City of Tshwane: Draft Water Scarcity Report · 3 DEFINING WATER SCARCITY The Food and Agriculture Organisation of the United Nations defines water scarcity when demand for freshwater

1

City of Tshwane: Draft Water Scarcity Report

2

Contents 1 PURPOSE ............................................................................................................................................... 5

2 STRATEGIC OBJECTIVES TO BE ADDRESSED .......................................................................................... 5

3 DEFINING WATER SCARCITY ................................................................................................................. 5

4 BACKGROUND ....................................................................................................................................... 7

4.1 THE CITY OF TSHWANE WATER RESOURCES MASTER PLAN (WRMP) .......................................... 8

4.2 PRESENT BULK WATER DISTRIBUTION SYSTEM ............................................................................ 8

4.3 CURRENT WATER SOURCES AND DEMAND .................................................................................. 9

4.4 FUTURE WATER DEMAND AND SEWER FLOWS ............................................................................ 9

4.5 STATUS QUO AND FUTURE WATER TREATMENT PLANT (WTP) EXPANSION. ............................ 10

4.5.1 Status Quo of Rietvlei WTP ................................................................................................. 10

4.5.2 Status Quo Roodeplaat WTP ............................................................................................... 13

4.5.3 Status Quo Themba WTP .................................................................................................... 15

4.5.4 Status Quo Bronkhorstspruit WTP ...................................................................................... 17

4.5.5 Status Quo Bronkhorstbaai WTP ........................................................................................ 20

4.5.6 Summary ............................................................................................................................. 22

5 CITY OF TSHWANE WATER REQUIREMENTS AS FROM WRMP ........................................................... 22

5.1 Crocodile West River Catchment ................................................................................................ 25

5.2 Upper Olifants River Catchment ................................................................................................. 26

5.3 WATER QUALITY AND PROCESSES REQUIRED FROM THE WRMP .............................................. 26

5.3.1 Water Reuse ........................................................................................................................ 27

5.3.2 REUSE PROCESSES ............................................................................................................... 28

6 INPUT MODELLING METHODOLOGY .................................................................................................. 31

6.1 GDP Modelling ............................................................................................................................ 31

6.2 Population Modelling .................................................................................................................. 33

6.2.1 Weighted Average ............................................................................................................... 33

6.2.2 The starting point ................................................................................................................ 34

6.2.3 Projection Methods ............................................................................................................ 34

6.2.4 Ratio Methods ..................................................................................................................... 35

6.2.5 Extended Term population Models .................................................................................... 40

6.3 Population Characteristics Modelling ......................................................................................... 41

6.3.1 Household Calculations ....................................................................................................... 41

6.3.2 Age Categories .................................................................................................................... 42

3

7 WATER DEMAND MODELLING ............................................................................................................ 43

7.1 Projections with normal GDP ...................................................................................................... 46

7.2 Projections with High GDP .......................................................................................................... 48

7.3 Infrastructure .............................................................................................................................. 50

7.4 Water Losses Costs. .................................................................................................................... 52

8 RISKS THAT MAY INFLUENCE THE CITY OF TSHWANE WATER RESOURCES MASTER PLAN (WRMP) 53

8.1 Population increase .................................................................................................................... 53

8.2 Climate change ............................................................................................................................ 54

8.3 ACID MINE DRAINAGE (AMD) ..................................................................................................... 56

8.3.1 AMD Generation ................................................................................................................. 58

8.4 Ground and surface water pollution ........................................................................................... 60

8.4.1 SOURCES OF GROUND WATER POLLUTION ........................................................................ 60

8.5 SPRINGS AND BOREHOLES .......................................................................................................... 61

8.5.1 Capture Zones ..................................................................................................................... 62

8.5.2 Aquifer Protection based on Capture Zones ....................................................................... 63

9 Age and Maintenance of existing infrastructure ................................................................................ 63

9.1 Leak repair reaction time ............................................................................................................ 65

9.1.1 Water pipe system results .................................................................................................. 65

10 Nonpayment of services ................................................................................................................. 66

11 Reporting of water leakages. .......................................................................................................... 68

12 Conclusion ....................................................................................................................................... 69

13 Recommendations .......................................................................................................................... 70

Table of Figures

Figure 5.1: Simplified schematic of the CoT water resource system .......................................................... 24

Figure 5.2: Treatment technologies are available to achieve any desired level of water quality (Taken

from EPA/600/R-12/618) ............................................................................................................................ 28

Figure 5.3: Possible process configurations ................................................................................................ 30

Figure 6.1: Mapping of CoT scenarios against WRP requirements ............................................................. 47

Figure 6.2: Water Demand balance calculations ........................................................................................ 48

Figure 6.3: Water requirements mapped High Growth Scenarios ............................................................. 49

Figure 6.4: High Growth Scenario Water Balance results ........................................................................... 50

Figure 6.5: Water system and supply mapping 2012 - 2055 ...................................................................... 50

Figure 6.6: Water system mapping High Growth scenario 2012 - 2055 ..................................................... 51

4

Figure 7.1: AMD formed owing to interaction between water and mine residue on the surface tailing

reclamation operation in the Western Basin. This site drains into pit operations that are directly

connected to mine void .............................................................................................................................. 57

Figure 7.2: acid drainage flowing into streams near Krugersdorp Source: Rachel, 2011 ........................... 57

Figure 8.1: Systems age analysis ................................................................................................................. 66

Figure 9.1: Non Revenue Water for the period 2007 - 2012 ...................................................................... 67

Figure 10.1: City wide leakages repair times .............................................................................................. 68

Figure 10.2: Average leaks per month for the period 2012 – 2015 ............................................................ 68

Figure 10.3: Technical staff per 100 000 people ......................................................................................... 69

Table of Tables

Table 6.1: Forecasting and Rationale .......................................................................................................... 44

Table 6.2: Household figures for the period 2012 - 2055 ........................................................................... 45

Table 6.3: Kilolitre per day utilized for each respective year...................................................................... 45

Table 6.4: Key input Variables ..................................................................................................................... 52

Table 6.5: Water losses results ................................................................................................................... 52

Table 7.1: City of Tshwane population figures, across three scenarios ...................................................... 53

Table 7.2: CoT Springs and Borehole Water Sources .................................................................................. 62

Table 8.1: Summary of replacement cost by regions ................................................................................. 66

5

1 PURPOSE

The City of Tshwane Water Resources Master Plan was approved by Mayco on 17 June

2015. This document is aimed to inform on the risks that potentially could negate the

objectives of the Water Resource Master Plan and result in water scarcity within the City

of Tshwane. Furthermore, the work aims to determine the water needs of the city and

compare it to the existing information to further assist decision making in the city.

