An Introduction to Brine Waste

1

An Introduction to Brine Waste

Christos Charisiadis 2014

2

Contents

1. Introduction 4

2. Feedwater intake and environmental impacts 4

3. Characteristics of desalination brine 6

3.1 Multi-Stage Flash Desalination (MSF) 6

3.1.1 Seawater Intake 6

3.1.2 Discharge of Brine Containing Additives 6

3.1.3 Physical Properties of Brine 8

3.1.4 Biocides 8

3.1.5 Antiscalants 9

3.1.6 Antifoaming Agents 12

3.1.7 Corrosion Inhibitors and Corrosion Products 13

3.2 Multi-Effect Distillation Desalination (MED) 14



3.2.3 Physical Properties of Brine 16

3.2.4 Biocides 16


3.2.6 Antifoaming Agents 18

3.2.7 Corrosion Inhibitors and Corrosion Products 19

3.3 Reverse Osmosis (RO) 19



3.3.3 Biocides 20

3.3.4 Coagulants 21


3.3.6 Membrane Cleaning Agents 22

3.3.7 Corrosion Products 23

3.3.8 Dechlorination 23

4. Receiving coastal environment 24

4.1 Discharge Mode 24

4.2 Receiving Environment 25

4.3 Discharge and Site Variability 27

4.4 Bathymetry and Gravity Currents 27

5. Brine disposal; A look at the possible options 29

3

5.1 Inland plants 29

5.1.1 Deepwell injection 29

5.1.2 Evaporation ponds 31

5.1.3 Solar ponds 33

5.1.4 Zero Liquid Discharge (ZLD)/ Degremont 34

5.2 Coastal options; Disposal strategies and near field effects 37

5.2.1 Coastal hydrodynamic concept 38

5.2.2 Disposal Alternatives 38

6. Summary & Conclusions 42

4

1. Introduction

Available supply of good quality water for drinking water and industrial has met a

steep increase primarily because of high demand and severe droughts. As a result

many companies are looking into brackish and seawater to satisfy these demands.

These sources typically require a treatment process that generates concentrated

brine as a liquid residual. Due to increasingly strict regulations, conventional brine

waste management strategies such as surface water discharge, deep well injection

or discharge to wastewater treatment facilities may not be feasible in the near

future.

Brine is a very loose term in water treatment. When it comes to desalination, brine

is the liquid residue generated by the treatment process. It contains high

concentrations of sodium chloride and other dissolved salts. This brine can either be

disposed of without any additional treatment or minimized prior to disposal.

2. Feedwater intake and environmental impacts

The selection of the seawater intake system depends on the raw water source, local

conditions, and plant capacity. The best seawater quality can be reached by beach

wells, but in these cases the amount of water that can be extracted from each beach

well is limited by the earth formation, and therefore the amount of water available

by beach wells is very often far below the demand of the desalination plant. For

small and medium reverse osmosis plants, a beach well is often used.

For seawater with a depth of less than 3 m, short seawater pipes or an open intake

are used for large capacities. Long seawater pipes are used for seawater with depths

of more than 30 m.

5

The seawater intake may cause losses of aquatic organisms by impingement and

entrainment. The effects of the construction of the intake piping result from the

disturbance of the seabed which causes re-suspension of sediments, nutrients or

pollutants into the water column. The extent of damage during operation depends

on the location of the intake piping, the intake rate and the overall volume of intake

water.

The second impact category is linked to the demand of energy and materials

inducing air pollution and contributing to climate change. The extent of impact

through energy demand is evaluated by life cycle assessment, LCA. The impacts

of this category can be mitigated effectively by replacing fossil energy supply by

renewable energy and using waste heat from power generation for the thermal

processes.

The third impact category comprises effects caused by the release of brine to the

natural water body. Effluents from desalination are not merely concentrated salts,

but include a variety of chemicals that come from the reverse osmosis process, such

as antiscalants and antifoulants, including chlorine and other disinfection by-

products that may be toxic, as well as chemicals present in the intake water. These

additives and their by-products can be toxic to marine organisms, persistent

and/or can accumulate in sediments and organisms.

Also the elevated salinity of the concentrate can cause it to behave differently than

traditional wastewater, stormwater and cooling water plumes. When the effluent

density exceeds that of the ambient seawater, the plume could settle on the ocean

floor and spread as a density current, resulting in increased exposure to bottom-

dwelling marine life. The elevated concentration of salts and other constituents in

these discharges may result in adverse impacts to sensitive components of the

ecosystem.

Apart from the chemical and

physical properties the

impact of the brine depends

on the hydrographical

situation which influences

brine dilution and on the

biological features of the

discharge site. For instance,

shallow sites are less

appropriate for dilution than

open-sea sites and sites with

abundant marine life are

more sensitive than hardly

6

populated sites. But dilution can only be a medium-term mitigation measure. In the

long run the pre-treatment of the feed water must be performed in an

environmentally friendly manner. Therefore alternatives to conventional chemical

pre-treatment must be identified.

3. Characteristics of desalination brine

The environmental impacts of seawater desalination will be discussed

separately for each technology because of differences in nature and magnitude

of impacts. The technologies regarded here are MSF, MED and RO.

3.1 Multi-Stage Flash Desalination (MSF)

3.1.1 Seawater Intake

Due to their high demand of cooling water, MSF desalination plants are

characterized by a low product water conversion rate of 10 to 20 %. Therefore the

required volume of seawater input per unit of product water is large, i.e. in the case

of a conversion rate of 10 %, 10 m³ of seawater are required for 1 m³ of produced

freshwater (see Figure 6-1). Combining the high demand of seawater input in

relative terms with the high demand of seawater input in absolute terms due to the

large average MSF plant size the risks of impingement and entrainment at the

seawater intake site must be regarded as high. Therefore, the seawater intake must

be designed in a way that the environmental impact is low.

3.1.2 Discharge of Brine Containing Additives

The discharge of brine represents a strong impact to the environment due to its

changed physical properties, i.e. salinity, temperature and density, and to the

residues of chemical additives or corrosion products. In MSF plants common

chemical additives are biocides, anti-scalants, antifoaming agents, and corrosion

inhibitors. The conditioning of permeate to gain palatable, stable drinking water

7

requires the addition of chlorine for disinfection, calcium, e.g. in form of calcium

hydroxide, for remineralisation and pH adjustment. In case of acidification as pre-

treatment removal of boron might be necessary.

Figure 6-2 shows where the chemicals are added, and indicates the concentrations

as well as the characteristics of the brine and its chemical load.

8

3.1.3 Physical Properties of Brine

The physical parameters of the brine are different compared to the intake seawater.

During the distillation process the temperature rises and salt accumulates in the

brine. Taking the reference process (Figure 6-1) with a conversion rate of approx. 10

% (related to the seawater flow) as example the salinity of the brine rises from 45 g/l

to 67.5 g/l (Figure 6-3). Brine and cooling water temperature rises by 9 and 7.5 K,

respectively. Salinity of the brine is reduced by blending with cooling water, but still

reaches a value of 5.4 g/l above ambient level. The resulting increase of density is

small what can be attributed to balancing effects of temperature and salinity rise. In

general, the increase of the seawater salinity in the sea caused by solar evaporation

is normally much higher than by desalination processes. However, the brine

discharge system must be designed in a way that the brine is well distributed and

locally high temperature and salinity values are avoided.

3.1.4 Biocides

Surface water contains organic matter, which comprises living or dead particulate

material and dissolved molecules, leads to biological growth and causes formation of

biofilm within the plant. Therefore the seawater intake flow is disinfected with the

help of biocides. The most common biocide in MSF plants is chlorine. A

concentration of up to 2000 μg/l in the seawater intake flow is sustained by a

continuous dosage. Chlorine reacts to hypochlorite and, in the case of seawater,

especially to hypobromite. Residual chlorine is released to the environment with the

effluents from cooling and distillation where it reaches values of 200-500 μg/l,

representing 10-25 % of the dosing concentration. Assuming a product-effluent-ratio

of 1:9 the specific discharge load of residual chlorine per m³ of product water is 1.8-

4.5 g/m³. For a plant with a desalination capacity of 24,000 m³/day, for instance, this

means a release of 43.2-108 kg of residual chlorine per day.

