80°C remains above a
pressure of 300 bar. The results of the investigation are displayed in Table 3-2.
Table 3-2: Tensile strength of polycarbonate and the effect of temperature Stratasys (2013)
Temperature Wet Dry 1 week 13 weeks 26 weeks 52 weeks
(°C) (bar) (bar) (bar) (bar) (bar) (bar)
40 492 517 527 502 540 495
60 414 456 461 446 493 449
80 377 401 429 396 426 395
100 336 351 391 341 361 335
It is thus apparent when examining the material properties from literature and the Solidworks design
simulation that PC thermoplastic yielded satisfactory results, indicating that the design’s structural integrity
will not be compromised when exposed to the target pressure reduction of 0.41 bar.
3.4.2 Vacuum pumps
The vacuum pump’s cooling system needed to be altered from a recirculating system to a ‘once through’
system to run the pumps for a longer period whilst preventing the risk of overheating. This was due to the
fact that they were operating near to or above recommended temperature ranges with the initial recycling
closed circuit water configuration.
3.4.3 Drip tray
The drip tray had to fulfil a dual purpose. Firstly, the drip tray plays a key part in the structural integrity of
the evaporation chamber and secondly, laminar flow of water onto the heat exchange surface. Multiple
factors influence the latter, with the size and distance in between the apertures, and the distance heat
exchange surfaces are below the drip tray being key factors when designing the evaporator.
Design of the drip tray for distributing the liquid on the heat exchanger tube was based on the equation for
the flow rate, 𝑄𝑟𝑒𝑐 , through an orifice with area 𝐴, using a water height ℎ as driving force for flow:
𝑄𝑟𝑒𝑐 = 0.61𝐴√2𝑔ℎ Equation 3-1
This equation was derived from the Bernoulli flow equation, assuming that the fluid contracts after the
orifice to an area 0.61 A as shown in Figure 3-6.
33
Figure 3-6: Flow contraction after an orifice with area A
The diameter of the 𝑁 orifices in a drip tray of height ℎ to generate a total recycle liquid flow rate of
𝑄𝑟𝑒𝑐 m3/s is then given by:
4𝜋𝑄𝑟𝑒𝑐
𝑑=√ Equation 3-2
0.61𝑁√2𝑔ℎ
The feed flow rate for falling film evaporators, , can be expressed as the mass flow rate per length of tube,
L, in kg/m/s. For horizontal films, this is usually expressed per tube side:
𝜌𝑄𝑟𝑒𝑐
Γ= 2𝐿
Equation 3-3
Where 𝑄𝑟𝑒𝑐 is the volumetric flow rate of the feed water recycled via the drip tray over the heat exchanger
tube and 𝜌 is the water density.
The falling film Reynolds number is then, according to Ouldhadda et al. (2002), defined as:
4Γ 2𝜌𝑄𝑟𝑒𝑐
𝑅𝑒 = 𝜇
= 𝜇𝐿
Equation 3-4
Flow can be considered to be laminar when the Reynolds number ranges between 400 and 4000 (Ouldhadda
et al., 2002), whilst typical MED conditions range from 500 to 2000, according to Shausberger et al. (2009),
with the significance of the Reynolds number being displayed in Figure 3-7. Table 3-3 displays the
approximate ranges for the film Reynolds number with respect to the flow pattern – the hydrodynamic
regime.
34
Figure 3-7: Inter tube falling-film modes: (A) droplet mode, (B) jet mode, (C) sheet mode
(Schausberger et al., 2009)
Table 3-3: Ranges for the film patterns and corresponding flow regimes
Scenario Reynolds Number
Flow Pattern (A)-(B) 150-200
Flow Pattern (B)-(C) 315-600
Flow regime Laminar – Turbulent 1000-2000
Typical MED Re range 500-2000
Other factors that influence the flow of liquid over the heat exchange surface that should be kept in mind
include flow rate of the water moving through the drip tray, turbulence of flow inside the drip tray and the
length of the heat exchange surface. Consequences when not designed correctly include decreased heat
exchange efficiency and increased scale formation.
Equations 1 - 4, in conjunction with the two heat exchange surfaces running through the chamber as well
as the need for additional structural integrity, resulted in the design depicted in Figures 3-8 to 3-10. The
two holes on top of the drip tray were meant for the inlet of water to be dripped onto the heat exchange
surfaces as well as an air vent to control the liquid level inside the drip tray. The four anchoring holes
allowed for the attachment of the stainless steel adjustable legs.
35
Figure 3-8: Drip tray prior to installation inside evaporation chamber
Figure 3-9: 3-D view of drip tray
Figure 3-10: View of aperture sizing and spacing of drip tray

36
3.4.4 Operation
The previously determined ideal operating conditions (Section 3.2) were initiated with the incorporation of
two Busch LX 0030-0430 B Vacuum Pumps. One was introduced to the evaporation chamber to lower the
pressure inside and to increase the rate of evaporation, whilst the other was introduced to lower the heat
exchange surface temperature by lowering the pressure of the steam flowing through the heat exchange
surface. It was determined that the temperature of the steam providing the heat needed to be lowered to
80°C, and a vacuum of 0.41 bar needed to be created inside the evaporation chamber.
3.4.5 Method of evaluating MED performance
An accurate measurement of the amount of water evaporated was challenging to obtain due to the once-
through cooling of the vacuum pump that removed the vapour from the evaporation chamber. PHREEQ-C
software and the PITZER database were used to verify the volumetric calculations and level differences
inside the MED to determine the amount of water evaporated.
3.5 Methodology
3.5.1 Assembly
Once the design was finalised and materials obtained, the assembly entailed milling the polycarbonate to
the right sizes, after which they were glued together (Figures 3-11 to 3-13).
Figure 3-11: Milled polycarbonate
37
Figure 3-12: Gluing of polycarbonate to form evaporation chamber.
Figure 3-13: Evaporation chamber assembly

38
Stainless steel fittings and unreactive reinforced plastic piping were used to assemble the closed water
circuit running from the evaporation chamber to the recycling pump and to the drip tray. The recirculation
piping was further reinforced with stainless steel springs to ensure the vacuum would not influence the flow
of the brine/water. The available steam supply piping had to be adapted and positioned to the inlet of the
MED with valves to control the supply and pressure with pressure and temperature gauges before, inside,
and after MED. A frame to house the different components (evaporation chamber, recirculation chamber,
recirculation pump, vacuum pumps, vacuum pump cooling and effluent tanks) was also designed and
constructed. A silicon sheet lining was used as a sealant beneath the lid of the MED to ensure an airtight
seal. The start-up of the evaporator (filling up and recirculation of RO brine) proved challenging due to the
closed and airtight nature of the evaporator. Various air vent valves were introduced to speed up and
improve the control during the start-up and initial stabilisation of the evaporator. A continuous cooling
liquid supply with buffer tank was incorporated into the design to ensure adequate vacuum pump operating
temperatures.
3.5.2 Water sampling and analysis
Water/RO-retentate was sampled directly from the Lethabo Power Plant’s RO-retentate stream using clean
and dry 25 L containers. Samples were sent for analysis to the Midvaal Water Company, as its Scientific
Services division and laboratory are SANAS (South African National Accreditation System) accredited.
The results from this analysis were used for scale formation modelling.
3.5.3 Operation and control of MED
The control of the MED utilised the manipulation of various valves. The pressure valve on the steam supply
line was used to alter the temperature of the steam entering the MED, as the steam supply was placed under
a vacuum. A pressure regulating valve on the MED was used to manipulate the pressure in the MED. A
valve regulating the flow from the recirculation pump was used to influence the flow to the drip tray –
ultimately the flow of water onto the heat exchange surface.
3.5.4 Determination of anti-scaling agents by manual calculations
Accurate calculation included manual calculations, which mainly involved the Langelier saturation and
Ryznar stability indices, as well as the developing Puckorius equilibrium index. These indices usually
produced conflicting results, and thus simulation software was used to verify the exact calculations.
39
3.5.5 Determination of anti-scaling agents with modelling software
Scaling potential can be viewed by using PHREEQ-C, an open-source modelling package developed by the
US Geological Survey. This software was initially created for the simulation and modelling of water
chemistry with regard to hydrogeology and geochemical modelling, but when using thermochemical
equilibrium calculations, the scaling potential of aqueous solutions can be accurately identified (USGS,
2016).
French Creek software is applicable in this study, as it combines the Langelier saturation and Ryznar
stability indices, the Puckorius equilibrium index, the calculation of saturation levels based on an extensive
free ion association/exchange model, as well as Momentary Excess and Larson-Skold indices, which are
deemed appropriate and realistic by the programme authors to accurately predict the scaling potential of
the minerals present in the water. This is similar to PHREEQ-C’s capabilities, but with the additional
capabilities of predicting which anti-scaling agent will yield the best performance. Furthermore, the French
Creek WaterCycle software was developed specifically to allow the user to view these popular and lesser
known indicators in order to display problems and potential problems specifically inside a cooling water
treatment system. This software incorporates more than 18 different relevant scales (such as calcium
carbonate, calcite, aragonite, barite, tricalcium phosphate, etc.), yielding accurate results and predictions. It
functions as a system simulator taking water chemistry and operating conditions into consideration and also
provides treatment plans and anti-scaling agent limitations and preferences (French Creek Software, 2017).
3.5.6 Anti-scaling agent dosage
The data obtained from the RO-retentate to be used in the MED were plugged into French Creek to
determine which anti-scaling agents and dosages were to be used (Table 3-4).
Table 3-4: Anti-scaling agent active ingredient dosages