2 STRATEGIC OBJECTIVES TO BE ADDRESSED

Projected water scarcity is one of the major problems faced by cities throughout the globe.

The city of Tshwane like other numerous cities is bound to plunge into water scarcity

problems that are most likely to be a consequence of climate change, population growth

and other factors that can potentially contribute. The City of Tshwane indicated liveability

as one of its objective through its Tshwane Vision 2055. This can only be achieved

through sustainable means of service delivery and sanitation. A strategic plan that will

address the impending water scarcity will help the municipality achieve strategic objective

1. The proposed project will assist the city in obtaining an in-depth understanding of the

current status of water demand and supply of water and the importance of drafting ways

to deal with the impending potential shortages.

Strategic Objective 1: Provide sustainable services infrastructure and human settlement

management

Strategic Objective 6: Continued organisational development, transformation and

innovation

This project at its core is to assist the City of Tshwane in developing a strategy that will

curb the impacts of the potential future water scarcity in the most efficient and effective

manner. Through this project, the city will be able to develop its mechanism to improve

on the water supply and management which will in turn contribute towards the

implementation of transformation and innovation in the city as stipulated in the Tshwane

2055 vision.

3 DEFINING WATER SCARCITY

The Food and Agriculture Organisation of the United Nations defines water scarcity when

demand for freshwater exceeds the supply. This arises as a consequence of changing

weather patterns as well as a high demand from all water-using sectors and is directly

related to human interference with the natural water cycle. The three main dimensions

that characterise water scarcity are: physical lack of water availability to satisfy demand;

6

level and quality of infrastructure that controls storage and distribution; and the

institutional capacity to provide the necessary water services.

Demographic pressures, the rate of economic development, urbanisation and water

pollution are all putting unprecedented pressure on existing water resources. Human

pressure on water resources increases as disposable incomes swell. Increasing incomers

lead to a rise in the per capita demand for food and personal care. The consumption in

meat and dairy products escalate and directly relates to an excess demand for water. Of

all economic sectors, agriculture accounts for 70 percent of global freshwater

withdrawals.

A study by Maddocks, Maddocks, Young and Reig, 2015, found that 33 countries face

extremely high water stress in 2040. These countries are located adjacent to the

Mediterranean and Caspian Seas and include Saudi Arabia, Libya, Iran and Pakistan.,

Namibia, Botswana and South Africa could face an especially significant increase in water

stress by 2040. This means that businesses, farms, and communities in these countries

in particular may be more vulnerable to scarcity than they are today.

With regional violence and political turmoil commanding global attention, water may seem

to be an absolute minor issue. However, drought and water shortages in Syria likely

contributed to the unrest that stoked the country’s 2011 civil war. Dwindling water

resources and chronic mismanagement forced 1.5 million people, primarily farmers and

7

herders, to lose their livelihoods and leave their land, move to urban areas, and magnify

Syria’s general destabilization.

The U.N. High Commissioner for Refugees says nearly 300,000 refugees and migrants

have arrived in Europe across the Mediterranean Sea in the first 8 months of 2015. The

chaotic scenes of thousands of desperate people trying to enter the former Yugoslav

Republic of Macedonia is continuing and is unstoppable. The majority of these migrants

fleeing violence and conflict in Syria, Afghanistan and Iraq. Although not substantiated,

the present mass migration into Europe might be driven by water scarcity in home

countries.

4 BACKGROUND

The purpose of the City of Tshwane Water Resource Master Plan (WRMP) was to

investigate the possible upgrading or extension of the City’s own water resources, with a

view to reduce the dependence on imports from the Vaal River basin (via Rand Water).

It also concerns the Crocodile River basin and the Olifants River basin, which both receive

significant sewer return flows from the City that influence the yields of the local water

resources and water allocations to downstream users. The recommendations of the

WRMP would promote the strategic objectives set by National Government relating to

water resource management, while ensuring the long term development objectives of the

City as proposed in the Tshwane 2055 Vision.

8

4.1 THE CITY OF TSHWANE WATER RESOURCES MASTER PLAN (WRMP)

The content in this section relies heavily on the WRMP and relate directly to sections

within the report. The implementation of the WRMP will enable the City to be more water

resourceful in order to cope with water interruptions from bulk water service providers. A

case in point is the Rand Water bulk water supply interruption during September and

October 2014 effecting the south-western parts of the City.

4.2 PRESENT BULK WATER DISTRIBUTION SYSTEM

The City of Tshwane existing bulk water distribution system as shown on Map 2.02 and

serving the current Average Annual Daily Demand of 987 Ml/d consists of:

50 Raw Water connections

9

4 Major own Water Treatment Plants (WTP’s)

2 Minor own WTP’s

3 MW owned WTP’s

A few privately owned package WTP’s

A number of fountains, Bore Holes and springs

700 km of bulk pipelines

140 storage reservoirs on 110 sites with total capacity 1 690 Ml

40 elevated water towers with total capacity 13 Ml

70 pumping stations

235 primary Water Distribution Zones (per reservoir, water tower, or direct link to bulk system)

The system has a replacement value of ±R 12 billion.

4.3 CURRENT WATER SOURCES AND DEMAND

The City of Tshwane currently has an average potable water demand of 987 Ml/d.

Approximately 72% of the demand is supplied by Rand Water Board, the main water

source being the Vaal River. The remainder is generated internally by CoT’s own

fountains, springs, boreholes and Water Treatment Plants (WTP), of which Rietvlei WTP

(40 Ml/d), Roodeplaat WTP (60 Ml/d), Bronkhorstspruit (54 Ml/d) and Temba WTP (60

Ml/d) are the largest. Magalies Water Board (MW) also owns and operates three WTP’s

which supply CoT, namely Klipdrift WTP (18 Ml/d), Wallmannsthal WTP (12 Ml/d) and

Cullinan WTP (16 Ml/d).

4.4 FUTURE WATER DEMAND AND SEWER FLOWS

In accordance with the City of Tshwane current water and sewer Master Plan which have

been based on the City’s approved Spatial Development Framework, the CoT potable

water demand is set to increase over the next 40 to 50 years from 987 Ml/d to 2600 Ml/d,

with an associated increase in sewer return flows to 1600 Ml/d. The anticipated future

water demands and sewer return flows are based on a population growth rate of ±2% p.a.

10

4.5 STATUS QUO AND FUTURE WATER TREATMENT PLANT (WTP) EXPANSION.

4.5.1 Status Quo of Rietvlei WTP

The main findings of the status quo analysis of Rietvlei WTP are listed in the table below.