Further degradation of available chlorine after the release to the water body will

lead to concentrations of 20-50 μg/l at the discharge site. Chlorine has effects on the

aquatic environment because of its high toxicity, which is expressed by the very low

value of long-term water quality criterion in seawater of 7.5 μg/l recommended by

the U.S. Environmental Protection Agency and the predicted no effect concentration

(PNEC) for saltwater species of 0.04 μg/l determined by the EU environmental risk

assessment. In Figure 6-4 the occurring concentrations near the outlet and at a

distance of 1 km are compared to ecotoxicity values determined through tests with

different aquatic species and to the EPA short term and long-term water quality

criteria. It is striking that most of the concentrations at which half of the tested

populations or the whole population is decimated at different exposure times or

show other effects are exceeded by the concentrations measured near the outlet

and even at the distance of 1 km.

9

Another aspect of chlorination is the formation of halogenated volatile liquid

hydrocarbons. An important species is bromoform, a trihalomethane volatile liquid

hydrocarbon. Concentrations of up to 10 μg/l of bromoform have been measured

near the outlet of the Kuwaiti MSF plant. The toxicity of bromoform has been proven

by an experiment with oysters which have been exposed to a bromoform

concentration of 25 μg/l and showed an increased respiration rate and a reduced

feeding rate and size of gonads. Larval oysters are even more sensitive to

bromoform, as significant mortality is caused by a concentration of 0.05-10 μg/l and

acute, 48 h exposures.

3.1.5 Antiscalants

A major problem of MSF plants is the scale formation on the heat exchanger surfaces

which impairs heat transfer. The most common scale is formed by precipitating

calcium carbonates due to increased temperatures and brine concentration. Other

scale forming species are magnesium hydroxide calcium sulphate, the latter being

very difficult to remove as it forms hard scales. Therefore sulphate scaling is avoided

in the first place by regulating the operation parameters temperature and

concentration in such a way that the saturation point of calcium sulphate is not

reached. Calcium carbonates and magnesium hydroxides, again, are chemically

controlled by adding acids and/or antiscalants.

In the past, acid treatment was commonly employed. With the help of acids the pH

(acidity value) of the feed water is lowered to 2 or 3 and hereby the bicarbonate and

carbonate ions chemically react to carbon dioxide which is released in a

10

decarbonator. Thus, the CaCO3 scale forming ions are removed from the feed water.

After acid treatment the pH of the seawater is re-adjusted. Commonly used acids are

sulphuric acid and hydrochloric acid, though the first is preferred because of

economic reasons. High concentrations and therefore large amounts of acids are

necessary for the stoichiometric reaction of the acid.. Apart from a high consumption

of acids further negative effects of using acids are the increased corrosion of the

construction materials and thus reduced lifetimes of the distillers as well as handling

and storage problems. The negative effects mentioned above have led to the

development of alternatives: Nowadays antiscalants are replacing acids during

operation. But before talking about antiscalants, the use of

acids as cleaning agents needs to be mentioned because that’s when significantly

acidic effluents occur. During this periodic cleaning procedure the pH is lowered to

2-3 by adding citric, sulfamic or sulphuric acid, for instance, to remove carbonate

and metal oxide scales. In this context Mabrook (1994, in /Lattemann and Hφpner,

2003/) explained an observed change in density and diversity of marine organisms

by a decreased pH of 5.8 compared to 8.3 in coastal waters. Eco-toxic pH values

range from 2-2.5 for starfish (LC50, HCl, 48 h) to 3-3.3 for salt water prawn (LC50,

H2SO4, 48 h) and show the sensitivity of marine organisms to low pH values. Little

mobile organisms, like starfish, are especially affected by an acid plume as they

cannot avoid this zone. To mitigate these possible effects the cleaning solution

should be neutralized before discharge or at least blended with the brine during

normal operation.

An antiscalant can suppress scale formation with very low dosages, typically below

10 ppm. Such low dosages are far from the stoichiometric concentration of the

scaling species. Hence inhibition phenomena do not entail chemical reactions

and stem from complex physical processes involving adsorption, nucleation and

crystal growth processes. Scale suppression in the presence of minute

concentrations of antiscalants is believed to involve several effects:

Threshold effect: An antiscalant can slow down the nucleation process

occurring in a supersaturated solution. Thereby, the induction period, which

precedes crystal growth, is increased. The inhibition effect of anti-scalants is

based on their ability to adsorb onto the surfaces of sub-microscopic crystal

nuclei, which prevents them from growing any further or, at least,

substantially slows down the growth process. Since anti-scalant molecules

with a low molecular weight are more mobile, the extension of the induction

period is more pronounced with molecules of comparatively low molecular

weight.

Crystal distortion effect: Adsorbed antiscalant molecules act to distort the

otherwise orderly crystal growth process. A different degree of adsorption

and retardation of the growth process on different crystal faces results in

11

alteration of the crystal structure. The scale structure can be considerably

distorted and weakened. The distorted crystals are less prone to adhere to

each other and to metal surfaces. When crystallisation has started either

further growth is inhibited or the precipitates form a soft sludge that can be

easily removed rather than hard scales.

Dispersive effect: Antiscalants with negatively charged groups can adsorb

onto the surfaces of crystals and particles in suspension and impart a like

charge, hence repelling neighbouring particles, thereby preventing

agglomeration and keeping the particles suspended in solution.

Sequestering effect: Antiscalants may act as chelating agents and suppress

the particle formation by binding free Ca2+ or Mg2+ ions in solution. Anti-

scalants with strong chelating characteristics cannot work at the sub-

stoichiometric level, as the anti-scalant is consumed by the scale-forming

ions. Sequestration is affected by chemicals that require relatively high

concentrations and is not a physical inhibition effect.

Polyphosphates represent the first generation of antiscalant agents with

sodium hexametaphosphate as most commonly used species. A procedural

disadvantage is the risk of calcium phosphate scale formation. Of major concern

to the aquatic environment is their hydrolytic decomposition at 60°C to

orthophosphate which acts as a nutrient and causes eutrophication. The

development of algae mats on the water body receiving the discharge could

be ascribed to the use of phosphates. Because of these reasons they have partly

been substituted by thermally stable phosphonates and polycarbonic acids, the

second generation of antiscalants. Where phosphates have been replaced by these

substances the problem of algae growth could be solved completely.

Main representatives of polycarbonic acids are polyacrylic and polymaleic acids.

Especially polyacrylic acid has to be dosed carefully if precipitation is to be avoided.

The reason for this is that, at lower concentrations, it enhances agglomeration and

therefore also serves as a coagulant in RO plants (see below). Discharge levels of

phosphonates and polycarbonic acids are classified as non-hazardous, as they are

far below concentrations with toxic or chronic effects. They resemble naturally

occurring humic substances when dispersed in the aquatic environment which

is expressed by their tendency to complexation and their half-life of about one

month, both properties similar to humic substances. Though they are generally

assumed to be of little environmental concern, there is a critical point related to

these properties. As they are rather persistent they will continue to complex metal

ions in the water body. Consequently, the influence on the dissolved metal

concentrations and therefore metal mobility naturally exerted by

humic substances is increased by polymer antiscalants. The long-term effect induced

hereby requires further research.

12

Experimental data on the bioaccumulation potential of polycarboxylates are not

available. However, polymers with a molecular weight > 700 are not readily taken up

into cells because of the steric hindrance at the cell membrane passage. Therefore a

bioaccumulation is unlikely. Copolymers have a favourable ecotoxicological profile.

Based upon the available short-term and long-term ecotoxocity data of all three

aquatic trophic levels (fish, daphnia, algae) for a variety of polycarboxylates, it is

considered that exposure does not indicate an environmental risk for the

compartments water, sediment and sewage treatment plants.

A MSF plant with a daily capacity of 24,000 m³ releases about 144 kg of antiscalants

per day if a dosage concentration of 2 mg per litre feedwater is assumed. This

represents a release of 6 g per cubic meter of product water.

3.1.6 Antifoaming Agents

Seawater contains dissolved organics that accumulate in the surface layer and are

responsible for foaming. The use of antifoaming agents is necessary in MSF plants,

because a surface film and foam-increase the risk of salt carry-over and

contamination of the distillate. A surface film derogates the thermal desalination

process by increasing the surface viscosity. An elevated

surface viscosity hampers deaeration. Furthermore, if the surface tension is too high,

brine droplets will burst into the vapour phase during flashing. Deaeration is

essential for thermal plants as it reduces corrosion; salt carry-over with brine

droplets must be avoided for a clean distillation.