Anti-Scaling Agent Dosage (ppm)
100% HEDP 0.169
100% PBTC 0.210
100% AMPS 0.369
Simulation results are displayed in Figures 3-14 to 3-16.
40
Figure 3-14: AMP dosage profile
Figure 3-15: HEDP dosage profile
41
Figure 3-16: PBTC dosage profile
3.6 Results
3.6.1 Performance of MED
Parameters
The targeted parameters were obtained. A steam supply temperature varying between 78°C and 80°C
(Figure 3-17), paired with a vacuum varying between 0.40 and 0.44 bar inside the evaporation chamber
(Figure 3-18) of the MED ensured rapid evaporation. The temperature inside the MED stabilised at 60°C
(Figure 3-19) with the steam exit temperature stabilising around 50°C.
Figure 3-17: Temperature gauge

displaying the temperature of steam
entering the evaporation chamber inside
the heat exchange piping
42
Figure 3-18: Vacuum gauge displaying vacuum
inside evaporation chamber
Figure 3-19: Temperature gauge displaying

temperature inside evaporation chamber
The assembled and functioning single effect evaporator is displayed in Figure 3-20.
43
Figure 3-20: Assembled, functioning single effect bench scale evaporator
Evaporation rate
When evaluating the water level pre-and post-evaporation, an 8 cm level difference translated into 24 L of
water being evaporated from the initial 70 L over a 6-hour run time. Displayed in Figure 3-21 is the water
flow onto the heat exchange surfaces.
44
Figure 3-21: Drip tray's laminar water flow
distribution onto heat exchange surfaces
The trial runs were performed and tabulated below in Table 3-5. It was the clear increase in the
concentrations of certain elements, portraying adequate functioning of the evaporator, that prompted further
and more in-depth investigation.
Table 3-5: Concentration in mg/L of a) RO-retentate before MED-evaporation and b) after MED-evaporation
ICP-OES RO- ICP-OES MED PHREEQ-C Simulation-
Retentate Reject Theoretical MED Reject
Ca 2830 3310 4304,38
K 10100 13400 15357,81
Mg 174,92 231,83 265,90
Anti-scaling agents were not used during trial runs, and the influence can be clearly seen in Table 3-5 when
comparing ICP-OES analysis of the MED reject with the PHREEQ-C simulation theoretical MED reject
concentrations. It is palpably clear that scaling and precipitation took place during the trial runs inside the
MED.
Analysis of water
The water both prior to and after the evaporation were analysed using Inductively Coupled Plasma Atomic
Emission Spectroscopy (ICP-OES) to determine the elements present and to determine the increase in
concentration. These analyses were then used for scale prediction and modelling. Displayed in Table 3-6 is
the increase in concentration of certain scale influencing and forming elements:
45
Table 3-6: Speciation of certain elements present in MED reject
Speciation of MED Reject without anti-scaling agents

Ba Ba2+ 92,41% K K+ 99,36%
BaSO4 7,60% KSO4- 0,65%
+ +
BaOH 0,00% Li Li 99,81%
Ca Ca2+ 95,83% LiSO4- 0,19%
CaSO4 4,19% Mg Mg2+ 87,52%
+
Cu(1) Cu 100,00% MgSO4 12,15%
Cu(2) Cu(OH)2 80,29% MgOH+ 0,34%
2+ 2+
Cu 9,88% Mn(2) Mn 91,33%
Cu2(OH)22+ 3,87% MnSO4 7,03%
+ +
CuOH 1,63% MnOH 1,68%
CuSO4 0,50% Mn(3) Mn3+ 100,00%
Cu(OH)3- 0,00% Na Na +
99,72%
Cu(OH)42- 0,00% NaSO4- 0,28%
Fe(2) Fe2+ 80,56% S(6) SO42- 48,72%
+
FeOH 13,30% CaSO4 22,91%
FeSO4 5,98% MgSO4 19,75%
-
Fe(OH)2 0,15% NaSO4 6,04%
Fe(OH)3- 0,00% KSO4- 2,49%
+
FeHSO4 0,00% SrSO4 0,06%
Fe(HS)2 0,00% BaSO4 0,01%
Fe(HS)3- 0,00% MnSO4 0,01%
-
Fe(3) Fe(OH)3 78,82% LiSO4 0,01%
Fe(OH)4- 17,82% Sr Sr2+ 94,45%
+
Fe(OH)2 3,36% SrSO4 5,57%
It is very clear that, when the water is concentrated, the use of PHREEQ-C modelling software displays the
abundance of scale forming ions present in the water (Ca+2 and Mg+2 for example) when the amount of
volumetrically determined water is evaporated from of the RO-retentate inside the MED.
Prediction of scale formation and modelling
Using the initial water analysis from Midvaal Water Company as input parameters combined with the
operating conditions inside the MED, Figures 3-22, 3-23 and 3-24 display the temperatures and pH-values
at which aragonite and calcite are supersaturated and the Langelier Saturation Index. These graphs were
generated using French Creek modelling software.
46
Figure 3-22: Aragonite saturation level vs. temperature vs. pH
Figure 3-23: Calcite saturaton level vs. temperature vs. pH
47
Figure 3-24: Langelier saturation index
French Creek software confirmed that scale will form and that various elements will become oversaturated
at 80°C, 0.48 bar pressure and a residence time of 6 hours.
When using the PHREEQ-C modelling software, the amount of water that evaporated was less than the
volumetrically determined amount, depending on which element’s balance was used. When evaporating 24
L of water with this modelling software, the results differed from the ICP-OES analysis, as displayed in
Table 3-5. The reason for this difference is that PHREEQ-C did not take the scale formed on the heat
exchange surface into account. According to the PHREEQ-C modelling, roughly 13 L of water should be
evaporated to obtain an identical calcium concentration to the ICP-OES analyses during the trial runs. When
taking the scale formation on the heat exchange surface into account and evaluating the increase of other
non-scaling elements, 24 L of water evaporated is confirmed to be a realistic and accurate measurement.
Furthermore, the speciation of elements before and after evaporation shows that more scale forming ions
such as Ca+2 and Mg+2 are likely to occur after evaporation of the feed water; implying that scale formation
will take place should the necessary precautions not be taken. Refer to Table 3-7 to compare the
performance of the MED with and without anti-scaling agents.
48
Table 3-7: PPM concentration of certain elements prior and post MED evaporation, with and without anti-
scaling agents
Before RO- ICP-OES ICP-OES ICP-OES After MED PHREEQC
MED Retentate to MED Reject MED MED Evaporation Simulation
be used with PBTC Reject with Reject No Dosage theoretical MED
inside MED (ppm) AMPS with HEDP (ppm) reject (ppm)
(ppm) (ppm) (ppm)
Ba 0,27 0,39 0,37 0,36 0,35 0,41
Ca 309,46 451,48 457,62 399,37 447,32 470,42
Cu 0,09 33,35 18,92 18,60 0,68 0,14
Fe 0,85 0,63 0,42 0,67 0,95 1,29
K 212,26 288,57 297,47 260,66 295,18 322,76
Li 0,27 0,35 0,36 0,32 0,37 0,41
Mg 55,85 65,08 65,82 57,99 79,25 84,92
Mn 0,07 0,20 0,09 0,11 0,32 0,11
Na 695,19 884,78 925,87 797,17 986,78 1057,03
S 136,00 112,29 97,93 97,93 197,89 206,69
Sr 1,35 1,96 1,97 1,72 1,90 2,05
pH 7,4 7,48
When the performance of the MED combined with the three anti-scaling agents was evaluated, it was
determined AMPS worked the best. When evaluating the performance of the MED paired with the three
anti-scaling agents as prescribed by French Creek software, it becomes apparent how well which anti-
scaling agent performed. The key to determine the best performance is to keep in mind that the purpose of
the anti-scaling agent is to keep ions in suspension whilst increasing their concentration. AMPS fared best
when specifically examining the amount of calcium present in suspension. The explanation as to why the
calcium count when dosing HEDP is lower than when no anti-scaling agent is dosed, is ascribed to the
manner in which it complexes with the precipitation prone ions in order to keep them in suspension not
being accurately measured as calcium ions by ICP-OES techniques.
49
3.7 Conclusion and Recommendations
The designed bench-scale single effect vacuum evaporator exhibited a performance that complied with all
the design requirements. The vacuum inside the vessel was maintained between 0.40 and 0.44 bar, with the
steam temperature varying between 78°C and 80°C. The evaporation rate of the RO-retentate was
volumetrically measured to be 4 L/h (22.3 L/m2/h), and this was verified by the PHREEQ-C modelling
software. The improved structural integrity of the evaporator and vacuum pump cooling adaptations paired
with the doubling of heat exchange surface area increased the evaporation rate from 1.25 L/h to 4 L/h. This
implied that roughly 34.5% of the feed water/RO-retentate was evaporated and can then theoretically be
recycled whilst absorbing some of the abundant heat present at the power station. This was more than the
33.3% reported when using MED to desalinate seawater, which may be attributed to the lower salinity of
the RO-retentate.
The speciation using PHREEQ-C modelling software further showed that evaporation increased the
concentration of scale forming agents. The dosage determined using the software showed that the anti-
scaling agents inside the MED yielded confirmation of their capability of keeping ions in suspension and
not allowing them to scale. When measuring the performance of the anti-scaling agents on this basis, AMPS
fared better than PBTC whilst HEDP yielded concentrations of scale forming ions lower than when no anti-
scaling agent was dosed. This may be ascribed to the complexes that HEDP forms with the scale forming
ions. It is recommended that the evaluation of the anti-scaling agents be further investigated by means of
heat transfer efficiencies, as the active ingredients of anti-scaling agents work with different anti-scaling
mechanisms through various publications (Li, 2010). Furthermore, calcium carbonate scaling is a complex
and poorly understood phenomenon, with the unique composition and nature of the water on which the
anti-scaling agents are being tested (MacAdam & Parsons, 2004).
The application of MED technology paired with the use of anti-scaling agents to treat inland waste waters
will see the MED-evaporator serving a dual purpose. Its primary purpose is to serve as water
producing/recycling technology that can treat brine solutions. The MED reject will be suitable for further
treatment using Eutectic Freeze Crystallisation technology to produce fresh water and salts. The secondary
purpose of the MED technology when applied to a power plant will be to dissipate some of the abundant
heat – alleviating the amount of heat energy the cooling towers must dissipate to some extent, and thus
lowering the amount of cooling water needed. A combination of a thermal technology and membrane
technology could be theoretically utilised to adhere to the zero liquid effluent discharge policy considered
at Eskom, whilst lowering the amount of fresh water needed for cooling purposes.
50
References
Copper Development Association. 2016. Copper: Properties and Applications. [Online]
Available at: copperalliance.org.uk/education-and-careers/education-resources/copper-properties-and-
applications
[Accessed 8 05 2017].
Drela, I., Falewicz, P. & Kuczkowska, S. 1998. New rapid test for evaluation of scale inhibitors, s.l.
Elsevier.
Everson, R., Okolo, G., Neomagus, H. & Dos Santos, J. 2013. X-ray diffraction parameters and reaction
rate modeling for gasification and combustion of chars derived from inertinite-rich coals. Fuel, 109: 148-
156.
Fosso-Kankeu, E., Maluba-Bafubiandi, A., Mamba, B. & Barnard, T. 2011. Prediction of metal-adsorption
behavior in the remediation of water contamination using indigenous micro-organisms. Journal of
Environmental Managment, 10(92): 2786-2793.
Fosso-Kankeu, E., Redelinghuys, J.G., Waanders, F., Rogers, D., Bruinsma, D & Gericke, G. 2016.
Fosso-Kankeu, E., Waanders, F., van Niekerk, D., Rogers, D., Bruinsma, D & Gericke, G. 2016.
Treatment of raw and processed waters from coal power plant using PACl supplemented with cationic
organic polymer and betonite. Parys, International Conference on Advances in Science, Engineering,
French Creek Software. 2017. Cooling Water Treatment - WaterCycle. [Online]