Evaluation Finding

Current capacity and utilisation The plant can treat 40 Ml/d at present. Average production from the plant is

37 Ml/d. Grootfontein adds another 5 to 10 Ml/d.

Raw water quality The water source remains eutrophic and is typical of a metropolitan water

source which receives a substantial return flow of treated waters from

municipal wastewater treatment plants. Typical characteristics are: low

turbidity, high organic carbon loads, high levels of eutrophication with a high

likelihood of taste and odour occurrence.

Process evaluation The installed process will be adequate for the raw water quality profile once

the ozone installation is finalised.

Final water quality Treatment plant performance for the years reviewed (2010 to 2012), was

excellent. Water samples need to be monitored more frequently for priority

compounds related to the debate on EDCs (Endocrine Disrupting

Compounds). Water quality performance can be expected to drop due to the

current unavailability of the GAC plant. A downward trend in aesthetic and

operational quality was noted in the data for 2012.

Licensing Rietvlei was licenced for the abstraction of 65.7Ml/d (Act 25 of 1929)

Request for additional 20.Ml/d Water Use in 1986 (Application Number

B190/1/134) was received by DWA but no decision was made on this request.

However when Act 25 of 1929 was repealed by the National Water Act (Act 36

of 1998), no licence for Rietvlei was issued by DWA

Licensing status for this plant must be re-established with the Department of

Water Affairs

State of repair and maintenance A maintenance and repair backlog is hampering production at the plant:

- The GAC facility is not operational.

- Two of six high lift pumps were operational at the time of the inspection.

- GAC furnace is not operational.

Staffing The plant does not comply with the staffing requirements of Regulation 2834

and Regulation 17.

Potential for expansion Hydrological studies indicate it may be possible to expand this plant to 240

Ml/d in 2055. This level of expansion will require the transfer of water from

the Kaalspruit (Olifantsfontein Waste Water Treatment Works) to the Rietvlei

catchment. The raw water abstraction facility must be expanded as well.

11

Evaluation Finding

The site is constrained and expansions will have to be placed to the north of

the current site.

Specific areas of concern 1. The ozonation and laboratory facility must be finalised and commissioned.

2. Staffing must be brought in line with Regulation 2834.

3. Maintenance planning and implementation must be improved.

4. Water quality monitoring must be expanded to cover EDCs.

4.5.1.1 Immediate interventions required

There is a critical lack of qualified process controllers to operate the Water Treatment

Works. The staffing philosophy should be investigated with a view to outsourcing the

entire water works or a staffing policy to ensure that the process control staff is

retained.

The GAC filtration unit has not been in operation due to flooding of the entire GAC

gallery on the 27th July 2012. Two of the six actuators on each GAC filter need to be

replaced to allow this unit to function properly once again.

Two of the six high lift pumps were not working at the time of the site visit and therefore

the Rietvlei WTW was unable to operate at full capacity for several weeks.

The newly built ozonation facility has not been commissioned yet and a maintenance

plan and contract will need to be drawn up.

No additional virgin/reactivated GAC is available to top up or replace the GAC filter

beds. The Mintek furnace has been problematic to operate and needs to be re-

commissioned.

The clear water tank where disinfection occurs should be retrofitted with baffles in

order to facilitate proper disinfection and this should be done before the

implementation of chloramination disinfection.

12

A maintenance plan needs to be drawn up to ensure that the infrastructure is well

maintained.

Persistent Organic Pollutant levels and other organic pollutants emanating from the

return flows from the sewage works upstream as described in the research undertaken

at Rietvlei, should be monitored and researched both before and after the ozonation

facility is put into operation.

4.5.1.2 Possible extensions

Water resource from natural flow and sewer return flows is available to increase the

capacity of the Rietvlei WTP from existing 40 Ml/d to 140 Ml/d over time. A transfer

scheme to pump the effluent from the Olifantsfontein Waste Water Treatment Works into

the Rietvlei basin can be considered. This will add additional resource which will allow

extending the Rietvlei WTP capacity to 240 Ml/d over time, in 3 phases:

2020 +100 Ml/d

2033 +50 Ml/d

2045 +50 Ml/d

13

4.5.2 Status Quo Roodeplaat WTP

The main findings of the status quo analysis of Rietvlei WTP are listed in the table

below.

Evaluation Finding


42.95 Ml/d.

Raw water quality The water source remains eutrophic and is typical of a metropolitan water

source which receives a substantial return flow of treated waters from

municipal wastewater treatment plants. Typical characteristics are: low

turbidity, high organic carbon loads, high levels of eutrophication with a high

likelihood of taste and odour occurrence. Iron and Manganese levels are

problematic. Microcystin levels are high and require special attention.

Process evaluation The installed process will be adequate for the raw water quality profile.

Manganese was the only determinand tested which exceeded the SANS

requirement on the 90th percentile. The Microcystin-LR problem will

probably be addressed through the recent addition of the ozone and GAC

process units.


excellent. A downward trend in aesthetic, operational and chronic health

determinands was noted in the data for 2012. Water samples need to be

monitored more frequently for priority compounds related to the debate on

EDCs (Endocrine Disrupting Compounds). Diligence is required in monitoring

and controlling algal toxin production in the raw water source.

Licensing WUL Certificate No. 26061886 is in place for 90 Ml/d.

State of repair and maintenance No maintenance plans are currently in place and several preventive and

reactive maintenance issues are evident on site.

Staffing The plant does comply with the staffing requirements of Regulation 2834 and

Regulation 17.

Potential for expansion Hydrological studies indicate that it may be possible to expand this plant to

240 Ml/d in 2055.

Initial expansions (to 120 Ml/d) may be accommodated on site but additional

land must be procured for the full expansion. The current abstraction facility

may be limited to 90 Ml/d and further expansion of the facility will be

required.

Specific areas of concern 1. A maintenance strategy has to be developed and implemented for this

site.

2. There are no stand-by chlorination facilities available on site. This presents

a major risk to public health.

3. Microcystin-LR levels are high and require special attention.

14

4. Water quality monitoring must be expanded to cover EDCs.


No standby chlorination facility is in place and this poses a significant risk to the drinking water

quality if the current unit requires maintenance or repairs.

UV irradiation facility has been out of order for several years and should be commissioned as

an additional barrier.

On-line turbidity monitoring devices are out of order and need to be recommissioned.

Ozone dosing facility is working as a “black box” and is not optimised by the process

controllers to ensure optimum dosage.

Maintenance plans are lacking and CoT should consider the development of a maintenance

strategy and maintenance contracts to ensure that the entire water infrastructure is maintained

in good order.