As the antifoaming agents are organic substances, too, they must carefully be

chosen and dosed. Blends of polyglycol are utilized, either containing polyethylene

glycol or polypropylene glycol. These substances are generally considered as non-

hazardous and low discharge concentrations of 40-50 µg per litre of effluent further

reduce the risk of environmental damage. However, highly polymerized

polyethylene glycol with a high molecular mass is rather resistant to

biodegradation. On this account it has been replaced in some industrial

applications by substances, such as dialkyl ethers, which show a better

biodegradability. Addition of usually less than 0.1 ppm of an antifoaming agent is

usually effective. Concentrations in the discharge were found to be half this level,

which is mainly due to mixing of brine with cooling water.While the brine contains

residual antifoaming agents, the cooling water is not treated and thereby reduces

the overall discharge concentration.

Under the assumption of a product-feedwater-ratio of 1:3 and 0.035-0.15 ppm

dosing 0.1-0.45 g per cubic meter of product water are released.

13

3.1.7 Corrosion Inhibitors and Corrosion Products

An important issue for MSF plants is the inhibition of corrosion of the metals

the heat exchangers are made of. The corrosive seawater, high process

temperatures, residual chlorine concentrations and corrosive gases are the reason

for this problem. Corrosion is controlled by the use of corrosion resistant materials,

by deaeration of the feed water, and sometimes by addition of corrosion inhibitors .

Especially during acidic cleaning corrosion control by use of corrosion inhibitors is

essential for copper-based tubing. In a first step oxygen levels are reduced by

physical deaeration. The addition of chemicals like the oxygen scavenger sodium

bisulfite can further reduce the oxygen content. Sodium bisulfite should be

dosed carefully as oxygen depletion harms marine organisms.

Corrosion inhibitors generally interact with the surfaces of the tubes. Ferrous

sulphate, for example, adheres to the surface after having hydrolized and oxidized

and hereby protects the alloy. Benzotriazole and its derivates are special

corrosion inhibitors required during acid cleaning. They possess elements like

selenium, nitrogen, sulphur and oxygen with electron pairs which interact with

metallic surfaces building a stable protective film. However, it is assumed

that in the end the major amount is discharged with the brine. Due to the slow

degradation of benzotriazole, it is persistent and might accumulate in sediments if

the pH is low enough to allow adsorption to suspended material. Acutely toxic

effects are improbable because the expected brine concentrations are well below

the LC50 values of trout and Daphnia magna. Still the

substance is classified as harmful for marine organisms. The release of benzotriazole

per cubic metre product water, corresponding to a continuous dosage of 3-5 ppm to

the feed water, amounts to 9-15 g.

The most important representative of heavy metals dissolved from the tubing

material is copper, because copper-nickel heat exchangers are widely used. In brines

from MSF plants it represents a major contaminant. Assuming a copper level of 15

ppb in the brine and a product-brine-ratio of 1:2, the resulting output from the

reference MSF plant with a capacity of 24,000 m³/d is 720 g copper per day.

Generally, the hazard to the ecosystem emanates from the toxicity of copper at high

levels. Here, levels are low enough not to harm the

marine biota, but accumulation of copper in sediments represents a latent risk as it

can be remobilised when conditions change from aerobic to anaerobic due to a

decreasing oxygen concentrations. To illustrate the latent risk posed by discharge of

untreated brine Figure 6-5 compares reported discharge levels to eco-toxicity values

and the EPA water quality criteria. The eco-toxicity values have been derived from

values which have been determined during tests with

copper sulphate under the assumption that copper sulphate is of less concern for

14

saltwater organisms. Diluting discharge water with cooling water does not produce

relief as reported levels are still above water quality criteria and total loads stay the

same.

3.2 Multi-Effect Distillation Desalination (MED)


The flow rate of the cooling water which is discharged at the outlet of the final

condenser depends on the design of the MED distiller and the operating conditions.

In the case of a conversion rate of 11 % (related to the seawater intake flow), 9 m³ of

seawater are required for 1 m³ of fresh water (Figure 6-6). Due to the smaller unit

sizes the seawater intake capacity for a single MED unit would be lower than for a

single MSF unit, but in the majority of cases the required distillate production is

reached by installing several units in parallel. Thus, the seawater

intake capacity for MED plants and MSF plants would be similar. Nevertheless, the

potential damage caused by impingement and entrainment at the seawater intake

must be regarded as high.

15



changed physical properties and to the residues of chemical additives or corrosion

products. In MED plants common chemical additives are biocides, antiscalants,

antifoaming agents at some plants, and corrosion inhibitors at some plants. The

conditioning of permeate to gain palatable, stable drinking water requires the

addition of chlorine for disinfection, calcium, e.g. in form of calcium hydroxide, for

remineralization and pH adjustment. Figure 6-7 shows where the chemicals are

added and at which concentrations as well as the characteristics of the brine and its

chemical load.

16

3.2.3 Physical Properties of Brine

The physical parameters of the brine are different compared to the intake seawater.

During the distillation process the temperature rises and salt accumulates in the

brine. Taking the reference process (Figure 6-6) with a conversion rate of approx.

11.2 % as example the salinity rises from 45 g/l to 66 g/l (Figure 6-8). Brine and

cooling water temperature rises by about 14 and 10 K, respectively. Salinity of the

brine is reduced by blending with cooling water, but still reaches a value of 5.6 g/l

above ambient level. The resulting decrease of density is very small what can be

attributed to balancing effects of temperature and salinity rise.

3.2.4 Biocides



biofilm within the plant. Therefore both the feed water and the cooling water are

disinfected with the help of biocides. The most common biocide in MED plants is

chlorine. A concentration of up to 2000 µg/l is sustained by a continuous dosage.

Chloride reacts to hypochlorite and, in the case of seawater,

especially to hypobromite. Residual chloride is released to the environment with the

brine where it reaches values of 200-500 µg/l, representing 10-25 % of the dosing

concentration. Assuming a product-effluent-ratio of 1:8 the specific discharge load of

residual chlorine per m³ of product water is 1.6-4.0 g/m³. For a plant with a daily

desalination capacity of 24,000 m³, for instance, this means a release of 38.4-96.0 kg

of residual chlorine per day. The effects of chlorine are described in 3.1.2.

17

3.2.5 Antiscalants

A major problem of MED plants is the scale formation on the heat exchanger

surfaces which impairs the heat transfer. The most common scale is formed by

precipitating calcium carbonates due to increased temperatures and brine

concentration. Other scale forming species are magnesium hydroxide, and

calcium sulphate, the latter being very difficult to remove as it forms hard scales.

Therefore sulphate scaling is avoided in the first place by regulating the operation

parameters temperature and concentration in such a way that the saturation point

of calcium sulphate is not reached. Calcium carbonates and magnesium hydroxides,

again, are chemically controlled by adding acids and/or antiscalants.

In the past, acid treatment was commonly employed. With the help of acids the pH

(acidity value) of the feed water is lowered to 2 or 3 and hereby the bicarbonate and

carbonate ions chemically react to carbon dioxide which is released in a

decarbonator. Thus, the CaCO3 scale forming ions are removed from the feed water.

After acid treartment the pH of the feed water is re-adjusted. Commonly used acids

are sulphuric acid and hydrochloric acid, though the first is

preferred because of economic reasons. High concentrations and therefore large

amounts of acids are necessary for the stoichiometric reaction of the acid. Apart

from a high consumption of acids further negative effects of using acids are the

increased corrosion of the construction materials and thus reduced lifetimes of the

distillers as well as handling and storage problems. The negative effects mentioned

above have led to the development of alternatives: Nowadays antiscalants are

replacing acids during operation. But before talking about antiscalants, the use of

acids as cleaning agents needs to be mentioned because that’s when significantly

acidic effluents occur. During this periodic cleaning procedure the pH is lowered to 2-

3 by adding citric or sulfamic acid, for instance, to remove carbonate and metal oxide

scales. In this context Mabrook explained an observed change in density and

diversity of marine organisms by a decreased pH of 5.8 compared to 8.3 in coastal

waters. Ecotoxic pH values range from 2-2.5 for starfish (LC50, HCl, 48 h) to 3-3.3 for

salt water prawn (LC50, H2SO4, 48 h) and show the sensitivity of marine organisms to

low pH values. Little mobile organisms, like starfish, are especially affected by an acid

plume as they cannot avoid this zone. To mitigate these possible effects the cleaning

solution should be neutralized before discharge or at least blended with the brine

during normal operation.