Available at: http://www.frenchcreeksoftware.com/WaterCycle/
[Accessed 15 10 2017].
Greffrath, D.R. 2002. Effect of Scaling and Operation of Thermal Seawater Desalination Plants Germany:
SERCK COMO GmbH.
Kavitha, A., Vasudevan, T. & Prabu, H.G. 2011. Evaluation of synthesized antiscaling agents for cooling
water system application. Desalination, 268: 38-45.
Kirk, W. & Tuthill, A. 1992. Copper-Nickel condenser and Heat Exchenge Systems. The Application of
Copper Nickel Alloys in Marine Systems.
Koretsky, M. 2004. Engineering and Chemical Thermodynamics. 2nd ed. s.l. John Wiley and Sons, Inc.
51
Li, H. 2010. Mineral precipitation and deposition in cooling systems using impaired waters: mechanisms,
kinetics, and inhibition. Pittsburgh: s.n.
Environmental Science and Bio/Technology, 3: 159–169
Ouldhadda, D., Idrissi, A. & Asbik, M. 2002. Heat transfer in non-Newtonian falling liquid film on a
horizontal circular cylinder. Heat and Mass Transfer Journal, 38(7): 719-721.
Roos, T. H., Rogers, D. E. C. & Gericke, G. 2017. Solar-assisted MED treatment of Eskom power station
Shausberger, P., Nowak, J. & Medek, O. 2009. Heat Transfer in Horizontal Falling Film Evaporators.
Dubai, UAE, IDA world congress on desalination and water reuse.
Stratasys, 2013. Fortus pc (polycarbonate) characterization of material properties. Loughborough, UK:

s.n.
Surenda, A. 2014. Eskom Water Planning, MegaWatt Park, Sandton: Eskom/NGO Forum workshop on
Water and Climate Change.
The International Nickel Company, INC. 1963. Corrosion Resistance of the Austenitic Chromium-Nickel
Stainless Steels in Chemical Environments New York: s.n.
USGS. 2016. PHREEQC. [Online]

Available at: https://www.usgs.gov/software/phreeqc
[Accessed 28 10 2017].
s.l. Elsevier.
52
Determination of Anti-Scaling Agent to be used inside
Bench Scale MED
Abstract—Increasing pollution and scarcity of fresh water have prompted the need to develop and
implement effective strategies to treat waste water and reduce water usage. One such a strategy is to
recirculate waste water in a power plant; however, this strategy necessitates additional treatment. The
considered water contains high levels of scale forming ions. Scale formation needs to be controlled to
practically increase the feasibility of recirculation processes. The selected test method to determine anti-
scaling agent induction times was found to be inefficient, and an improvement on the current method was
proposed. The possible causes for the inefficiency of the selected induction time test method was mainly
ascribed to an insufficient heating apparatus setup. The main finding in this study was observed at 75°C
where the anti-scaling agent PBTC performed better than ATMP; the former exhibiting a longer induction
time in reverse osmosis reject water.
Keywords—Anti-scaling agent, Scale, PBTC (2-Phosphonobutane-1,2,4-Tricarboxylic Acid), AMTP

(Amino Trimethylene Phosphonic Acid), Cooling water.
53
4.1 Introduction
Water is one of the most important resources to sustain biological life. It is needed for various applications,
including human consumption, power generation, industrial processes and agriculture. Thus, water
conservation is critical. To further emphasise the need for water conservation, South Africa has recently
experienced a severe drought that led to eight of the nine provinces in the country being declared disaster
areas (Zhuwakinyu, 2016).
Effective water management has consequently become a priority. Power and processing plants are currently
looking for methods that allow for water to be used more efficiently, thus optimising water use. The high
concentrations of scale forming ions in certain waste water streams may pose significant problems when
considering treatment through membrane or thermal technology. Scale formation and build-up in heat
transfer units (heat exchangers) as well as membrane filters causes a decrease in treatment efficiency, which
results in lower water reclamation/re-use. Furthermore, this causes high levels of maintenance, increased
maintenance costs and more downtime to perform maintenance in order to restore efficiency.
Eskom, South Africa’s national power utility, is allocated 2-3% of the national available fresh water to
generate electricity (Surenda, 2014). Eskom is under pressure to minimise its water usage and has set a
target of lowering its usage to 1.34 L of water per kilowatt hour electricity produced by 2020, from the 1.39
L/kWh presently (Fosso-Kankeu et al., 2016). Of the total water intake, 98% is lost via evaporation in the
cooling towers of the power plants. The re-usage of certain streams may potentially reduce the amount of
fresh water intake needed to replenish the amount of water lost due to evaporation. Eskom’s water usage
licences clearly stipulate that they have to adhere to a zero liquid effluent discharge policy. This results in
Eskom investigating all potential methods to aid in adhering to this policy.
The use of multi-effect distillation, amongst various other technologies, is being investigated and thus the
investigation of curbing the most common challenge facing these technologies, namely scaling (Wildebrand
et al., 2007), needs to be better examined and explored.
Scale formation occurs due to changes occuring in the physical conditions of a system. Anti-scaling agents
(a chemical and proven method of countering scale formation) inhibit scale formation by interfering with
scale formation mechanisms. By adding a anti-scaling agent or a combination of more than one anti-scaling
agent, industries in turn could decrease maintenance costs on affected equipment. Anti-scaling agent
performance is limited thus more than one anti-scaling agent can be utilized to increase performance
(MacAdam & Parsons, 2004).
The purpose of this study was to identify problems related to the induction time test method and to establish
anti-scaling agent performance of PBTC and ATMP in terms of induction time for application in the
treatment of waste waters such as reverse osmosis retentate and cooling water from Eskom’s power plants.
54
4.2 Materials and Methods
4.2.1 Sampling and sample storage
Different water streams were sampled at the Grootvlei Power Plant. A cooling water blowdown/bleed off
stream was sampled (concentrated water effluent from the power plant’s cooling towers). The reason for
sampling the cooling water blowdown/bleed off stream was due to its high scale forming ion/mineral
content, hence the consideration of scale inhibition treatment on this stream. To further save costs and
decrease usage of fresh make-up water, the RO-retentate stream was considered for treatment and sampling
was also considered. Samples were stored in a cold environment, which aided in decreasing the sampled
water’s scaling ability as well as preserved the water chemistry.
The cooling water blowdown/bleed off stream (CW-blowdown) and reverse osmosis retentate (RO-
retentate) stream conductivities and pH are displayed in Table 4-1 below and were established for the
sample batch on which static induction tests were performed. Conductivity measurements were performed
with the use of the Lovibond SensoDirect 150 (Lovibond, Germany). pH measurements were conducted
with the Hanna HI8424 portable pH/ORP meter (HANNA instruments, USA).
Table 4-1: Table displaying conductivities and pH's of CW-blowdown and RO-retentate
Water Source: Conductivity pH
CW Blowdown 1.98 mS/cm 7.4
RO Retentate 1.08 mS/cm 7.8
4.2.2 Synthetic control water preparation
A synthetic water sample (control) was prepared by dissolving salts with a concentration of 0.005 mol/dm3
CaCl2 and 0.005 mol/dm3 Ca(HCO3)2 in demineralised water. The pH of the control solution was 7.5. The
purpose of the synthetic water was to compare the difference between the sampled cooling water blowdown,
reverse osmosis reject stream and the synthetic water sample. This helped to note if any of the minerals,
besides those associated with the scale forming mechanism in the sampled water, interacted with the scale
forming mechanism. Each water sample was analysed for additional minerals not associated with scale
formation.
4.2.3 Spiked sample preparation
500 mL of the RO-reject and CW-blowdown samples were spiked with 0.105 g Ca(HCO3)2 and 0.138 g
CaCl2; amounting to the same content as the control solution. Spiked solutions were analysed to investigate
55
the assumption that scale forming ions in the raw samples might have formed scale before static induction
tests are performed.
4.2.4 ICP-OES analysis
An ICP analysis using ICP Expert II, Agilent Technologies 720 ICP-OES (Agilent Technologies, USA)
was performed to determine the concentration of calcium present in the two sampled waters.
4.2.5 EDTA titration
An EDTA (ethylene-diamine-tetra-acetic acid) titration was performed to determine the hardness of the
CW-blowdown and RO-reject water samples. Two titrations were performed, one to determine the total
hardness and the other to determine the calcium hardness. 25 mL of the samples were diluted to 50 mL for
the total and calcium hardness titrations. 0.01 M EDTA solution was used as titrant. An ammonium buffer
solution was used for the total hardness titration and was prepared by adding 16.9 g of ammonium chloride
(NH4Cl) to 143 mL of concentrated ammonium hydroxide (NH4OH). The buffer solution used for calcium
titration was 0.1 M sodium hydroxide (NaOH) solution.
Total hardness was established using an EDTA titration with the pH of the samples maintained at 10±0.1.
1 – 2 mL of ammonium buffer solution was added to the diluted samples. A pH of 10 allowed the EDTA
to react with both calcium and magnesium:
𝐶𝑎2+ + 𝐸𝐷𝑇𝐴4− → 𝐶𝑎𝐸𝐷𝑇𝐴2− Equation 4-1
𝑀𝑔2+ + 𝐸𝐷𝑇𝐴4− → 𝑀𝑔𝐸𝐷𝑇𝐴2− Equation 4-2
Calmagite indicator was used, which changed from a red to a blue at the end point.
Calcium hardness determination was done by EDTA titration, with the pH of the sample maintained
between 12 and 13. 1 - 2 mL of ammonium hydroxide (NH4OH) buffer was added to the solution to maintain
a pH of 12 - 13. The pH was maintained at 13 so that magnesium would not react with EDTA and precipitate
from solution, so only the reaction using calcium would take place. Murexide indicator was used for
calcium titration, which changed from a pink to a purple colour at the end point.
The difference between the total hardness and calcium hardness yielded the amount of magnesium hardness
as:
𝑇𝑜𝑡𝑎𝑙 ℎ𝑎𝑟𝑑𝑛𝑒𝑠𝑠 = 𝐶𝑎 ℎ𝑎𝑟𝑑𝑛𝑒𝑠𝑠 + 𝑀𝑔 ℎ𝑎𝑟𝑑𝑛𝑒𝑠𝑠 Equation 4-3
The total hardness was calculated by using the following equation:
𝐸𝐷𝑇𝐴 𝑉𝑜𝑙𝑢𝑚𝑒 × 1000