Due to the impact of both Zeekoegat and Baviaanspoort WWTW effluent which is discharged

into the Roodeplaat dam, increased monitoring of emerging contaminants such as estradiol,

lindane and DDT should be done and this increased monitoring should include algal species

and algal toxins (Microcystin) as this has been found to be problematic in this catchment.

4.5.2.2 Possible extensions

Yield from natural flow and sewer return flows is available to increase the capacity of the

Roodeplaat WTP from existing 60 Ml/d to 240 Ml/d in 4 phases:

2014 +30 Ml/d (already in process)

2025 +50 Ml/d

2035 +50 Ml/d

2045 +50 Ml/d

15

4.5.3 Status Quo Themba WTP


below.

Evaluation Finding


52.44 Ml/d.

Raw water quality Temba’s source water is problematic due the nature of the main inflow into

the Leeukraal Dam. The dam is mainly fed by the Rooiwal WWTW. Problem

parameters include: colour, turbidity total organic carbon, aluminium, iron,

lead, manganese, zinc, ammonia, nitrite and all the microbiological

parameters such as E.coli, heterotrophic plate count and total coliforms. The

Leeukraal Dam is eutrophic. Tests in the raw water indicate that EDC’s are a

concern.

Process evaluation The currently installed process is not adequate for the present raw water

quality profile. Process upgrades are currently being planned for the plant

and the inclusion of ozone and GAC will greatly improve the performance of

the plant.


inadequate. Some improvements were noted in 2012, possibly due to the

temporary inclusion of chlorine dioxide in the process. The application of

chlorine dioxide has however been halted due to procurement issues. In

prior years the plants performance as measured in terms of aesthetic,

operational and acute health determinands was poor. Water samples need

to be monitored more frequently for priority compounds related to the

debate on EDCs (Endocrine Disrupting Compounds).

Licensing WUL No. 27/2/2/A623/103/1 is in place for 130 Ml/d.

State of repair and maintenance No maintenance plans are currently in place and several preventative and

reactive maintenance issues are evident on site.

16

Evaluation Finding

Staffing The plant does comply with the staffing requirements of Regulation 2834 or

Regulation 17.

Potential for expansion Hydrological studies indicate it may be possible to expand this plant to 180

Ml/d in 2055. This is inclusive of a transfer of production responsibility of 28

Ml/d from the Temba WTW to the Klipdrift WTW for delivery into the

Moretele area.

Initial expansions (to 120 Ml/d) may be accommodated on site but additional

land must be procured for the full expansion.

Specific areas of concern 1. Plant performance is inadequate.

2. The raw water quality is problematic. The retention time in the Leeukraal

Dam is very low and this is problematic in terms of providing a buffer

between the dam inflows and the abstraction for potable water treatment.

3. A number of problem determinands can be linked directly to the Rooiwal

WWTW. It is therefore imperative that the WWTW delivers the best water

quality possible through process refinement and diligent operation.

4. The site has been shown to have problematic EDC levels. This must be

studied and dealt with urgently.

5. A maintenance strategy has to be developed and implemented for this

site.


The raw water quality is concerning as the ammonia levels are regularly 6mg/L and this makes disinfection exceptionally difficult and also forms disinfection by-products such as Tri-halo-methanes.

Spillage of water into the road reserve from the sludge lagoons needs to be investigated.

Sludge lagoons are very full and there is no schedule available for emptying the lagoons.

Ultraviolet radiation dosing is in place at the final water produced before chlorination but has been discontinued.

The endocrine disrupter chemical content of the raw water supplied to the Temba WTP needs to be studied in more detail and on a regular basis to ensure that the GAC and ozonation technology which is to be installed during the coming upgrade, will be sufficient to remove all the organic pollutants within the raw water supply to protect the health of the consumer.

17

Maintenance is neglected and a maintenance plan should be compiled and provision made for timeous replacement or repair of broken equipment.

4.5.3.2 Possible Expansions

The Temba WTP with augmentation form the MW Klipdrift WTP serves the Temba,

Kudube, Stinkwater and New Eersterus areas, with cross boundary supply into the

southern portions of Moretele.

In line with this requirement, there is sufficient yield available from natural flow and sewer

return flow to increase the capacity of the Temba WTP from existing 60 Ml/d to 180 Ml/d

in 3 phases:

2016 +60 Ml/d

2035 +30 Ml/d

2045 + 30 Ml/d

4.5.4 Status Quo Bronkhorstspruit WTP


below.

Evaluation Finding

Current capacity and utilisation The plant can treat 54 Ml/d at present. Average production from the plant at

capacity for the period 2010 to 2012 was 55.32 Ml/d.

Raw water quality The raw water is characterised by low turbidity and medium levels of organic

carbon. Low levels of eutrophication with low levels of microbiological

contamination are evident.

Process evaluation The installed process is adequate for the present raw water quality profile.

18


adequate with increased compliance failures noted in 2012 for microbiological

determinands. Micro compliance in 2012 was only 72.7% and this will result in

increased health risks and poor performance during the Blue Drop assessments.

The microbiological compliance was related to chlorination failures. An overall

downward trend was noted in the compliance data for the period 2010 to 2012.

Turbidity removal is problematic during rainy seasons due to the poor condition

of the filters. This is a result of the insufficient backwash capability of the filters.

Other problem parameters include colour, aluminium, iron and manganese.

Licensing WUL are in place for a total of 27.4 Ml/d for Bronkhorstspruit, with additional

allocations for Summerplace and Bronkhorstbaai This is not sufficient to cover

the current abstraction and an additional WULA must be submitted.

State of repair and maintenance Maintenance services are currently being provided by Rand Water.

Due to the maintenance backlogs on the plant the current situation remains

unsatisfactory. Although Tshwane has started to address this, several process

critical items remain problematic. A detailed maintenance plan must be

developed and implemented for this site.

Staffing The plant does comply with the staffing requirements of Regulation 2834 or

Regulation 17.

Potential for expansion Due to a limitation in raw water, no expansions are possible at this plant. The

site itself is not a limiting factor as space is available.

Specific areas of concern 1. Plant performance has deteriorated in 2012.

2. Several process critical equipment items are not operational or do not have

back-up. This presents high levels of performance risk to the plant. This includes

raw water pumps and transformers.

3. The lack of proper air scours for the filters result in dirty filter media. The

result of this is poor turbidity compliance performance.

4. The backwash water return bypasses coagulation and flocculation as it is

returned to the process stream leading to increased water quality failures.

5. Poor air distribution in the DAF facility is leads to poor reactor performance.

6. A maintenance strategy has to be developed and implemented for this site.

19


Raw water pumps were out of order and this has been a problem for a significant period of time. A preventative maintenance plan is required to ensure that breakdowns are attended to timeously.