The mode of action of antiscalants is described in 3.1.2. They react substoichio-

metrically which is the reason why they are effective at very low concentrations.

Polyphosphates represent the first generation of antiscalant agents with sodium

hexametaphosphate as most commonly used species. A procedural disadvantage is

the risk of calcium phosphate scale formation. Of major concern to the aquatic

18

environment is their hydrolytic decomposition at 60°C to orthophosphate which acts

as a nutrient and causes eutrophication. The development of algae mats on the

water body receiving the discharge could be ascribed to the use of phosphates.

Because of these reasons they have partly been substituted by thermally stable

phosphonates and polycarbonic acids, the second generation of antiscalants. Where

phosphates have been replaced by these substances the problem of algae growth

could be solved completely. Main representatives of polycarbonic acids are

polyacrylic and polymaleic acids. Especially polyacrylic acid has to be dosed carefully

if precipitation is to be avoided. The reason for this is that, at lower

concentrations, it enhances agglomeration and therefore also serves as a coagulant

in RO plants. Discharge levels of phosphonates and polycarbonic acids are classified

as non-hazardous, as they are far below concentrations with toxic or chronic effects.

They resemble naturally occurring humic substances when dispersed in the aquatic

environment which is expressed by their tendency to complexation and their half-life

of about one month, both properties similar to humic substances. Though they are

generally assumed to be of little environmental concern, there is a

critical point related to these properties. As they are rather persistent they will

continue to complex metal ions in the water body. Consequently, the influence on

the dissolved metal concentrations and therefore metal mobility naturally exerted

by humic substances is increased by polymer antiscalants. The long-term effect

induced hereby requires further research.

A MED plant with a daily capacity of 24,000 m³ releases about 144-288 kg of

antiscalants per day if a dosage concentration of 2-4 mg per litre feedwater is

assumed. This represents a release of 6 g per cubic meter of product water.

3.2.6 Antifoaming Agents

MED plants also use antifoaming agents, but compared to MSF plants, it’s less usual.

The use of antifoaming agents can be necessary if foam forms in the presence of

organic substances concentrated on the water surface which derogates the

thermal desalination process by hampering the falling film flow onto the horizontal

evaporator tubes and thus the wetting of the tubes.

As the agents are organic substances, too, they must carefully be chosen and dosed.

Blends of polyglycol are utilized, either containing polyethylene glycol or

polypropylene glycol. These substances are generally considered as non-hazardous

and low discharge concentrations of 40-50 µg/l per litre brine further reduce the

risk of environmental damage. However, highly polymerized polyethylene glycol

with a high molecular mass is rather resistant to biodegradation. On this

account it has been replaced in some industrial applications by substances, such

as dialkyl ethers, which show a better biodegradability.

19

Under the assumption of a product-feedwater-ratio of 1:3 and 0.035-0.15 ppm

dosing 0.1-0.45 g per cubic meter of product water are released.

3.2.7 Corrosion Inhibitors and Corrosion Products

The corrosion inhibitors that are used in MSF plants are also necessary in MED

plants. However, it is assumed that the copper load is smaller compared to MSF

plants as operation temperatures are lower and piping material with lower copper

contents are used, such as titanium and aluminium-brass.

3.3 Reverse Osmosis (RO)


The conversion rate of RO processes ranges between 20 and 50 % /Goebel 2007/,

signifying an intake volume of less than 5 m³ of seawater per cubic meter of

freshwater. Therefore, compared to the thermal processes the mechanical process

of RO requires significantly less intake water for the same amount of product water.

Consequently the loss of organisms through impingement and

entrainment is lower. The flows, shown in Figure 6-9, result from a conversion rate

of 33 %.



changed physical properties and to the residues of chemical additives or corrosion

products. In RO plants common chemical additives are biocides, eventually acids

if not yet substituted by antiscalants, coagulants, and, in the case of polyamide

membranes, chlorine deactivators. The conditioning of permeate to gain palatable,

stable drinking water requires the addition of chlorine for

disinfection, calcium, e.g. in form of calcium hydroxide, for remineralization and pH

adjustment. Figure 6-10 shows where the chemicals are added and at which

concentrations as well as the characteristics of the brine and its chemical load.

20

Physical Properties of Brine

The salinity of the brine is increased significantly due to high conversion rates of 30

to 45 %. The conversion rate of 32 % of the process presented in Figure 6-9 leads to a

brine salinity of 66.2 g/l (Figure 6-11). As the temperature stays the same during the

whole process, also density increases significantly from 1028 g/l to 1044 g/l. If the

RO process is coupled with electricity generation and the effluent streams are

blended, the warmed cooling water from the power plant reduces the overall

density slightly compared to the ambient value and the overall salinity is almost

reduced to the ambient level.

3.3.3 Biocides



biofilm within the plant.

Therefore the RO feed water is disinfected with the help of biocides. The most

common biocide in RO plants is chlorine. A concentration of up to 1000 µg/l is

21

sustained by a continuous dosage. Chloride reacts to hypochlorite and, in the case of

seawater, especially to hypobromite. In RO desalination plants operating with

polyamide membranes dechlorination is necessary to prevent

membrane oxidation. Therefore the issue of chlorine discharge is restricted to the

smaller portion of plants which use cellulose acetate membranes. Regarding these

plants residual chlorine is released to the environment with the effluents where

it reaches values of 100-250 µg/l, representing 10-25 % of the dosing

concentration. Assuming a product-effluent-ratio of 1:2 the specific discharge load of

residual chlorine per m³ of product water is 0.2-0.5 g/m³. For a plant with a daily

desalination capacity of 24,000 m³, for instance, this means a release of 4.8-12 kg of

residual chlorine per day. Again, the problem of chlorine discharge is restricted to

plants with cellulose acetate membranes. In contrast, the release of chlorination by-

products is an issue at all RO plants regardless of the material of their membranes,

as by-products form up to the point of dechlorination. The effects of chlorine are

described in 3.1.2.

3.3.4 Coagulants

The removal of suspended material, especially colloids, beforehand is essential for a

good membrane performance. For this purpose coagulants and polyelectrolytes

are added for coagulation-flocculation and the resulting flocs are hold back by dual

media sand-anthracite filters. Coagulant substances are ferric chloride, ferrous

sulphate, and ferric chloride sulphate or aluminium chloride. To sustain the

efficiency of the filters, they are backwashed regularly. Common practice is to

discharge the backwash brines to the sea. This may affect marine life as

the brines are colored by the coagulants and carry the flocs (see Figure 6-12). On the

one hand the decreased light penetration might impair photosynthesis. On the

other hand increased sedimentation could bury sessile organisms, especially corals.

The dosage is proportional to the natural water turbidity and can be high as 30 mg/l.

This extreme dosage results in a specific load of 90 g per m³ of product water and a

daily load of a 24,000 m³/d plant of 2200 kg which adds to the natural turbidity.

Polyelectrolytes support the flocculation process by connecting the colloids. Possible

substances are polyphosphates or polyacrylic acids and polyacrylamides respectively,

which are also used as antiscalants. The concentration decides whether they have a

dispersive or coagulative effect. Compared to their use as antiscalants the dosage of

polyelectrolytes is about a tenth of the concentration required for dispersion. These

substances are not toxic; the impact they cause is connected to the increased

turbidity. A dosage of 500 µg/l implies a discharge of 1.5 g per m³ of

product water and a daily load of a 24,000 m³/d plant of 36 kg which adds to the

natural turbidity.

22

3.3.5 Antiscalants

The main scale forming species in RO plants are calcium carbonate, calcium sulphate

and barium sulphate. Acid treatment and antiscalant dosage are used for scale

control. Here, sulphuric acid is most commonly used and dosed with a range of 30-

100 mg/l. During normal operation the alternative use of antiscalants, such as

polyphosphates, phosphonates or polycarbonic acids, has become very common in

RO plants due to the negative effects of inorganic acid treatment explained in 3.1.2.

As it is explained there, these antiscalants react substoichiometrically and therefore

low concentrations of about 2 mg/l are sufficient.

A RO plant with a daily capacity of 24,000 m³ releases about 144 kg of antiscalants

per day if dosage concentration of 2 mg per litre feedwater and product-feedwater-

ratio of 1:3 are assumed. This represents a releaseof 6 g per cubic meter of product

water.