𝑇𝑜𝑡𝑎𝑙 𝑜𝑟 𝑐𝑎𝑙𝑐𝑖𝑢𝑚 ℎ𝑎𝑟𝑑𝑛𝑒𝑠𝑠 = 𝑆𝑎𝑚𝑝𝑙𝑒 𝑉𝑜𝑙𝑢𝑚𝑒
Equation 4-4
56
The EDTA volume was calculated by subtracting the volume of EDTA inside the burette before titration
from the EDTA left inside the burette after titration.
4.2.6 Alkalinity test
As the measurement of the amount of carbonates and bicarbonates of the water was necessary to determine
the scaling potential of the water, an alkalinity test was performed. Alkalinity is a measurement of the
concentration of carbonates and bicarbonates in the water as well as a measurement of the buffer capability
to resist changes in pH from the addition of acidic or basic solutions. The presence and concentration of
bicarbonates were determined through the alkalinity test, by titrating water samples with 0.1 M sulphuric
acid solution. The sample batch for which alkalinity was tested had a pH of 8.6 for both CW-blowdown
and RO-reject. A volume of 25 mL was used for the alkalinity titration. First, the sulphuric acid was added
to a titration point of pH 8.3, which represented the elimination of hydroxyl ions (OH-). The volume of acid
in the burette was noted. Sulphuric acid was further added to the water sample until an end point of pH 4.5
was reached. The volume left in the burette was noted and the volume of acid added from 8.3 to 4.5 was
calculated. The total alkalinity was calculated by using the following equation:
Sulphuric acid volume ×1000

𝐴𝑙𝑘𝑎𝑙𝑖𝑛𝑖𝑡𝑦 = 𝑆𝑎𝑚𝑝𝑙𝑒 𝑣𝑜𝑙𝑢𝑚𝑒
Equation 4-5
4.2.7 Sulphate content in samples
Sulphate content was analysed with the use of the HANNA instruments HI 83099 COD and multi-
parameter photometer. The sulphate content was analysed three times for both the CW-blowdown and RO-
reject samples, of which the mean was reported.
4.2.8 Scale potential modelling
The MineSat™ application of French Creek software® was used to model the scaling potential of the water
samples. The model aided in selecting a dosage for which the performance of anti-scaling agents in terms
of induction times could be evaluated within a reasonable time. The software calculated the degree of
supersaturation for a range of pH and temperatures. The model was additionally used for comparison and
interpretation of static induction tests, as well as to draw a thermodynamic result of scaling.
4.2.9 Jar heating system
Jars were heated to temperatures of 25, 40 and 75°C respectively, using a water bath comprised of a heating
element and temperature control. These temperatures were selected to determine the performance of anti-
scaling agents at room temperature (25°), at slightly elevated temperature (40°) and a temperature
mimicking conditions that would occur in the constructed bench scale single effect evaporator (75°). The
57
heated water was then circulated using a pump and recirculation piping to maintain the temperature
surrounding the jars (Figure 4-1 and 4-2).
Figure 4-1: Jar test apparatus with a tray
Figure 4-2: Jar heating system
A jar test apparatus was used to allow for stirring the contents of the glass jars. A tray was filled with the
heated water provided by a water heating bath and placed under the impellers of the jar test apparatus in
which jars were placed. Warm water flowed around the jars, which allowed transfer of heat to the contents
of the glass jar. The circulation maintained a constant temperature inside the tray, and thus the glass jars
once the required temperature was reached.
4.2.10 Data logging method
An Arduino Uno micro-controller was used as a data logging and measurement tool for the measurement
of the resistivity of the samples. The Arduino Uno measured the voltage drop over the water sample, from
which a resistivity can be determined according to a voltage divider calculation:
𝑉𝑜𝑢𝑡 𝑅1
𝑅𝑤𝑎𝑡𝑒𝑟 = Equation 4-6
𝑉𝑠 −𝑉𝑜𝑢𝑡
58
Vs was the voltage of the source, which was 5 V (in this case the output voltage of the Arduino Uno); R 1
was a resistance chosen such that the voltage over the water remained within the range of 0-5 V, in order
for the Arduino to have the capability to read a resistivity for the water. R1 was chosen as 1740 Ω.
Vout, the output voltage, was a variable voltage value, which varied according to the change in resistivity of
the water sample. Having known Vout, R1 and Vs values, Rwater, which was the resistivity of the water (Ω.cm),
was calculated. Conductivity is the inverse of resistivity, which was the measurement of consideration for
the static induction test.
Two copper wires were placed in the water samples, with the objective of measuring the resistivity of the
water. The copper wires were placed 1 cm apart. The measured resistivity was converted to conductivity,
then data was treated with a Savitzky–Golay smoothing algorithm to eliminate noise.
4.2.11 Static induction test procedure
A static induction test was performed in order to determine the difference in time induction performance of
two anti-scaling agents.
Anti-scaling agents were selected according to their effectivity after simulation and evaluation of dosage
profiles using French Creek software. The resulting anti-scalants differ from those used in Chapter 3, as the
water sampled and used to complete Chapter 4 originated from the Grootvlei Power Plant, as opposed to
the Lethabo Power Plant for Chapter 3.
• PBTC (2-Phosphonobutane-1,2,4-Tricarboxylic Acid)

• Belclene 640 (Amino Trimethylene Phosphonic Acid)
The kinetics of scale formation of the synthetic water sample (control solution), CW-blowdown and RO-
reject samples (in the presence as well as absence of ATMP and PBTC) and the spiked CW-blowdown and
RO-reject sample (in the presence as well as absence of ATMP and PBTC) were tested using the static
induction test. The procedure was initiated by heating the bath and was allowed to stabilise at a temperature
such that the temperature inside the tray was either 25, 40 or 75°C respectively.
After experimentation of the modelled dosage, a final dosage of 0.2 ppm of PBTC and ATMP was chosen
for the RO-reject sample and 0.1 ppm dosage of PBTC and ATMP for the cooling water blowdown sample
as well as the control solution, as these dosages indicated a drop in conductivity within a reasonable time.
Glass jars were filled with 500 mL of the water sample and placed in the heating tray. The samples were
dosed with anti-scaling agent while being stirred at a speed of 60 rpm using the jar test apparatus. Two
samples were dosed with ATMP, two with PBTC and two samples were left blank, which amounts to six
samples per experimental. Logging was initiated and allowed to continue for 30 seconds, after which the
pump was switched on and heated water allowed to circulate around the jars. Resistivity was logged every
59
1 minute in the case of 25°C and every 10 seconds for the case of 40°C and 75°C. Experimental runs of 12
hours were performed for 25°C, 6 hours for 40°C and 2 hours for 75°C.
4.3 Results and Discussion
4.3.1 Sample pre-analysis
ICP-OES
The calcium content of the samples was determined using ICP-OES, the value of 83.61 ppm was recorded
in the CW-blowdown sample and 170.4 ppm in the RO-reject sample. The high ion concentration and
hardness of the CW blowdown will potentially cause fouling/scaling, and this is then treated to a certain
extent with RO keeping this in mind. The RO reject generated from the CW Blowdown has a higher
fouling/scaling potential than the CW blowdown and will potentially lead to extensive fouling of
membranes and heat exchange surfaces if it comes in contact when being treated further in downstream
processes. This will effect the choice and complexity of downstream processes. The concentration of other
metals was also determined using ICP-OES and is summarised in Table 4-2:
Table 4-2: Concentration of metals contained in samples