A Class V Supervisor/ Plant Manager should be available on-site to manage this WTP as well as sufficient number of qualified staff.

Transformer on the site has no back up facility and the current transformer is on loan from another location. Should this transformer fail, the water supply will fail.

Sand filtration system does not backwash efficiently and there is no facility for air scour. This is a significant critical control point which should be upgraded to ensure good quality water from this WTP.

Recovered water from the backwash is not chemically dosed before being added to the process train again which may concentrate unwanted protozoan parasites and other pathogenic organisms in the treatment process.

The cause of the poor air distribution at the DAF units should be investigated as this does not allow for good separation of the flocculated materials and puts undue pressure on the sand filtration system.

A good maintenance plan and team should be available to ensure that maintenance and repairs are carried out timeously.

20

4.5.5 Status Quo Bronkhorstbaai WTP


below.

Evaluation Finding

Current capacity and utilisation The plant can treat 0.5 Ml/d at present. The average production for

this plant is not known at this time.

Raw water quality The raw water is characterised by low turbidity and medium levels of

organic carbon. Low levels of eutrophication but high (non blue-

green) algal counts with low levels of microbiological contamination

are evident.

Process evaluation The installed process is not adequate for the present raw water

quality profile. The addition of DAF must be considered to prevent

algal blinding of the filters.

Final water quality Treatment performance has been poor. Problem determinands

include turbidity, total coliforms, colour, uranium and aluminium.

Poor compliance in these determinands is generally related to

operational failures.

Licensing WUL is in place for a total of 0.1 Ml/d. This is not sufficient to cover

the current abstraction and additional WULA must be submitted.

State of repair and maintenance A detailed maintenance plan must be developed and implemented for

this site.

Staffing The plant complies with the staffing requirements of Regulation 2834

or Regulation 17.

Potential for expansion Expansion of the plant to 5.5 Ml/d by 2055 must be considered. The

site allows for such an expansion. The plant will be used to augment

supplies from other smaller plants in the area (such as Summerplace)

which will be closed going forward. This is considered despite the

shortage of raw water in this catchment as this plant places a very

small demand on the overall system.

Specific areas of concern

1. Overall plant performance is poor.

2. The raw water abstraction system requires refinement and a risk

review.

21

3. Overall chemical dosing needs to be refined. Dosing control needs

to be reviewed as well as chemical management and storage.

4. Disposal of sludge into the dam close to the abstraction point is

problematic.

5. A maintenance strategy has to be developed and implemented for

this site.


Raw water pump and methodology of extraction needs to be reviewed to allow for sustainable raw water supply despite floods or droughts.

The dosing protocol, pumps and storage facilities for dosing the flocculent needs to be replaced with duty and standby pumps, new storage facilities for flocculent as well as jar testing apparatus.

Laboratory space at the Water Treatment Works is not available and makes testing difficult.

No internet access is available and this makes communication very difficult.

The disposal of sludge and wash water into the dam can impact on the quality of the incoming raw water and it is recommended that an alternative method be investigated for disposal of the wash water and sludge.

Replacement of the final water tank with two new tanks should be considered as the current tank is rusting through and no alternative water storage is in place.

A maintenance plan and contracts need to be implemented in order to prevent poor service delivery.

Staffing is insufficient and additional staff needs to be employed with sufficient skills and knowledge to enable them to operate and maintain the WTP.

4.5.5.2 Possible Expansions

The Bronkhorstbaai WTP currently serves only the Bronkhorstbaai resort (0,35 Ml/d

AADD), but since it is under ownership of CoT, it has been earmarked for further

development in order to consolidate all the water treatment for all the resorts around the

Bronkhorstspruit dam in one facility.

There is no water resource available in the Olifants River catchment to allow for the

required extension of the WTP. However, it may be justified by consolidation of all the

22

existing water rights of all the resorts around the dam, and the argument that just a little

extra water should be augmented from the RW system to Cullinan and Bronkhorstspruit,

in effect freeing up some of the water resource in the Olifants catchment. In line with this

argument it is proposed to increase the capacity of the Bronkhorstbaai WTP from the

existing 0,55 Ml/d to 5,5 Ml/d in 3 phases:

2016 + 2,00 Ml/d

2030 + 1,75 Ml/d

2045 + 1,20 Ml/d

4.5.6 Summary

The WRMP will ensure sustainable water provision to the City by reducing the

dependency on Rand Water to 1816 Ml/day by 2022. The capacity of the water treatment

plants will be increased as follows:

Stepwise increase the Rietvlei Water Treatment plant (WTP) capacity from the 40

Ml/d to 140 Ml/d;

A further expansion the Rietvlei WTP to 240 Ml/d once water is transferred from

Ekurhuleni’s Olifantsfontein Waste Water Treatment Works to Rietvlei Dam;

Expand the Roodeplaar WTP capacity from 60 Ml/day to 240 Ml/day;

Expand Temba WTP from 60 Ml/day to 180Ml/day to address peak summer

demand;

The core objective of the WRMP is to ensure sustainable water supply to the City of

Tshwane. This will be done by expanding the water and waste water treatment plants

within the City. Although infrastructure is critical for water provision, there are unforeseen

risks that potentially could negate all the objectives of the WRMP.

5 CITY OF TSHWANE WATER REQUIREMENTS AS FROM WRMP

23

The section below aims to briefly outline the results of the city’s water requirements

from both catchments. This analysis was done as part of the Water Resource Master

Plan and is provided to further give context of the interdependency between the sources

of water and how infrastructure interacts.

The city of Tshwane sits between two catchment areas, see Figure 5.1. Thus the

analysis focused on each catchment separately.

24

Figure 5.1: Simplified schematic of the CoT water resource system

25

5.1 Crocodile West River Catchment

The updated CoT future water requirement and return flow projections are both higher than the

Crocodile Reconciliation Study projections. The growth is however dependent on developments that

may or may not realise as anticipated, which could have a significant effect on sewer return flows and

therefore on the yield of the local water resources. A conservative approach was thus adopted by

assuming a projection halfway between the CoT and Reconciliation Study projections for both water

requirements and return flows as illustrated in the figures (Scenario 1).

The analysis results based on the adopted scenario concluded that the total current (2013) surplus

yields available from Rietvlei Dam, Roodeplaat Dam and Olifantsfontein WWTW (Hartebeespoort Dam)

are 4.7 Mm3/a, 9.2 Mm3/a and 27.6 m3/a respectively and are projected to increase to 33.0 Mm3/a,

63.6 Mm3/a and 44.5 Mm3/a respectively by 2057.

The surplus yields available from the analysis are dependent and can be influenced by the following:

The future return flows generated by the CoT and Ekurhuleni WWTWs (Hartebeesfontein &

Olifantsfontein) which in turn are effected by developments in some large areas, which may or

may not realise as anticipated in the projections.