3.3.6 Membrane Cleaning Agents

Apart from acid cleaning, which is carried out with citric acid or hydrochloric acid,

membranes are additionally treated with sodium hydroxide, detergents and

complex-forming species to remove biofilms and silt deposits.

By adding sodium hydroxide the pH is raised to about 12 where the removal of

biofilms and silt deposits is achieved. Alkaline cleaning solutions should be

neutralized before discharge, e.g. by blending with the brine.

Detergents, such as organo-sulfates and -sulfonates, also support the removal of dirt

particles with the help of both their lipophilic and hydrophilic residues. Regarding

their behaviour in the marine environment, organo-sulfates, e.g. sodium

dodecylsulfate (SDS), and organo-sulfates, e.g. sodium dodecylbenzene sulfonate

(Na-DBS), are quickly biodegraded. Apart from the general classification of

detergents as toxic no further information is available on toxicity of NaDBS, but it’s

assumed to be relatively low once the decomposition has started with cutting off the

hydrophilic group. In contrast, LC50 for fish, Daphnia magna and algae are available

in the case of SDS confirming the categorization as toxic substance. But, again, fast

degradation reduces the risk for marine life. This risk could be further reduced by

microbial waste treatment which destroys the surface active properties and

degrades the alkyl-chain.

Complex-forming species, such as EDTA (Ethylendiamine tetraacetic acid) are

employed for the removal of inorganic colloids and biofouling. From comparing the

calculated maximum estimate of discharge concentration (46 mg/l) and an LC50 for

bluegill (159 mg/l, 96 h) it can be deduced that in the case of EDTA direct toxicity is

of minor concern. In contrast, persistent residual EDTA in the marine environment

23

might provoke long-term effects in connection with its chelating and dispersing

properties. Consequences of increased metal solubility and mobility and thereby

reduced bioavailability still need further investigation. Generally, total amounts are

of bigger interest than concentrations.

During the periodic membrane cleaning process also further disinfectants such as

formaldehyde, glutaraldehyde, isothiazole, and sodium perborate, are used. These

substances are toxic to highly toxic and reach toxic concentrations if discharged all at

once. Therefore deactivation should be compulsory. Several deactivation substances

are available: formaldehyde can be deactivated with hydrogen peroxide and calcium

hydroxide or sodium hydroxide and isothiazole is neutralized with sodium bisulfite.

Sodium perborate has to be handled carefully as it breaks down to sodium borate

and hydrogen peroxide. The latter is the actual biocide and therefore may not

be overdosed, also for reasons of membrane protection as it has an oxidizing effect.

3.3.7 Corrosion Products

In RO plants corrosion is a minor problem because stainless steels and non-metal

equipment predominate. There are traces of iron, nickel, chromium and

molybdenum being released to the water body, but they do not reach critical levels.

Nevertheless, an environmentally sound process should not discharge heavy metals

at all; therefore alternatives to commonly used material need to be found.

3.3.8 Dechlorination

The removal of chlorine is performed with sodium bisulfite, which is continuously

added to reach a concentration three to four times higher than the chlorine

concentration, the former amounting to 1500-4000 µg/l. The corresponding amount

per cubic metre of product water is 4.5-12 g/m³. As this substance is a biocide itself

and harms marine life through depletion of oxygen, overdosing should be prevented.

Alternatively sodium metabisulfite is used.

24

4. Receiving coastal environment

4.1 Discharge Mode

The mode of discharge controls the physical properties of the discharge plume, the

most significant of which is the net buoyancy of the plume. There are three principal

discharge modes to be considered: 1) positively buoyant combined discharges that

blend concentrate with thermal or wastewater effluents using existing

infrastructure, 2) negatively buoyant discharge using dedicated brine discharge

infrastructure, and 3) sometimes or weakly-negatively buoyant combined discharge

when brine is the predominant effluent constituent. Each of these discharge modes

will interact differently with the receiving environment, producing different near

field different characteristics and dimensions in both the near and far fields.

Fig.5.1 Factors influencing discharge site scenarios.

25

4.2 Receiving Environment

The physical boundaries, oceanography, and geomorphology of the receiving

environment affect the fate of the plume and also determine the nature of the

biological communities potentially affected by the discharge. Key boundary

conditions that should be considered include coastal type, bathymetry and coastal

structures, sediment properties, and water mass properties (salinity and

temperature). The coastal type includes collision coasts, which are exposed open

coastlines that are accompanied by either sandy or rocky intertidal and subtidal

habitats. The geomorphology of both the sandy and rocky collision coastal types

creates high-energy coastal environments with vigorous ambient mixing and

advection that contributes to rapid dilution that limits dispersion and accelerates

extinction of brine discharge plumes.

Concentrate releases into the open ocean will be influenced by different currents as

a function of the depth (i.e., location) of the discharge. The three primary circulation

regimes that can be expected in the coastal setting are shown in Figure 5-1: 1) surf

zone, 2) inner shelf, and 3) deep waters. These regimes are distinguished by the

different processes that dominate their currents. The surf zone is that shallow water

province at the shoreline in which currents are dominated by the effects of breaking

waves. The inner shelf is that region from the surf zone to the offshore location

where incident waves begin to refract and shoal and where surface and bottom

boundary layers (also known as Ekman layers) first intersect. Finally, the deep water

is the offshore region for which surface and bottom boundary layers exist but the

bulk of the flow is dominated by the geostrophic acceleration balance between

horizontal pressure gradients and the effect of Earth's rotation (i.e., the Coriolis

acceleration). The exact depths that determine the boundaries between the three

offshore regimes vary from place to place and, at one location, in time. The

boundary between the surf zone and the inner shelf is around 10 m and between the

inner shelf and the deep water is around 30 m.

26

Geomorphology also influences the resident biological communities of a particular

coastal environment. Open coasts with sandy environments support soft-bottom

habitat species (particularly benthic macroinvertebrates), while rocky coasts provide

substrate for kelp-based marine communities. Rocky coasts support tide pool

environments, and kelp beds and sea grasses in the offshore environments, both of

which are often protected. Estuarine embayments are generally complex and highly

productive ecosystems, likely to have tidal marshes in the intertidal zone, which is

another sensitive habitat type protected. The subtidal areas of embayments are

generally nurseries for a variety of juvenile fish species and are considered to be

sensitive habitats.

Forcing Functions

Forcing functions affect the strength of ocean mixing, ventilation and available

dilution volume in shallow water, including: waves, currents, ocean water levels

(tides and sea level anomalies), and winds. Movement of material in inner shelf and

surf zone, including the average and low-frequency movements, is controlled by the

waves approaching the beach and the shape of the bottom. Waves striking the

beach at an oblique angle drive mean currents up or down the beach (cross-shore

currents) and parallel to the beach (long-shore currents) as a function of the

incidence angle. When waves approach the beach straight on, local rip currents are

more common, which are associated with complex circulation cells that control the

amount of mixing with waters outside the surf zone. Hence, any assessment or

monitoring of a shoreline (i.e., surf zone) discharge must include local wave statistics

and bathymetry data.

Of the three current regimes, the inner shelf has received the least amount of study.

By definition, it is a transition region between the surf zone and deep waters. In

recent years it has been recognized that the inner shelf is a critical region with

regard to cross-shore exchange of material and better characterization of inner shelf

dynamics has been shown. A good review of the important dynamics in the inner

shelf is given by Lentz and Fewings (2012).

The deep water currents outside the inner shelf are a combination of flow driven by

local winds and geostrophic currents. Accurate modeling of velocity statistics for a

given location can be difficult because, in all cases, flow is highly dependent on the

conditions at the boundary of any local model. (For a review of numerical ocean

circulation models see Miller 2007). Direct measurements in the region may be

adequate to describe the velocity variability if the data records are appropriately

long. Care must be taken, however, to focus model results and observations on the

local bottom currents when assessing the fate of concentrate in a negatively buoyant

plume. In particular, deep water bottom currents will include the effects of the

frictional bottom Ekman (boundary) layer.

27

4.3 Discharge and Site Variability

None of the initial conditions, forcing functions or the boundary conditions of the

far-field are constants over time, and consequently the dilution and dispersion of

concentrate discharge can have considerable variability, which complicates the

determination of "natural" conditions and prediction of discharge dispersion. There

may be short-term or seasonal changes in RO operations resulting in variations in the

concentrate discharge rates, salinities and temperatures. On the other hand, the

temporal variation in boundary conditions and forcing functions of the far field

receiving water can vary over a vast range of time scales that are related to

geophysical, atmospheric, and climatic processes, including: diurnal variations

related to tides, solar heating and coastal winds, monthly variations related to tidal

spring/neap cycles, semi-annual variability related to summer/winter equilibrium

transitions, and longer term variability related to climate (e.g., El Nino/Southern

Oscillation (ENSO)). Variations in salinity, climate, and bathymetry are discussed here

to illustrate some of the most important factors.