Element CW-blowdown (ppm) RO-reject (ppm)
Al 0.11 0.15
Ba 0.21 0.45
Ca 83.61 170.40
Co 0.01 0.00
Cu 0.14 0.27
K 26.84 53.78
Li 0.15 0.35
Mg 37.95 76.32
Mo 0.10 0.15
Na 87.63 159.76
Pb (220.353 nm) 0.00 0.20
Pb (283.305 nm) 0.05 0.34
Se 1.75 0.04
Sr 0.60 1.27
Tl 1.14 0.01
Zn 0.15 0.20
Water hardness determined by EDTA titration
The results of the EDTA titrations for the CW-blowdown and RO-reject are summarised in Table 4-3,
where the results are given as CaCO3 concentration equivalent. The CW blowdown is treated with RO
hence an increase in hardness is observed. This hardness potentially can cause fouling and scaling when
60
treated further and complicate downstream processes; the impact of hardness on scale formation has been
reported in previous studies (Ben Amor et al., 2004).
Table 4-3: Hardness of water samples

Mg Hardness (mg Total hardness (mg CaCO3/L)
Sample Ca Hardness (mg CaCO3/L)
CaCO3/L)
CW-blowdown 160 115 275
RO-reject 348 210 558
4.3.2 Bicarbonate and carbonate content from alkalinity test
The alkalinity was reported as a concentration equivalent to CaCO3 and determined by means of an
alkalinity test. The concentration of bicarbonates in the CW-blowdown stream was found to be 124.4 mg
CaCO3/L and 60 mg CaCO3/L for the RO-reject sample.
4.3.3 Sulphate content
The sulphate content of the RO-reject was determined to be 400 mg/L, while the sulphate content of the
CW-blowdown was 208 mg/L.
4.3.4 Scaling potential model
The French Creek MineSat™ modelling application was used for scale potential modelling. Scaling
potential models were calculated for a temperature range of 15 - 85⁰C. The pH ranged from 7 - 9 for the
calculation of the scaling potential models. Scaling potential is proportionally related to the degree of
supersaturation.
(Note: For descriptive purposes, the following colour codes of Figures 4-3 and 4-4 are assigned for scaling
potential: blue - low, green - moderate, yellow - moderately high, pink - high and red - severe)
CW-blowdown stream
Figure 4-3 depicts the scaling potential of the CW-blowdown stream calculated at 25⁰C.
61
Degree of supersaturation
Figure 4-3: French Creek MineSat® model, degree of supersaturation for CW
The expected low degree of supersaturation at low temperatures and low pH-values was evident from
Figure 4-3. Below a pH of 7.67, the scaling potential was low. Between a pH of 7.67 and 8, scaling was
moderately low. Above pH 8.33, scaling was excessively bad for a temperature range of 15⁰C - 85⁰C.
1 CW-blowdown at 25⁰C:
At a temperature of 25⁰C and pH 7.4, scaling should not form easily as the scaling potential was low
according to the MineSat™ model, and scaling was kinetically and thermodynamically unfavored in this
case, as scaling could take long and the driving force to form scale was found to be low. The concentration
of Ca2+ and HCO3- was too low to indicate signs of CaCO3 formation.
The scaling potential of the CW-blowdown stream was moderate at a temperature of 40⁰C and a pH of 7.4,
according to Figure 4-3. Although, according to the MineSat®, the scaling potential was not severe; the
potential for scaling in the cooling water system could occur from a thermodynamic perspective, as it was
still moderately high.
According to the model, scaling became severe when calculated at 75⁰C and pH 7.4. The considered severe
scaling potential was indicative of scaling being most likely. The solubility of Ca2+ ions and carbonate ions
was decreased with temperature, as was indicated by the increase in scale potential with temperature in
Figure 4-3. The driving force for scale formation was high in this case, hence favourable from the
thermodynamic and kinetic perspective.
62
RO-reject stream
The scale potential model for the RO-reject stream is displayed in Figure 4-4.
Figure 4-4: French Creek MineSat® model, degree of supersaturation for RO
Similar to the CW-blowdown stream, Figure 4-4 indicated a low scaling potential at low temperatures and
low pH-values. The degree of supersaturation was low, below a 7.67 pH and high above a pH of 8.67.
Scaling was moderate below 27⁰C and between a pH of 7.8 and 8.
4 RO-reject at 25⁰C:
At a pH of 7.8 and at a temperature of 25⁰C, scaling potential was low and unlikely to occur, therefore
thermodynamically and kinetically unfavoured. The concentration of Ca2+ and HCO3- was too low to show
signs of CaCO3 formation at this temperature.
The RO-reject stream showed a moderate potential to form CaCO3 scale at 40⁰C with a pH of 7.8, although
not severe; yet a high scaling potential was considered, as there was a thermodynamic drive to form scale
according to the French Creek simulation model shown in Figure 4-4.
Evaluation of the scaling potential viewed in Figure 4-4 at 75⁰C and pH 7.8 showed a severe tendency to
scale according to the high degree of supersaturation. As a result, the stream should be considered as highly
thermodynamic and kinetically favourable.
63
Summary of scaling potential
The description of scaling potential and degree of supersaturation are summarised in Table 4-4:
Table 4-4: Scaling potential and degree of supersaturation for CaCO3 for CW-blowdown and RO-reject
streams
Scaling potential Degree of supersaturation (Calcite)
Temperature (⁰C)
CW-blowdown RO-reject CW-blowdown RO-reject
25 Low low 0.98 1.2
40 Moderately high high 1.59 1.93
75 Severe severe 3.79 4.36
* Note: For descriptive purposes in Table 4-4, the following colour codes of Figures 4-3 and 4-4 are assigned for
scaling potential: blue - low, green - moderate, yellow - moderately high, pink - high and red - severe.
The degree of supersaturation was determined using the French Creek MineSat™ application. The scaling
tendency for the RO-reject stream was determined to be higher than the CW-blowdown at all three
temperatures according to the degree of supersaturation; this was due to the concentration of the calcium
ions being greater in the RO-reject stream. The calcium ion concentration was determined by ICP-OES as
170.40 ppm present in the RO-reject stream and 83.61 ppm present in the CW-blowdown stream.
4.3.5 Dosage profile modelling for ATMP and PBTC anti-scaling agents
In addition to the scaling potential model, a dosage profile model was generated by French Creek’s software
MineSat™ application and illustrated in Figures 4-5 and 4-6. The dosage profile gave a general idea of the
dosage that was needed for the static induction tests. The dosage was considered at the highest temperature
of experimentation.
64
Figure 4-5: French Creek MineSat® model, dosage profile of ATMP for RO, calculated at 75⁰C
Figure 4-6: French Creek MineSat® model, dosage profile of PBTC for RO, calculated at 75⁰C
4.3.6 Static induction test:
Anti-scaling agents (PBTC & ATMP) were evaluated at 25⁰C, 40⁰C and 75⁰C respectively, to determine
the difference in performance in terms of induction time. Each of the samples was divided into three
batches, a blank (without anti-scaling agent dosage), ATMP dosed, and a PBTC dosed batch.
The first gradual increase in conductivity was due to the increase in energy in the form of heat, whereby
the samples were heated from ambient temperature to the considered temperature of either 25⁰C, 40⁰C, or
75⁰, as observed in Figure 4-7 to Figure 4-21. Ions moved more readily upon elevated temperatures, thereby
increasing conductivity. In some cases, the induction test at 25⁰C showed no initial rise in conductivity; this
was due to the ambient temperature varying with time and in the case where no initial rise in conductivity
was observed; the ambient temperature was much closer to 25⁰C than in other cases. A decrease in
65
conductivity occurred when the ion content of the water decreased, which was the case when scale
precipitated from the water.
The compositions of potential scale forming waters and the methods of testing anti-scaling agent
performance vary greatly throughout the literature (MacAdam & Parsons, 2004). As a result, comparison
to the scaling potential model generated by French Creek’s MineSat™ application provided valuable
insight. The general trend of what was expected can be related to the plot of conductivity over time given
by Muryanto et al. (2014). The difference observed is that an initial rise in conductivity was not observed
in the study done by Muryanto et al. (2014), although the trend after the rise became similar to the study.
There was a sudden rise of conductivity which eventually stabilised and then dropped due to the
precipitation of CaCO3 scale.
Note: Conductivities are not as accurate due to noise, polarisation on the wire probes and electro-magnetic
induction present from the use of an Arduino set-up; the changes in conductivity can however be considered
for determining an induction time in the case of this experiment.
4.3.7 Static induction test at 25⁰C
Figure 4-7 represents an induction test of the control solution at 25⁰C.
Figure 4-7: Conductivity over time (control solution 25°C)
66
The induction time test showed no drop in conductivity for the synthetic water control solution; this was
due to the low scaling potential of the control solution at 25°C and a moderately low initial pH of 7.5, which