The MCWAP water requirement projection which historically has been adjusted according to the

uptake in planned developments.

It is recommended that the return flows and water requirements are continuously monitored and

that the analysis is revisited, should noticeable differences occur between the actual recorded

figures and the assumed scenario.

Changes in water requirements and return flows in the catchment outside of the CoT supply also

have an impact on the overall water balance of the system. It is also recommended the analysis

is revisited, should noticeable differences occur between the actual recorded figures and figures

used in the analysis.

The monitoring of all the water requirements and return flows are reviewed as part of overarching

DWA Crocodile Reconciliation Strategy, which is continuously updated.

26

5.2 Upper Olifants River Catchment

The total CoT projection is significantly higher than the Olifants Reconciliation Study projection. The

main differences lie in the water requirement projections for Bronkhorstspruit Town and surrounds and

also the Western Highveld South projection. The information from the Development of a Reconciliation

Strategy for All Towns in the Northern Region proved to be more in line with the CoT Projection.

The results of the Olifants Reconciliation Study WRPM analysis (based on the Olifants Reconciliation

Study water requirement projection with WCDM) showed that the total surplus yield from Rust de Winter

Dam is required to ensure users are not restricted.

The results of the WRPM analysis with the CoT projection showed that the users cannot be supplied

according to their required assurance criteria and the following interventions will be required to ensure

sufficient water resource availability:

Total surplus yield from Rust de Winter Dam required as support.

The successful implementation of WCDM initiatives to achieve total savings of 12.8 Mm3/a.

Additional augmentation of approximately 14 Mm3/a.

The current Rust de Winter Dam yield is of low confidence and will be confirmed by the DWA

Olifants Reconciliation Strategy which is continuously updated.

It is envisaged that the feasible option for the required support will most likely be additional

supply from Rand Water.

5.3 WATER QUALITY AND PROCESSES REQUIRED FROM THE WRMP

It was noted in the previous section that the additional water resources available to CoT emanate from

upstream municipal discharges. This situation is not much different from its situation previously but,

given the immense public and scientific interest in the implications of this situation, detailed

consideration of the status quo is required.

27

5.3.1 Water Reuse

CoT will have to acknowledge that the exploitation of water in its catchment, which contains substantial

return flows, as proposed, goes beyond conventional treatment, towards the formal practice of water

reuse. Water reuse practices are now formally categorised, namely1:

De facto reuse: A situation where the reuse of treated wastewater is, in fact, practiced but is not officially recognized (e.g., a drinking water supply intake located downstream from a WWTW discharge point).

Direct potable reuse (DPR): The introduction of reclaimed water (with or without retention in an engineered storage buffer) directly into a drinking WTP, either co-located or remote from the advanced WWTW.

Indirect potable reuse (IPR): Augmentation of a drinking water source (surface or groundwater) with reclaimed water followed by an environmental buffer that precedes drinking water treatment.

These boundaries between the various modes of reuse are marginal. Questions regarding what

constitutes an environmental buffer, for example, have not yet been addressed adequately, although it

is an important issue among engineers and scientists active in the reuse field at present as will be

indicated later.

What is important however is that the CoT acknowledges the current and future reality that it has,

without much intent in this regard, been thrust into a reuse situation and that the City will have to adapt,

where necessary, its water management practices professionally and accordingly.

A successful and professional potable water reuse strategy, by universal international acceptance,

hinges upon a number of supporting technical pillars:

Close control, and knowledge of the wastewater being collected for later potable water reuse

Comprehensive, consistent wastewater treatment to ensure effluent of predictable, acceptable quality

An environmental buffer which acts both for blending of wastewater effluent and other contributing sources, as well as a means to dispel the public “yuck factor” perception associated with domestic wastewater

Water treatment technology specifically chosen and designed to ensure a safe, acceptable water supply

Proper distribution of the water to a large part of the population to dispel the notion of a few having to “suffer” the indignity of being the only ones targeted for potable reuse

A comprehensive water quality operational monitoring programme across the full urban water use cycle to detect potential hazards in time, and to ensure the best water quality at all times

A comprehensive water quality compliance monitoring programme with full transparency, regular review by independent experts and regular reporting to the public. This programme should be extended beyond the normal drinking water standards where necessary

1 http://www.scgcorp.com/WaterReuse2013/Definitions.pdf

28

5.3.2 REUSE PROCESSES

The implication is that proven treatment technologies are available, and continue to be refined, which

allow for the reliable treatment of any water to any level of quality. In the case of CoT, the intent is

potable reuse. The onus is therefore on CoT to ensure that the correct treatment trains are selected

and implemented, and that these are operated appropriately, in order to derive the desired outcome.

The graphic in Figure 5.2 below summarises this principle.

Figure 5.2: Treatment technologies are available to achieve any desired level of water quality (Taken from EPA/600/R-12/6182)

Figure 5.3 illustrates the full reuse process as employed in Windhoek which includes pre-ozone,

enhanced coagulation and flocculation, dissolved air flotation, rapid gravity sand filtration, ozone, 2

applications of biologically active carbon, granular activated carbon, ultrafiltration, stabilisation and

finally disinfection. The City’s current WTP’s at Rietvlei and Roodeplaat share a very similar process

with the following exceptions: limited application of enhanced coagulation and flocculation, no

deliberate application of biologically active carbon and no application of membrane filtration. It must

be noted that it may be possible to retrofit both these WTPs to the specification of the full reuse plant if

required. Given Tshwane’s familiarity with the Rietvlei and Roodeplaat process, as well as the proven

performance of this process, it was decided to base an analysis of Tshwane’s future expansions on the

adoption of the same process. It should be noted that the proposed process is also supportive of the

process analysis performed as part of the preparation of this report’s companion report3. A cost

sensitivity analysis was performed on the basis of the full reuse process train in order to assess pricing

differences between the full and “partial” reuse processes.

2 EPA/600/R-12/618 (2012) Guidelines for Water Reuse. U.S. Environmental Protection Agency. September 2012. 3 City of Tshwane Water Resources: Status Quo Assessment of Current Potable Water Supply Facilities. Project No CB82/2010. Dated December 2013.