Salinity

Ocean salinity variation exerts a modulating effect on the impact of sea salts

discharged from a desalination plant. The RO process produces a concentrated sea

water reject (brine) that is a fixed multiple of the instantaneous source-water salinity

(generally 1.8 to 2 times ambient). However, the ambient ocean salinity has

considerably different degrees of variability. So a water quality objective for salinity

should not be a fixed limit in terms of absolute salinity units. Rather, a water quality

objective should be stated in terms of some relative measure of deviation from

natural background, such as % deviation from background or a minimum initial

dilution producing equivalent results.

Ocean climate

The plume dilution and dispersion processes in the far field are influenced by ocean

temperature, salinity and the wave climate. These features vary as a result of

seasonal weather cycles and can also be severely modified by global ocean climate

events.

4.4 Bathymetry and Gravity Currents

The dynamics of negatively buoyant plumes are fundamentally different that those

of positively buoyant plumes. The fate of positively buoyant plumes is primarily

controlled by background currents, density stratification, and wave or wind-induced

mixing. They will either reach the water surface or be trapped by ambient

stratification. Negatively buoyant plumes, on the other hand, will generate density

(i.e., gravity) currents along the seabed by virtue of their density anomaly compared

28

with the ambient bottom waters. The magnitude of these density currents will

depend on the magnitude of the density anomaly and the bottom slope.

There may be environmental concerns with respect to density currents, however.

The presence of rocky outcrops and reefs offshore from the discharge site may block

the offshore dispersion of brine by gravity. Therefore, discharge sites with

bathymetric barriers (offshore rocky reefs and outcrops) should be avoided with

negatively buoyant discharges. Depending on the mixing rates with ambient waters

outside of the density layer, the dissolved oxygen (DO) supply to the density layer

may not meet the net oxygen demand of the benthic fauna within the layer. In this

case, DO will decrease over time and, if the layer persists long enough, hypoxia or

anoxia within the bottom layer can produce lethal effects in the far field well away

from the discharge. This is unlikely to occur with a well-designed discharge, however.

Many factors control the development of hypoxia or anoxia, including the

stratification between the ambient waters and the density layer, the thickness of the

layer, the water depth, the slope of the bottom, the strength of the wind, the

vertical velocity shear across the layer, and the height of the surface waves. The

general situation and many of these factors are addressed in the excellent study by

Hodges et al. (2011).

Other far field bathymetric features to be avoided for the siting of a negatively

buoyant brine discharge are bathymetric depressions (hollows). These are not

generally features found along the exposed open coast of California, but can be

common in embayments, either from natural shoaling effects or from man-induced

activities such as the dredging of navigation channels and berthing areas. When such

features are located in embayments with low mixing, a bathymetric depression can

fill with brine and displace the lighter ambient seawater from the depression. This

situation can result in stratification and stagnation of the bottom layer, leading to

hypoxia and increased exposure of the benthos to the plume contaminants. Sites

with topographic depressions should be avoided as locations for negatively buoyant

discharges.

29

5. Brine disposal; A look at the possible options

5.1 Inland plants

The major strategies for brine disposal at inland sites are limited to three general

categories;

1) Deep Well Injection

2) Evaporation Ponds and

3) Solar Ponds.

Several other systems for utilization of waste brine have been proposed, which

include, among others, irrigation of salt tolerant plants (halophytic crops) and brine

shrimp harvesting. Such approaches have been limited and are certainly not

applicable to very large volumes of wastewater. Recovery of inorganic salts with

potential commercial value has also been suggested, but construction of chemical

separation facilities would indeed result in a costly venture. It is important to

recognize the prime mission of any desalination facility, which is to upgrade water

quality - not to market by-products. To date, proposed by-product recovery systems

have not demonstrated economical viability and it seems likely that the cost of by-

product recovery would far exceed the cost of the principal product – water.

A system known as zero discharge has also been used in certain situations where

waste brine streams are relatively small and available land is limited. This technology

has generally been applied to wastewater disposal from power plants, oil refineries

and certain mining operations. The final stage of such a system involves thermal

evaporation, subsequently providing a solid residue. Energy requirements are large

and overall economics unfavorable for handling large brine volumes.

5.1.1 Deepwell injection

Deep well injection is presently applied worldwide for disposal of industrial,

municipal and liquid hazardous wastes . In recent years this technology has been

given serious consideration as an option for brine disposal from land based

desalination plants. Deep well injection has been applied successfully for brine

disposal from several membrane plants in Florida. Injection wells may vary in depth

from a few hundred feet to several thousand feet depending on geological

considerations at the selected site. Several factors contribute to the overall

performance and reliability of an injection well. In general, however, this method of

brine disposal is considered the most cost effective as compared with other systems

in practice for land based desalination plants.

30

The nature of subterranean strata

must be carefully considered in

selecting a suitable well location.

Most stratifications consist of

sand (porous medium) and shale

(confining layers). Mathematical

models, are used to determine

the permeability and solution

confinement capability of strata

with different ratios of sand to

shale. They are primarily

applicable to hazardous waste

disposal, can also be used as a

model for deep well injection of

any concentrate where

inadequate waste solution

transport and confinement

could result in contamination of

surface and groundwater

resources.

A measure of the effects of

plugging and damage to

subterranean formations on

injection well performance was

expressed by injectivity (I),

defined as the ratio of injection rate (q) to the difference between well flowing

pressure (Pwf) and the average formation pressure (Pr) given by the following

equation.

I = q / (Pwf - Pr) (1)

Injectivity is impacted by several factors, which include the chemical and physical

quality of the injected fluid, injection rate and pressure, as well as the nature and

physical properties of subterranean strata. One of the most important constraints on

stable injectivity is the presence of suspended solids in the injection fluid. Frequent

measurements of total suspended solids (TSS) are required to insure steady well

performance. High TSS in process fluids, low injection rate, low injection pressure,

and low porosity and permeability of the well strata all contribute to rapid well

plugging and diminished injectivity.

31

Deep well injection is a reasonable method for brine disposal provided that long-

term operation can be maintained, in order to dispose of large volumes of process

fluid. Drawbacks of this technology are:

1) selection of a suitable well site;

2) costs involved in conditioning the waste brine

3) possibility of corrosion and subsequent leakage in the well casing;

4) seismic activity which could cause damage to the well and subsequently result in

ground water contamination; and

5) uncertainty of the well half-life which can only be estimated using mathematical

simulation techniques.

5.1.2 Evaporation ponds

Evaporation pond technology is practiced primarily in the Middle East and to a lesser

extent in arid regions of Australia. At this time, it is probably the most widespread

method of brine disposal from inland-based desalination facilities. This disposal

system is especially effective in regions with low rainfall, and where climatic

conditions are favorable for steady, and relatively rapid evaporation rates. In

addition, desalination plants are often sited at locations where the cost of adjacent

level land is relatively low.

The pond open surface area (A) and minimum pond depth (d) can be estimated from

A = (V x f1) / Eave (2)

d = Eave x f2 (3)

where V is volume of reject water, Eave is evaporation rate, f1 is an empirical safety

factor to allow for lower than average evaporation rate and f2 is an empirical factor

32

that accounts for the length of the winter season. The designer of evaporation ponds

must carefully consider the surface area, depth and freeboard of such installations,

since these factors are determined by rates of concentrate discharge relative to

surface evaporation rates. It is noted that freeboard is especially difficult to estimate

since it depends on average rainfall and wind velocity in the pond area. It is clear

from the above relationship that the area needed is directly proportional to volume

of reject water and inversely proportional to the evaporation rate. Although other

empirical factors and relations have been suggested for calculating the impact of

brine salinity on surface evaporation rates, in the judgment of the authors of this

report, evaporation pond design optimization could be best developed

experimentally by circulating typical model brine solutions through small

evaporation ponds.

The principal environmental concern associated with evaporation pond disposal is

pond leakage, which may result in subsequent aquifer contamination. All current

installations are lined with polyethylene or various other polymeric sheets. Liner

installation must be carried out with care since sealing of joints is critical in order to

prevent leakage. Double lining is strongly recommended with leakage sensing probes

installed between layers of pond lining.