further added to a low scaling potential.
In Figure 4-8, the induction test performed on the CW-blowdown sample at 25⁰C is given.
Figure 4-8: Conductivity over time (CW sample 25°C)
As depicted in Figure 4-8, a slight drop in conductivity was observed between 226 min and 351 min, and
although this had occurred, it was not large enough to draw a concrete conclusion in terms of CaCO3
precipitation . The drop in conductivity was due to indeterminable external influences in this case. In
comparison to Figure 4-10, which represents the induction test for the spiked CW-blowdown sample; it had
a higher scaling potential than that of the standard CW-blowdown sample, and no drop in conductivity
occurred. Furthermore, Figure 4-3 showed that, according to the scale potential model for the CW-
blowdown stream, the scaling potential was low; therefore no precipitation occurred for the CW-blowdown
stream at 25°C.
An induction test performed on the RO-reject sample at 25°C is given in Figure 4-9:
67
Figure 4-9: Conductivity over time (RO sample 25°C)
The drop in conductivity observed in Figure 4-9 exists only for the blank batch. The drop in conductivity
was not accredited to the precipitation, as compared to Figure 4-11, which represents the spike RO-reject
sample, which had a higher concentration of calcium and carbonate ions. Hence, a higher scaling potential
than the standard RO-reject sample showed no drop in conductivity for the non-dosed blank batch. The
scale potential model depicted in Figure 4-4 showed a low potential for scale formation, which further
infers that the conductivity drop was not accredited to the formation of CaCO3, but rather to indeterminable
external influences in this case.
Figure 4-10 represents the induction test for the spiked CW-blowdown sample at 25°C.
68
Figure 4-10: Conductivity over time (CW spiked sample 25°C)
Figure 4-11 represents the induction test for the spiked RO-reject sample at 25°C.
Figure 4-11: Conductivity over time (RO spike sample 25°C)
There was no significant change in conductivity for either the CW-blowdown or RO-reject samples and
blank batches at a temperature of 25°C (Figures 4-10 and 4-11). At 25°C and for the considered time, all
samples (control solution, CW-blowdown, RO-reject, spiked CW-blowdown and spiked RO-reject) were
undersaturated with potential scale forming ions (Ca2+ and CO32-).
For each sample, the blank, PBTC dosed and ATMP dosed batches were duplicated. In some cases, spikes
associated with the electronics of the data logging method occurred, and in these cases, the duplicate where
the electronic spike did not occur was selected for representation. The trends were similar for most cases,
except in the case of electronic spikes. However, this trend remained similar prior to and after a spike had
occurred for the duration of the experimental run time.
pH and actual conductivity and summary of results for 25⁰C
Table 4-5 details the results of the pH and actual conductivity measurements before the start of the induction
test as well as directly after the induction test was completed.
69
Table 4-5: Initial and final measurements of pH and actual conductivity at 25⁰C
Samples at 25⁰C pH Conductivity
(mS/cm)
Initial Final Initial Final
Control solution 7.5 7.6 0.6 0.8
ATMP dosed control solution 7.4 7.5 0.7 0.8
PBTC dosed control solution 7.5 7.5 0.7 0.8
CW-blowdown 7.4 7.5 1.9 2.0
ATMP dosed CW-blowdown 7.5 7.5 2.0 2.0
PBTC dosed CW-blowdown 7.4 7.4 1.9 1.9
RO-reject 7.8 7.8 1.1 1.2
ATMP dosed RO-reject 7.7 7.8 1.0 1.0
PBTC dosed RO-reject 7.7 7.8 1.0 1.0
Spiked CW-blowdown 7.7 7.6 2.5 2.5
ATMP dosed spiked CW-blowdown 7.7 7.7 2.5 2.5
PBTC dosed spiked CW-blowdown 7.8 7.6 2.5 2.5
Spiked RO-reject 7.8 7.7 1.7 1.7
ATMP dosed spiked RO-reject 7.8 7.8 1.8 1.8
PBTC dosed spiked RO-reject 7.8 7.8 1.7 1.7
* Note: pH accuracy within ±0.1 and conductivity accuracy was determined within
±0.1 mS/cm
The pH and conductivity remained relatively similar before and after the duration of the induction test. This
is indicative that no change in conditions of any of the samples occurred throughout the duration (12 hours)
of the induction test in the case of 25⁰C.
70
The initial increase in conductivity observed at 40⁰C was higher than that observed at 25°C. Figure 4-12
represents an induction test of the control solution at 40⁰C.
The control solution showed no signs of CaCO3 formation, as a drop in conductivity was absent. The
amount of calcium and carbonate ions was too low for the temperature and pH conditions during an
experimental run time of 6 hours.
A representation of the induction test performed on the CW-blowdown sample at 40°C is given in
Figure 4-13.
71
Figure 4-14 represents an induction test of the RO-reject sample at 40⁰C.
In Figure 4-13 and Figure 4-14, the rate of change in conductivity decreased during the initial increase in
temperature of the water, which could be indicative of the occurrence of precipitation, and could be the
case for the CW-blowdown sample (Figure 4-13). This behaviour occurred in the blank batch. In contrast,
in the RO-reject sample (Figure 4-14), this behaviour occurred in the PBTC dosed batch and not in the
blank batch, which was inconsistent and contradicted the idea that anti-scaling agents increase induction
time.
72
Figure 4-3 (scaling potential model for CW-blowdown stream) and Figure 4-4 (scaling potential model for
RO-reject stream) showed a moderately high and high potential to form scale. The spiked samples for the
CW-blowdown (Figure 4-15) and RO- reject (Figure 4-16) had a higher potential to form scale than the
standard samples due to the additional spiked amounts of calcium and bicarbonates, which showed no signs
of scale formation, as there existed no observed drop in conductivity. Although the potential to form scale
existed according to the scale potential models, a solid conclusion that scale forms could not be drawn due
to the inconsistency between the spiked and standard samples in both CW-blowdown and RO-reject cases.
Figure 4-15 represents an induction test of the spiked CW-blowdown sample at 40⁰C.
Figure 4-15: Conductivity over time (CW spike sample 40°C)
Figure 4-16 represents an induction test of the spiked RO-reject sample at 40⁰C.
73
No signs of CaCO3 formation existed in the spiked CW-blowdown or RO-reject samples, as there was no
drop in conductivity. The amount of calcium and carbonate ions was too low for the temperature and pH
conditions during an experimental run time of 6 hours. At 40°C, there was no observable change in
conductivity after temperature set point was reached, and given the same situation as for 25⁰C; the solution
and samples were both below supersaturation at 40⁰C and unable to undergo precipitation of CaCO3 for the
considered time of 6 hours.
All samples were run in duplicate, and the results observed were in most cases similar in both replicates. In
some cases, spikes and noise associated with the electronics of the data logging method occurred, and in
these cases, the duplicate where the electronic spikes and electronic noise did not occur were selected for
representation. The trends were similar for most cases, except in the case of electronic spikes, but the trend
remained similar prior to and after a spike had occurred for the duration of the experimental run. Some
trends did not show similarity, and in these cases the most representative trend of the theory of anti-scaling
agent performance was considered.
The initial and final measurements of pH and actual conductivity for the induction time test at 40⁰C are
detailed in Table 4-6.
74
Samples at 40⁰C pH Conductivity (mS/cm)
CW-blowdown 7.5 8.3 2.0 2.4
RO-reject 7.8 8.2 1.1 1.2
* Note: pH accuracy within ±0.1 and conductivity accuracy was determined within ±0.1
mS/cm
Conductivity was determined to be higher for all samples after the experimental run relative to the start of
the experimental run. The increase in conductivity was due to a temperature increase of the samples, as ions
were able to move more freely with a temperature elevation; hence the conductivity increased.
The pH was noticeably increased after the induction test. Coetzee et al. (2017) claim that this was due to
the release of CO2 gas as temperature was increased. Coetzee et al. (2017) further explains that a decrease
in the pH of the solution signified an onset of CaCO3 precipitation, and in their experiment the pH dropped
well below the initial level. Although the pH was lower than the initial value for the Coetzee et al. (2017)
75
experiments, a conclusion based on pH cannot be inferred, as the considered method did not entail
intermediate logging of the pH, as the possibility of a rise and fall in pH may have existed. In terms of
conductivity, no signs of precipitation had occurred at 40⁰C.
Experimental induction tests on the samples and solutions at 75⁰C produced a diversity of results.
Figure 4-17 represents an induction test of the control solution at 75⁰C.
As seen in Figure 4-17, ATMP performed better than PBTC in terms of induction time for the control