Prepared by CSVwater Consulting Engineers (CSV Project 1306)

29

It is quite possible that, over time, other technologies may arise, or that current technologies are refined,

which will be more cost effective to install and operate. The newer technologies may be considered at

a time when the WTPs become due for further upgrading or expansion. For the purposes of comparison

and costing however, a fairly standard process, which will be able to deliver the required quality of

potable water, will be used. This allows for a balanced evaluation and feasibility study. The challenge

to improve on the recommendations made in this report, lies with the engineers and scientists that will

design and install the WTPs at the appropriate time.

30

Figure 5.3: Possible process configurations

Ultrafiltration

Stabilisation

Disinfection

Ultrafiltration

Stabilisation

(Enhanced) Coagulation

& Flocculation

(Enhanced) Coagulation

& Flocculation

Settling

Biologically Active

Carbon

Granular Activated

Carbon

Distribution Distribution

Pre - ozone

Dissolved Air Flotation

Settling

Biologically Active

Carbon

Granular Activated

Carbon

Pre - ozone

Dissolved Air Flotation

Rapid Gravity Sand

Filtration

Ozone contact

Biologically Active

Carbon

Rapid Gravity Sand

Filtration

Ozone contact

Biologically Active

Carbon

Disinfection

TREATMENT OPTIONS

FULL RE-USE COMPLIANCE

INTERIMPROCESS

31

6 INPUT MODELLING METHODOLOGY

In this section the methodology of the population model and GDP model will be discussed. The aim of

this section is to provide a clear understaning of the technical approach applied that ultimately formed

key inputs into the water Demand modelling in the following section. The water demand modelling

Framework below outlines the approaches applied.

6.1 GDP Modelling

The method utilised to forecast GDP was inspired by (Mikhael, Kamel and Khoury, 2010) who utilised

a Vector Autoregressive Model with Exogenous Variables (VARX) which is a variant of the VAR

model. Based on the modelling and selection of variables the following variables were selected for

this model.

M3 – It represents the impact of monetary policy and Liquidity;

Real Exports of Goods and Services – Represents an outflow of economic activity to the

global market;

Real GVA of the Construction Industry – Used as a proxy for investment in real estate and

infrastructure;

Real Final Household Consumption – There is a clear linkage between household

consumption and GDP; and

Real GDP – GDP at constant 2005 prices

Wat

er D

eman

d M

od

ellin

g

Water Category Breakdown

GDP Forecast

Categories:

Businesses

Farms and Agricultural Holdings

Industrial

Large consumers

Population Forecast

Categories:

Residential houses

Flats

Clusters

Internal Growth Rate

Categories

Education

Government Institutions

Parks

Unknown (labelled as provided in data

32

The Granger Causality test was used to determine the endogenous variables for the model. From this

test it was observed that GDP and M3 were endogenous and all other variables were exogenous.

Based on preliminary checks it was determined that only 1 lag is necessary. The model was then

utilised to forecast GDP to the year 2055.

Then utilising the constant share method over the base period of available data the GDP for Tshwane

was forecasted. To obtain a similar GDP growth pattern. The interest here is not the nominal value of

GDP but rather the rate at which it grows. This growth rate was then applied to the relevant GDP

forecast categories as indicated in the framework.

Table 1: Parameter estimates of the VARX model with applied Restrictions and fit statistics

Model Parameter Estimates

Equation Parameter Estimate Standard

Error t Value Pr > |t| Variable

GDP XL0_1_1 0.37235 0.07447 5.00 0.0007 Exp(t)

XL0_1_2 0.85742 0.10428 8.22 0.0001 HH(t)

XL0_1_3 0.00000 0.00000 GVA(t)

XL1_1_1 0.00000 0.00000 Exp(t-1)

XL1_1_2 -0.66882 0.15503 -4.31 0.0019 HH(t-1)

XL1_1_3 -0.00000 0.00000 GVA(t-1)

AR1_1_1 0.79081 0.06866 11.52 0.0001 GDP(t-1)

AR1_1_2 -0.00000 0.00000 M3(t-1)

M3 XL0_2_1 1.84461 0.42830 4.31 0.0020 Exp(t)

XL0_2_2 0.00000 0.00000 HH(t)

XL0_2_3 14.77738 3.28323 4.50 0.0015 GVA(t)

XL1_2_1 0.00000 0.00000 0.00 1.0000 Exp(t-1)

XL1_2_2 0.00000 0.00000 HH(t-1)

XL1_2_3 0.00000 0.00000 0.00 1.0000 GVA(t-1)

AR1_2_1 -1.15257 0.17528 -6.58 0.0001 GDP(t-1)

AR1_2_2 0.89049 0.05979 14.89 0.0001 M3(t-1)

Testing of the Restricted Parameters

Parameter Estimate Standard

Error t Value Pr > |t| Equation

Restrict0 1.84006 1.24051 1.48 0.1763 XL0_1_3 = 0

Restrict1 -3.88839 8.69215 -0.45 0.6665 XL1_1_1 = 0

Restrict2 1.01421 1.09127 0.93 0.3799 XL1_1_3 = 0

Restrict3 77.30776 44.44917 1.74 0.1202 AR1_1_2 = 0

Restrict4 4.97546 2.80015 1.78 0.1135 XL1_2_1 = 0

Restrict5 0.81230 1.09010 0.75 0.4775 XL1_2_2 = 0

Restrict6 -0.26129 0.23770 -1.10 0.3036 XL1_2_3 = 0

Information Criteria

AICC 41.31748

HQC 39.72223

AIC 39.64428

SBC 40.42848

FPEC 1.937E17

33

6.2 Population Modelling

This section briefly outlines the population modelling approach followed.

6.2.1 Weighted Average

A large part of the calculations and growth in elements were calculated using the weighted average

and therefore it is imperative that a brief insight into this approach be provided. The Weighted average

formula is calculated as follows:

Which means:

It has to be noted that the weights have to be non negative.

The assigning of weights was done by assuming that weights increase systematically as the years

progress, thus resulting in the movements in recent years being weighed more.

In this modelling the weights were assigned as follows:

Year Weight

1996 1

1997 2

…

…

2011 16

The resulting mean was used to extend trends and growth patterns in the model as it was conservative

and well behaved.

34

6.2.2 The starting point

In order to start the population analysis it was imperative that the following data be obtained

Census 1996

Census 2001

Census 2011

Community Survey 2007

Mid-Year population Estimates 2002 – 2030

The Census data was used to obtain Tshwane population Shares for 1996, 2001 and 2011. Then using

linear interpolation the estimates for all the years in-between was calculated using the following formula

(Kaw, 2011)

𝑑 = 𝑑1 + 𝑔 − 𝑔1

𝑔2 − 𝑔2 (𝑑2 − 𝑑1)

The period from 2011 from 2030 was calculated using the weighted average of the adjustments in the

period 1996 to 2011 and added to the final period of the base year and so on till 2030.