A system for enhancement of evaporation pond performance has recently been

developed at Ben Gurion University in Israel. This new technology described as the

WAIVER process involves periodic circulation of pond brine over “wetable surfaces,”

designed to increase the effective evaporative surface area. This results in enhanced

evaporation, which, of course, also depends on wind speed and direction in addition

to relative humidity. It has been estimated in this study that evaporation rates can

be increased by 50% in a typical Middle East dry climate. If proven effective in

practice, evaporation pond size could be significantly reduced.

Existing literature indicates that application of evaporation ponds is a relatively

simple and straightforward method of brine disposal. This technology, however, is

limited to relatively small desalination plants (less than 5MGD) and generally

33

restricted to arid climatic conditions. Capital costs arise primarily from acquisition of

land.

5.1.3 Solar ponds

Development of salt

gradient solar ponds as a

renewable energy source

began in Israel more than

thirty years ago. Although

limited in scope,

successful power

generation by this

technology has been

demonstrated primarily in

arid and semi-arid parts of

the world. Recent

technical papers have also appeared, describing experimental studies in Italy and

Switzerland, in which solar ponds are coupled with thermal desalination systems. In

these experimental studies, the pond is used as a heat source for small multistage

flash evaporator units. Ongoing solar pond studies at the University of Texas involve

power generation and thermal desalination coupled with brine disposal for recharge

of the bottom (hot) layer of the pond.

Perhaps one of the most innovative system for utilization of solar ponds is an

experimental study, conducted by the California Department of Water Resources at

Los Banos, CA. in the early 1980’s, which developed an integrated system for

membrane desalination coupled with brine disposal. A salt gradient solar pond

provided hot water cycled through a dual media heat exchanger, driving a turbine

for electric power generation. The power was then utilized for electrically driven

pumps to provide necessary pressure for reverse osmosis desalination of agricultural

drainage water. The resulting RO concentrate was then injected into the hot bottom

layer of the solar pond. The overall objective of this system was to provide for brine

disposal, heat recovery, and salinity augmentation of the solar pond. Unfortunately,

the Los Banos desalting operation was shut down in 1986 as a result of an EPA order

for termination of agricultural drainage, which provided feed water for this

experimental facility.

34

5.1.4 Zero Liquid Discharge (ZLD)/ Degremont

The ZLD System removes dissolved solids from the wastewater and returns distilled

water to the process (source). Reverse osmosis (membrane filtration) may be used

to concentrate a portion of the waste stream and return the clean permeate to the

process. In this case, a much smaller volume (the reject) will require evaporation,

thus enhancing performance and reducing power consumption. In many cases,

falling film evaporation is used to further concentrate the brine prior to

crystallization.

Falling film evaporation is an energy

efficient method of evaporation,

typically to concentrate the water up to

the initial crystallization point. The

resultant brine then enters a forced-

circulation crystallizer where the water

concentrates beyond the solubility of the

contaminants and crystals are formed.

The crystal-laden brine is dewatered in a

filter press or centrifuge and the filtrate

or centrate (also called “mother liquor”)

is returned to the crystallizer. The

collected condensate from the

membranes, falling film evaporator and

forced-circulation crystallizer is returned to the process eliminating the discharge of

liquids. If any organics are present, condensate polishing may be required for final

cleanup prior to reuse.

35

Pretreatment: In the majority of cases, it is more costeffective to remove

contaminants prior to evaporation/ crystallization. This pretreatment step often

includes limesoda softening which requires clarifiers to remove calcium and

magnesium compounds.

Membrane Filtration: Where possible, membrane filtration such as reverse osmosis

can be used to treat the wastewater. The permeate (clean water) is reused in the

process and the reject/concentrate is sent on to evaporation.

Evaporation: When a significant amount of water needs to be evaporated prior to

the crystallization step, pre-concentration in a falling film evaporator is the most

efficient solution. These evaporators require less heat/power per unit of water

evaporated.

Crystallization: The crystallizer is the heart of the ZLD process. Typically, forced-

circulation crystallizers are used, which evaporate the water past the crystallization

point. Crystals are mechanically dewatered and the resulting filtrate/ centrate is

returned to the crystallizer. The crystallizer usually requires corrosion resistant

materials due to the extremely high salt concentrations present. In some cases, part

of the crystallization can be achieved by spray driers to overcome high solubility of

certain salts. The clean condensate is returned to the process for reuse and the

dewatered crystals are transported off-site for reuse or disposal. This crystallization

process is extremely sensitive to the wastewater chemistry as the ions present will

determine the boiling point elevation which has a major impact on the power

consumption, impacting both capital and operating costs.

Solids Recovery: Sludge generated by the pretreatment phase, is generally

mechanically dewatered in a plate-and-frame filter press. A solids concentration of

20-50% dry solids can usually be achieved and the filtrate is simply recycled back to

the beginning of the pretreatment system. The crystals from crystallization phase

can also be mechanically dewatered, but corrosion resistant materials are usually

necessary due to the high salt concentrations present. The crystals can be dewatered

in a filter press (belt or recessed chamber) or centrifuge and as a result, much higher

solids concentrations can be achieved. The filtrate (or centrate) is then returned to

the crystallizer.

System Enhancements

Multiple Effects: Evaporation processes such as falling film evaporators can be

installed in series such that the water vapour from one is reused in the next. In this

way, the efficiency of the evaporation process can be increased and almost doubled,

tripled, etc. based on the number of evaporator effects installed. This increases the

36

capital cost of the system, but it is more economical for larger flow operations based

on the energy saved.

Waste Heat Usage: Economics are enhanced when waste heat found in most

industrial applications can be productively reused in the ZLD design. This can take

the form of dryer exhaust gas or low pressure return steam. The evaporator designs

can make use of these ‘waste’ heat sources, significantly saving on energy

requirements.

Integration Effects: The chemistry of ZLD systems is very complex. The waste

streams from the different zones are very interdependent. Having a single point of

contact for the entire system is therefore crucial.

Mechanical Vapor Recompression: For some applications, turbofans or high speed

compressors can be more economical by reducing the steam usage with electricity.

Water vapor produced in the evaporator or crystallizer can be compressed and

reused as the heating source. Depending on the boiling point elevation of the

wastewater, single- or multi-stage vapor compression systems can be used. MVR

systems are typically implemented where high evaporation rates are necessary.

A ZLD system is potentially more cost-effective when the nearest discharge location

(surface water or sewer) is far from the conventional treatment plant, since the ZLD

system eliminates the need for long conveyance lines.

The disadvantage of such a system is the potentially higher capital costs for

treatment equipment and more complex operations. A brine ZLD system uses heat

to force the majority of the water to evaporate quickly, leaving behind thick slurry

that can be sent to a small evaporation pond for final solar drying or mechanical

dewatering. Figure 1 shows examples of the brine condition before and after the

processes.

37

There are three general types of equipment for a brine ZLD system:

• Brine concentrator – Also known as a brine evaporator. The concentrator reduces

the original brine volume by 95 to 98%, and effluent is a two-phase solution – a high

TDS liquid and crystalline salt particles.

• Brine crystallizer – This further heats the concentrated solution from a

concentrator to reduce the liquid to slurry that consists of highly saline free liquids

and solid crystalline salt particles.

• Spray dryer – Alternative to the crystallizer that uses heat and forced air. ZLD

systems for brine applications are complex and energy intensive mechanical systems

with lifecycle costs on the same order as the desalination process itself. Such

systems have been used for decades in the pulp, petrochemical and power

industries. Given the costs, ZLD systems have rarely been used in municipal

applications in the United States.

However, ZLD systems will be used more frequently as salinity control and water

recovery become more pressing issues. Given the cost of this equipment, alternative

processes and brine minimization treatment equipment have been developed as

ways to reduce the costs of evaporating the brine. A major research and

development effort has been initiated by the water industry to develop new

methods to reduce the brine stream so that the ZLD systems can be smaller and less

expensive to install and operate.

5.2 Coastal options; Disposal strategies and near field effects

For a direct discharge of brine, the use of a diffuser is preferred. For flows typical of

major desalination plants, a multiport diffuser will probably be required that results

in high dilutions and rapid reductions of salinity in the near field. The diffuser should

be designed so that the jets do not impact the water surface and the effects of jet

merging should be carefully modeled (see later discussion of modeling techniques).