solution, as no drop in conductivity occurred for ATMP. The sample containing PBTC showed a very slight
decrease in conductivity, whereas the decrease in conductivity of the blank was steeper, starting at 25 min.
The slight decrease in conductivity could be an indication of the occurrence of precipitation, but the drop
was not large enough to draw a solid conclusion.
Figure 4-18 represents an induction test of the CW-blowdown at 75⁰C.
76
In the case of the CW-blowdown sample (Figure 4-18), it appeared that PBTC performed better than
ATMP, as the conductivity dropped much later (62 mins) than that of the ATMP sample (12 mins), while
an earlier drop (10 mins) was recorded for the blank batch, after which the conductivity remained constant
for the rest of the experiment. However, there was no significant difference between the trend observed in
the presence of ATMP and the one observed for the blank batch. PBTC performed better than ATMP in the
case of the spiked CW-blowdown solution, further supporting the performance of PBTC over ATMP.
The scale potential model predicted that scaling was severe at 75°C and a pH of 7.4 (depicted in Figure
4-3), which was confirmed by the results of the induction test for the CW-blowdown sample at 75°C, as
scaling was observed in all samples.
Figure 4-19 represents an induction test of the RO-reject at 75⁰C.
77
For the RO-reject sample (Figure 4-19), the conductivity decreased before the temperature reached 75°C.
At high temperatures, scaling becomes significant, which was further supported by the scaling potential
model at the conditions of pH 7.8 and 75°C (depicted in Figure 4-4). It can be observed in Figure 4-19 that
all the major drops in conductivity were recorded almost simultaneously, although a slight delay by the
PBTC could also be identified.
Figure 4-20 represents an induction test of the spiked CW-blowdown at 75⁰C.
Figure 4-20: Conductivity over time (CW spike sample 75°C)
78
In Figure 4-20, the blank CW spiked solution showed precipitation at 62°C and 11 mins. In samples
containing ATMP, conductivity drops started at 60°C and 12 mins, whereas samples containing PBTC
showed no drop in conductivity throughout; hence there were no signs of precipitation.
Figure 4-21 represents an induction test of the spiked RO-reject at 75⁰C.
The spiked RO-reject sample showed no signs of drop in conductivity for the two considered anti-scaling
agent dosed samples. The blank spiked RO-reject sample started to drop in conductivity at 65 mins.
Although the experiments were run in duplicate, in some cases, spikes and noise associated with the
electronics of the data logging method occurred, and in these cases the duplicate where the electronic spikes
and electronic noise did not occur were selected for representation. The trends were similar for most cases,
except in the case of electronic spikes, but the trend remained similar before and after a spike had occurred
for the duration of the experimental run.
The initial and final measurements of pH and actual conductivity for the induction time test at 75⁰C are
given in Table 4-7.
79
Samples at 75⁰C pH Conductivity (mS/cm)
CW-blowdown 7.4 8.0 2.0 2.2
RO-reject 7.8 8.2 1.1 1.2
* Note: pH accuracy within ±0.1 and conductivity accuracy was determined within ±0.1
mS/cm
Both the pH and actual conductivity were higher after the duration of the induction test at 75⁰C in relation
to the initial conductivity and pH. As was the case for the induction test at 40⁰C, a rise in pH was due to the
slow release of CO2 during the heating stage of the induction test (Coetzee et al., 2017).
The final pH and conductivity were lower than the initial parameters for the spiked RO-reject, where the
drop in pH of the solution signified an onset of CaCO3 precipitation. In terms of conductivity, precipitation
had occurred in the blank batches of all the samples for the case of 75⁰C, as can be seen in Figure 4-17 -
80
Figure 4-21. While in the presence of PBTC and ATMP, inhibition of precipitation was observed, with
PBTC exhibiting an overall better performance in terms of longer induction time.
PBTC and ATMP are both phosphonate type scale inhibitors, and the mechanism/reason for PBTC
outperforming ATMP is not clear, as research on the comparison between PBTC and ATMP is not widely
available. Väisänen (2011) mentions that the smaller the molecular size of a phosphonate, the better its
efficiency due to the adsorption of the inhibitor onto crystal growth sites. This could lead to a blockage for
the adsorption of other molecules in the case of larger molecules.
The formula for density, the molecular size in terms of volume of one mole of ATMP and PBTC, is as
follows:
𝑚
𝜌= 𝑉
Equation 4-7
Where 𝜌 is density (kg/m3), 𝑚 is mass (kg) and 𝑉 is volume (m3). The density of ATMP is 1300 kg/m3 and
has a molecular weight of 299 g/mol, which correlates to a volume of 2.3 × 10−3 m3. The density of PBTC
is 1250 kg/m3 and has a molecular weight of 270.13 g/mol, which correlates to a volume of 2.16 × 10−3
m3. The volume of PBTC was smaller than ATMP, which means that PBTC performed better due to the
ability of PBTC blocking more adsorption of other molecules onto the scale forming crystals.
4.3.10 General discussion of possible causes of results
The rate of heat transfer was not high enough, which affected the prediction of the performance of anti-
scaling agents within reasonable time, as precipitation may have occurred before reaching the targeted
temperature. Thus, an accurate evaluation of the performance of anti-scaling agents could not be achieved,
as in the case of this study, precipitation occurred during the heating process.
4.4 Conclusions
Assessing the scaling potential of the different water streams, there was little change in water sample
conditions when the temperature of the sample was changed from ambient temperature to 25°C. The
conductivity remained constant for all samples, which indicated that the samples were below supersaturated
conditions. When temperatures reached 40°C, conductivities remained fairly constant for all samples.
Increasing the temperature to 75°C showed a diversity of trends. In some cases, conductivity dropped
during the initial temperature increase phase, while in others a drop in conductivity was observed only when
the targeted temperature was reached. However, in general the anti-scaling agent PBTC performed better
than ATMP, both for RO-reject water and CW-blowdown. Overall, it is important to note that, in the context
of a laboratory study aiming to compare the performance of anti-scaling agents, one must ensure that the
solutions considered are supersaturated if results are to be obtained in a short time.
81
References
Al-Roomi, Y.M. and Hussain, K.F. 2016. Potential kinetic model for scaling and scale inhibition
mechanism. Desalination, 393: 186-195.
Ben Amor, M., Tlili, M.M., Manzola, A.S. 2004. Influence of water hardness, substrate nature and
temperature on heterogenous calcium carbonate nucleation. Desalination, 166: 79-84.
Coetzee, P., Yacoby, M., Howell, S. and Mubenga, S. 2017. Scale reduction and scale modification effects
induced by Zn and other metal species in physical water treatment. Water SA, 24 (1 January 1998): 79.
Demadis, K. & Ketsetzi, A. 2007. Degradation of Phosphonate‐Based Scale Inhibitor Additives in the
Presence of Oxidizing Biocides: “Collateral Damages” in Industrial Water Systems. Separation Science
and Technology, 42(7): 1639-1649.
Ferguson, R. 2017. MineSAT - Mining, Process Water Modeling, Chemistry & Treatment. [online]
Frenchcreeksoftware.com. Available at: http://www.frenchcreeksoftware.com/MineSAT/ [Accessed 29
Oct. 2017].
Ferguson, R., Ferguson, B. and Stancavage, R. 2011. Modeling Scale Formation and Optimizing Scale-
Inhibitor Dosages. IDA Journal of Desalination and Water Reuse, 3(2): 30-39.
Fosso-Kankeu, E., Redelinghuys, J.G., Waanders, F., Rogers, D., Bruinsma, D & Gericke, G. 2016.
Flynn, D. 2009. The Nalco water handbook. 3rd ed. New York: McGraw-Hill.
Huchler, L. 2008. Update your industrial water treatment operations. Hydrocarbon Processing, 87(12):
83-83.
Koutsoukos, P., Kofina, A. and Kanellopoulou, D. 2007. Solubility of Salts in Water: Key Issue for Crystal
Growth and Dissolution Processes. ChemInform, 38(40).
Landie, C. 2016. South Africa's water supply status. Water & Sanitation Africa, (3): 58-60.
Liu, G., Xue, M., Liu, Q. and Zhou, Y. 2016. Acrylic acid–allylpolyethoxy carboxylate copolymer as a
environmentally friendly scale inhibitor (part II). Clean Technologies and Environmental Policy, 19(3):
917-924.
Environmental Science and Bio/Technology, 3: 159–169
82
Muryanto, S., Bayuseno, A., Ma’mun, H., Usamah, M., Jotho. 2014. Calcium Carbonate Scale Formation
in Pipes: Effect of Flow Rates, Temperature, and Malic Acid as Additives on the Mass and Morphology of
the Scale. Procedia Chemistry, 9: 69-76.
Nehrke, G. 2007. Calcite Precipitation from Aqueous Solution: Transformation from Vaterite and Role of
Solution Stoichiometry. Dissertation. Utrecht University.
Oilfieldwiki.com. 2017. Scale inhibitor - OilfieldWiki. [online] Available at:

http://www.oilfieldwiki.com/wiki/Scale_inhibitor [Accessed 12 Jul. 2017].
Rieger, J., Thieme, J. and Schmidt, C. 2000. Study of Precipitation Reactions by X-ray Microscopy:
CaCO3 Precipitation and the Effect of Polycarboxylates. Langmuir, 16(22), pp.8300-8305.
Sigma-Aldrich. 2017. Poly(acrylic acid) 323667. [online] Available at:

http://www.sigmaaldrich.com/catalog/product/aldrich/323667?lang=en&region=ZA [Accessed 12 Jul.
2017].
Sumin, Z. & Kan, Z. 2015. Formation Mechanism and Control Techniques of Calcium Carbonate Scale
in the Langjiu Geothermal Field Tibet. In Proceedings World Geothermal Congress. Melbourne,
Australia.
Surenda, A. 2014. Eskom Water Planning. MegaWatt Park, Sandton: Eskom/NGO Forum workshop on
Water and Climate Change.
Väisänen, H. 2011. CaCO3 Scale inhibition in paper making processes - Evaluation of testing methods
and inhibitor performance. Undergraduate. Tampere University of Technology.
Wilson, D. & Harris, K. 2017. Steps for Jar Testing Scale Inhibitors in Oil, Gas Applications. [online]
Waterworld.com. Available at: http://www.waterworld.com/articles/iww/print/volume-11/issue-1/feature-
editorial/steps-for-jar-testing-scale-inhibitors-in-oil-gas-applications.html [Accessed 6 Jul. 2017].
Zhang, Y., Yin, H., Zhang, Q., Li, Y. & Yao, P. 2016. Synthesis and characterization of novel polyaspartic
acid/urea graft copolymer with acylamino group and its scale inhibition performance. Desalination, 395:
92-98.
s.l. Elsevier.
83
Conclusions and Recommendations
5.1 Conclusions
Eskom currently uses nano-filtration and reverse osmosis as water reclamation and purification
technologies in its plants, producing saline water streams that need to be disposed of to adhere to the zero
liquid effluent discharge stipulations in its water usage licenses. Due to the abundant heat present at
Eskom’s power plants and the versatility of Multi-Effect Distillation, a possible theoretical sequence of
membrane and thermal technologies could be implemented to approach a zero waste water reclamation
process: Firstly, the use of nano-filtration to remove bivalent scaling ions from feed water then treating
nano-filtration permeate using RO; secondly, the treatment of nano-filtration retentate to precipitate scaling
ions; thirdly, the combination of treated nano-filtration retentate with RO-retentate to be treated using
Multi-Effect Distillation – using a method of heating that harnesses some of the abundant heat that needs
to be dissipated on the power plant, as well as the correct selection and dosage of anti-scaling agents. The
last step of this sequence would be to make use of Eutectic freeze crystallisation to separate the reject from
the multi-effect distillation into fresh water and dry salt.
In this study, a bench scale single effect evaporator (using the same principles and conditions produced in
multi-effect distillation) was designed and constructed to examine the water recovery potential from waste
water streams produced at Eskom’s power plants when using the correct anti-scaling agents. Reverse
osmosis retentate from the Lethabo Power Plant was treated in the bench scale single effect evaporator with
and without the use of scale inhibiting agents.
The bench scale single effect vacuum evaporator exhibited a performance that complied with all the design
requirements. The vacuum inside the vessel was maintained between 0.4 and 0.44 bar, with the steam
temperature varying between 78°C and 80°C. The evaporation rate of the RO-retentate was volumetrically
measured to be 4l/h (22.3L/m2/h), and this was verified by PHREEQ-C modelling software. The improved
structural integrity of the evaporator and vacuum pump cooling adaptations combined with the doubling of
heat exchange surface area increased the evaporation rate from 1.25l/h to 4l/h. This implies that roughly
34.5% of the feed water/RO-retentate was evaporated that may theoretically be recycled while absorbing
some of the abundant heat present at the power station to in turn lower the amount of cooling water needed.
The speciation using PHREEQ-C modelling software further showed that evaporation increased the
concentration of scale forming agents. The dosage of the software (French Creek) determined anti-scaling
agents inside the MED yielded confirmation of their capability of keeping scale forming ions in suspension
and not allowing them to scale. When measuring the performance of the anti-scaling agents on this basis,
AMPS fared better than PBTC, while HEDP yielded concentrations of scale forming ions lower than when
no anti-scaling agent was used. This may be ascribed to the nature of complexes that HEDP forms with the
scale forming ions.
84
The MED reject will be suitable for further treatment using a technology such as Eutectic freeze
crystallisation technology to produce fresh water and salts. Thus, based on the literature, a combination of
a membrane technology and a thermal technology could be theoretically utilised in order to adhere to the
zero liquid effluent discharge policy adopted by Eskom, whilst lowering the amount of fresh water needed
for cooling purposes.
When evaluating selected anti-scaling agents in the laboratory, certain conclusions came to light, and these
will be described according to the temperatures at which the experiments were performed:
There was little change in conditions when the temperature of the sample was changed from ambient
temperature to 25°C. The conductivity remained constant for all samples, which indicated that the samples
were below supersaturated conditions.
When the temperature reached 40°C, the conductivities remained fairly constant for all samples.
Increasing the temperature to 75°C showed a diversity of trends. In some cases, conductivity dropped
during the initial temperature increasing phase, while in others a drop in conductivity was observed only
when the targeted temperature was reached.
However, in general, the anti-scaling agent PBTC performed better than ATMP, both for RO-retentate and
CW-blowdown. Overall, it is important to note that, in the context of laboratory study, aiming to compare
the performance of anti-scaling agents, one must ensure that the solutions considered are supersaturated if
results are to be obtained in a short time.
5.2 Recommendations
With regard to the MED paired with anti-scaling agents, it is recommended that the evaluation of the anti-
scaling agents be further investigated by means of heat transfer efficiencies, as the active ingredients of
anti-scaling agents remain poorly understood when considering the unique composition and nature of the
water on which the anti-scaling agents are being tested.
An improved method of testing the anti-scaling agents is proposed, whereby the induction tests are
performed using a closed jar cylinder. The water should be at the conditions required prior to dosing anti-
scaling agents. A closed jar can better aid in controlling the temperature and pressure while keeping the
contents constant by overcoming excessive evaporation of water. An accurate and reliable measuring device
should be used, and more variables should be taken into consideration, such as pH, which can allow
monitoring of the behaviour of critical elements or compounds in solution, such as bicarbonate, carbonate
and carbon dioxide content. Increasing the rate of heat transfer could allow the prediction of the
performance of anti-scaling agents within reasonable time, as precipitation may be occurring before
reaching the targeted temperature. Accurate evaluation of the performance of anti-scaling agents could be
achieved by designing a system that will shorten the heating time required to reach the targeted temperature,
85
as in the case of this study, precipitation occurred during the heating process, affecting proper evaluation
of the performance of anti-scaling agents.
86
Appendix A: MED Designs
Figure A-1: 3D Assembly of designed MED
87
Figure A-2: MED detail design box/evaporation chamber
88
Figure A-3: MED detail design lid of box/evaporation chamber
89
Figure A-4: MED detail design drip tray
90
Figure A-5: MED detail design rib structure dimensions
91
Appendix B: Academic and Technical Outputs
Figure B-1: Scaling conditions inside improved bench-scale single effect vacuum evaporator
92
93
94
95
96
97
Figure B-2: Determination of suitable anti-scaling agent to inhibit scale formation of water
processed in coal power plant system
98
99
100
101
102
Figure B-3: Determination of water evaporation rate in an assembled bench scale MED and
impact of anti-scaling agents on the morphology of scale
103
104
105
106
107

7.1.11.7.4 Redelinghuys JG 24071919

Uploaded by

Copyright:

Available Formats

7.1.11.7.4 Redelinghuys JG 24071919

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

7.1.11.7.4 Redelinghuys JG 24071919

Uploaded by

Copyright:

Available Formats

Design and application of a bench scale

evaporator for the treatment of coal

Dissertation submitted in fulfilment of the requirements for the

Supervisor: Prof E Fosso-Kankeu

Graduation Ceremony: May 2019

Signed on this, 7th day of November in Potchefstroom.

INITIALS AND SURNAME

Academic and Technical Outputs

Redelinghuys J, Fosso-Kankeu E, Waanders F, Rogers D, Gericke G. 2017. Scaling conditions inside

ESKOM, for sponsoring and funding this investigation.

4.3.2 Bicarbonate and carbonate content from alkalinity test ......................................................... 61

Figure 3-24: Langelier saturation index .................................................................................................48

EDTA Ethylene-diamine-tetra-acetic acid

PAA Polyacrylic acids

PASP Polyaspartic acid

AMPS Acrylic Acid-2-Acrylamide-2-Methylpopane Sulfonic Acid Copolymer

HEDP 1-Hydroxy ethylidene-1,1-Diphosphonic Acid

PBTC 2-Phosphonobutane-1,2,4-Tricarboxylic Acid

PESA Polyepoxysuccinic acid

PMA Polymaleic acids

MED Multi-Effect Distillation

1.2 Problem Statement

1.3 Aim and Objectives

This aim was achieved by the following objectives:

1.4 Research Questions

The research was aimed at addressing the following questions:

1.5 Dissertation Structure

CHAPTER 1: Background and Motivation

CHAPTER 2: Literature Review

CHAPTER 3: Design, configuration and construction of the bench scale MED

CHAPTER 5: Conclusions and Recommendations

2.2 Mineral Scale Formation

2.3 Scale Formation Mechanisms

𝛥𝐺𝑗 = 𝛥𝐺𝑏𝑢𝑙𝑘 + 𝛥𝐺𝑠𝑢𝑟𝑓𝑎𝑐𝑒 Equation 2-1

𝛥𝐺𝑏𝑢𝑙𝑘 = −𝑗𝑘𝑇ln𝑆 Equation 2-2

𝑆 which is the saturation ratio can be expressed as:

𝛥𝐺𝑠𝑢𝑟𝑓𝑎𝑐𝑒 = 4𝜋𝑟 2 γ Equation 2-5

2.3.3 Crystal growth

∆𝜇 = 𝜇𝑚𝑒𝑑𝑖𝑢𝑚 − 𝜇𝑐𝑟𝑦𝑠𝑡𝑎𝑙 Equation 2-7

2.3.5 Kinetics of Sulphate Removal through Gypsum Crystallization

2.4.2 Crystal distortion

2.4.3 Dispersion and surface conditioning

2.5 Characteristics of Anti-Scaling Agents

2-phosphonobutane-1,2,3- Scale formation reduction by 25 mg/l Yang et al. (2001)

Inhibition of precipitation 10 mg/l Harris and Marshall (1981)

2.6 Principles of the different treatment methods:

Figure 2-3: MED schematic

Figure 2-4: Multi-stage flash distillation (MSF) schematic diagram

2.6.2 Eutectic freeze distillation

• No need for chemical pre-treatment, eliminating the discharge of toxic chemicals

2.6.3 Membrane technology

Figure 2-5: Nano-filtration schematic

Figure 2-6: Reverse osmosis process diagram

2.6.4 Comparison of different technologies

Keywords—MED, scale formation, RO-retentate, PHREEQ-C, French Creek.

3.2 Operating Conditions

Table 3-1: Operating parameter targets of evaporation chamber

These three key components are depicted in Figure 3-2:

3.4 Design and Configuration

3.4.1 Evaporation chamber