For co-discharges with power plant cooling water, existing shoreline surface

discharges, multiport diffusers, or single-port risers can probably be used. In most

cases, however, near field dilution alone may not suffice to meet water quality

standards and in-pipe dilution will also be needed. If the discharge is negatively

buoyant, the dilution from horizontal nozzles must be carefully evaluated to ensure

adequate initial dilution. Small amounts of concentrate can probably be discharged

through existing municipal wastewater outfall diffusers. However, the dilution must

be re-evaluated to account for the change in effluent density and flow rates, and

carefully evaluated if negatively buoyant.

38

5.2.1 Coastal hydrodynamic concept

It is important to understand the distinctions between near field, mixing zones, and

other related terms that are often associated with wastewater discharges. The near

field is a hydrodynamic, or physical, concept. It is the region where mixing of the

effluent is influenced and affected by discharge parameters. The physical processes

are primarily entrainment caused by shear between the buoyant jet (either

positively or negatively buoyant), an internal hydraulic jump where the plume

impacts a boundary (e.g., sea floor) or water surface and transitions to horizontal

flow, and entrainment in the horizontally spreading layer. The near field ends where

the self-induced turbulence collapses under the influence of the induced density

stratification. The layer then spreads as a density current of some finite thickness.

Ultimately, ambient diffusion due to oceanic turbulence is responsible for most

mixing and dilution; this region is known as the far field. The rate of mixing and

dilution in the far field is much slower than in the near field. A mixing zone is a

regulatory concept that will generally encompass most, or all, of the near field.

The near field characteristics of negatively buoyant discharges are primarily

determined by the orientation of the discharge port or nozzle to the horizontal, the

jet exit velocity, and the density difference between the effluent and receiving

water. Flowing currents will generally increase the dilution in the near field. For

larger discharges a multiport diffuser consisting of many nozzles will be needed. In

that case, an additional parameter is the port spacing and orientation of the diffuser

axis to the prevailing currents.

5.2.2 Disposal Alternatives

Examples of common concentrate discharge scenarios are shown in Figure 6-1.

Concentrates can be disposed of in several ways. They can be discharged as a surface

stream at the shoreline, co-mixed (and pre-diluted) with other effluent such as

municipal wastewater or power plant cooling water, or directly into the ocean as a

“pure” brine stream. For shoreline surface discharges (Figures 6-1a and b), the near

field results primarily from entrainment into the surface layer (for a positively

buoyant flow), or the bottom density current (for a negatively buoyant flow). This

entrainment is dependent on the source velocity, as entrainment due to the

spreading density currents is quite slow. Also, the density stratification reduces

vertical mixing in the far field. Because of these effects, near field dilution is quite

small, of order 5 times or less.

39

40

Shoreline discharge of raw (negatively buoyant) concentrate (Figure 6-1a) will result

in a density current that runs down the bottom slope. Because the resulting density

stratification inhibits vertical mixing, dilution is relatively small and benthic

organisms could be exposed to relatively high salinities. Shoreline disposal of pure

concentrate by this means in California is discouraged.

Co-discharge is another disposal strategy that involves diluting the concentrate to

below potentially toxic levels prior to discharge into the receiving water body. This

strategy involves blending brine with an existing effluent stream to achieve what is

referred to as in-the-pipe dilution or in-plant dilution. Co-discharge is permitted by

California water quality regulations and is currently used by several facilities.

Shoreline discharges are practical if co-discharged with a much larger flow for pre-

dilution, such as power plant cooling water. In this case, the effluent is likely to be

positively buoyant because of the elevated temperature of the cooling water (Figure

6-1b).

There are two common means for achieving in-plant dilution: 1) co-locating the

desalination plant with a wastewater plant, in which the dilution water is generally

of very low salinity; or 2) colocating the desalination plant with a power plant where

the dilution water is cooling water taken from the receiving water body, typically the

ocean. Dilution with wastewater produces a discharge salinity lower than ambient

seawater, even at relatively low wastewater discharge rates because the treated

effluent is fresh water. This is a means of reducing or eliminating hypersalinity

impacts on marine life from brine discharge.

Concentrates that are blended with other effluents are typically discharged though

existing ocean outfalls and diffusers (Figure 6-1c). Discharge through an existing

outfall and diffuser will generally be at “low” pressure, i.e. the jet exit velocity is

relatively low and the jet momentum flux will be quite small. If the effluent is

positively buoyant as a result of the elevated effluent temperature, the jets will

ascend towards the surface. If the ambient stratification is strong enough the plumes

will be trapped below the water surface, if not the plumes will reach the water

surface. The near field is primarily the rising plume region and has dispersion

characteristics similar to other buoyant plumes currently addressed in the Ocean

Plan. Because multiport diffusers for positively buoyant effluents are predominantly

horizontal, they may not be suitable for a negatively buoyant discharge and will have

to be carefully evaluated. A possible solution is to open more ports on the diffuser

and fit the portswith variable-area check valves which give higher velocity at low

flow rates. Again, the dilution must be carefully modeled and evaluated.

For co-discharge though a single large vertical riser (such as used for some power

plants) the exit dimensions may be very large, such as a square opening 25 ft on side,

which is comparable to the local water depth. In that case, the initial dilution can be

41

quite small and mixing in the spreading layer should be incorporated into the near

field. These types of discharges should include in-pipe dilution of the brine with

larger flows of seawater in order to achieve adequate dilution of the brine within the

mixing zone.

The use of seawater to achieve in-plant dilution requires a much larger volume,

relative to municipal wastewater effluent, to achieve a comparable level of reduction

in the salinity of the brine discharge. The intake of seawater used for in-plant

dilution (e.g., as power plant cooling water) causes additional mortality to marine

organisms through velocity shear and turbulence in the confined flows through

pumps and impellors of the (older design) once-through sea water circulation

systems. However, recent work on hydroelectric turbines by Cada (2001) and Cada et

al. (2006) has shown pump-induced turbulence mortality can be reduced by

employing low speed impellors after the Kaplan turbine and Archimedes screw

pump that reduce the shear stresses on entrained organisms to levels they can

tolerate. Low-stress water wheel technologies are also being considered as

alternatives to seawater circulation pumps of legacy power plants to reduce impacts

on marine life. The practicality of these technologies for the applications considered

here remains to be demonstrated, however.

The final case is direct discharge of negatively buoyant brine concentrate by means

of high velocity jets inclined upwards. This could be either a single jet for a small

discharge or a multiport diffuser for larger discharges. Multiport diffusers are used

for the Perth and Sydney (Australia) desalination plants. The high jet velocities result

in entrainment of ambient seawater into the jets and rapid dilution and reduction of

salinity. The processes are illustrated in Figure 6-2. Dilutions exceeding 30:1 can be

readily accomplished by such a diffuser.

A multiport diffuser with multiport “rosette” risers is shown in Figure 6-3. In this

example, the rosettes each consist of four nozzles. Other diffusers may have the

nozzles distributed uniformly along one or both sides of the diffuser.

42

In turbulent environments, physical damage can occur to delicate eggs and larvae.

The effect of turbulence on larval mortality was studied in the field by Jessopp

(2007), who found that even turbulent tidal flows produce significantly increased

mortality to thin-shelled veligers of gastropods and bivalves. While there is presently

no known published evidence of mortality to marine species for diffuser jets, the

cause and effect relations demonstrated by prior studies certainly raises that

possibility. Threshold shear stress tolerances of marine organisms to diffuser

discharges could be established by combining data from laboratory tests,

computational fluid dynamics modeling, and field studies of diffuser systems.

6. Summary & Conclusions

Available technologies for brine disposal

43

Future technologies,

In summary:

• There are no ‘residuals-free’ treatment processes. While some residuals streams

are more difficult and costly to dispose of than others, the disposal of residuals

should be considered from the beginning of the project.

• The contaminants requiring treatment may not be those most difficult to dispose.

Residual disposal should consider the full spectrum of contaminants in the source

water, not just those being treated.

• If surface water or sewer discharge is not feasible, the cost and complexity of

residual disposal can equal that of treatment for high pressure membrane systems.

This economic reality must be addressed in the funding for the project.

44

• Because of the high cost and complexity of brine disposal, the selection of the

disposal method may dictate the selection of the treatment process.

• Brine management is a dynamic field that’s rapidly changing as vendors try to fill

the void.

Documents

An Introduction to Brine Waste