This would allow for a progressive share in the Gauteng Population. Then using the same procedure

of linear interpolation the 2002 estimate for GP from the mid-year estimates and the 1996 census value

for Gauteng, this was then combined with the mid-year estimates to create a smoothed Gauteng 1996

– 2030.

Using these estimates and the Tshwane share calculations the first projection for Tshwane is done for

the period 1996 – 2030. However, this was not sufficient for the development of a proper estimate for

the period. The next section will articulate the further work done to get population estimates.

6.2.3 Projection Methods

In order to get a thorough estimates various approaches were utilised to obtain variance in the data.

35

6.2.3.1 ReX Estimates

Utilising estimates from Regional Explorer for Tshwane population and calculating the share of the

base period 1996 – 2011 and then using the weighted average to determine the growth trend. Using

the Gauteng estimates and the share values the ReX Tshwane projection was done.

Further estimates were created through utilising the Rex population data split by population group and

then determining the growth rates for the period 1996 to 2013. Then taking the difference between

respective growth rates for each population group and calculating the weighted moving average for

each group and adding it as a constant increment every year. This provided the third population

estimate set.

6.2.4 Ratio Methods

There are many means of trend extrapolation. Given the Gauteng population and the Tshwane base

population calculated. This was used to calculate various estimates using the methods that are outlined

below.

6.2.4.1 Constant Share Method

In the constant-share method, the smaller area’s share, Tshwane in this case, of the larger area’s

population, Gauteng, is held constant with respect to the share of population observed in the launch

year, in this case 1996. A projection can then be obtained by applying the share to Gauteng’s projected

population.

𝑃𝑖𝑡 = (𝑃𝑖𝑙

𝑃𝑗𝑙⁄ ) 𝑃𝑗𝑡

Where:

Pit = population projection for smaller area in target year

Pil = population for smaller area in launch year

Pjl = population for larger area in launch year

Pjt = population of larger area in target year

36

6.2.4.2 Shift Share Method

In contrast to the constant-share method, the shift-share method accounts for changes in population

shares over time.

Where





Pib = base year population for smaller area

Pjb = base year population for large are

Z = years in projection horizon

Y = years in the base period

The base year and the horizon projection variables used were both equal to 1 and the base year was

1996 and the projection horizon was a consecutive year at a time. Again, the larger area data required

is taken from the Gauteng 2030 projection set and the Tshwane data is obtained from the Census

interpolated data.

6.2.4.3 Share of Growth Merhod

The third ratio method focuses on shares of population growth. In this method, it is assumed that the

smaller area’s, Tshwane, share of population growth will be the same over the projection period as it

was during the base period (reference

Where:





Pib = base year population for smaller area

Pjb = base year population for large are

37

6.2.4.4 Population Regressions

The 6 population projections obtained formed the basis for the regression estimates. The regressions

were simple linear regressions with the aim to obtain further variation in the data to assist in creating a

wider spectrum of possible population figures that come from the same N.

Then the Census figures and the community survey figure for 2007 were taken and a new set of

interpolated figures were created using Linear interpolation, Newton’s Divided Difference Interpolation

and Lagrange interpolation was used to obtain 3 additional base estimates for the regression.

The Linear interpolation formula was provided the earlier section, below are the formulas for the Newton

Interpolation and the Lagrange interpolation.

6.2.4.5 Newton’s Divided Difference

The general form of the Newton’s Divided Difference can be expressed as follows (Kaw, 2011: 11)

Where

Assuming that we have a quadratic function the formula will look as follows

Where

38

Thus rewriting the above

A Difference Table to the fifth order is provided below for ease of reference

Source: Indian Institute of Technology,

6.2.4.6 Lagrange Interpolation

The idea behind the Lagrange Interpolation is to construct the interpolated polynomial in the form:

Thus in general the Lagrange Interpolation is represented by

39

Expanded it looks as follows

All the population estimates were taken as a base from 1996 to 2011. Then a trend and quadratic trend

variable was added to the dataset for the time period 1996 to 2030. This allowed for establishing a low

and high population examples and the further taking the midpoint as the medium scenario.

It was established that the appropriate models to utilise are autoregressive models:

𝑃𝑜𝑝𝑡 = 𝛽0 + 𝛽1𝑡𝑖𝑚𝑒𝑡 + 𝜌𝑃𝑜𝑝𝑡−1 + 𝜇𝑡

Where

Popt = population in year t

𝛽0 = Constant Intercept

Timet = Time trend variable in year

Popt-1 = population in year t - 1

𝜇𝑡= Stochastic Disturbance term

And

𝑃𝑜𝑝𝑡 = 𝛽0 + 𝛽1𝑡𝑖𝑚𝑒𝑡 + 𝛽1𝑡𝑖𝑚𝑒_2𝑡 + 𝜌𝑃𝑜𝑝𝑡−1 + 𝜇𝑡

Where

Popt = population in year t

𝛽0 = Constant Intercept

Timet = Time trend variable in year

Popt-1 = population in year t - 1

𝜇𝑡= Stochastic Disturbance term

Time_2t = Quadratic Time trend variable in year t

40

These two model sets were utilised alongside the original models to the 3 different population figure

sets. However, both the high and low population sets were subjected to an ANOVA test and a

Homogeneity of variance test.

The Homogeneity of Variance test on the low population shows that null hypothesis of equal variances

cannot be rejected (p = 0.84). The ANOVA also shows that the selected populations have equal means

(p = x) thus we can’t reject the null hypothesis. The fact that the means of the groups do not differ is

further corroborated by a pairwise mean comparison.

The High population scenario was also evaluated in the same manner and similar results were found,

it has to be noted that in the quadratic model some estimates were excluded due to unsuitability, once

these outlier results were removed the results were confirmed as shown below.

Furthermore, the average for the high and medium population projections were compared to an external

study done for the city recently by a demographics firm projecting the city’s population to 2035 from

2014. The ANOVA comparisons as shown below confirm that there is no significant difference between

the estimates.

6.2.5 Extended Term population Models

From the initial analysis is became apparent that an extended regression will have to be run. Thus the

original results were used and set to be the in sample values giving a base period of 1996 to 2030.

This was then used alongside a trend variable as before and population estimates were calculated for

the period 2031 to 2060.

The updated predictions resulted in the following problem. The initial periods clearly indicates that there

are different linear angles of population growth. This was addressed by replacing the information for

each period, 1996 – 2009, with known figures and thus a diversion will only take place after that period.

41

6.3 Population Characteristics Modelling

Given that an overall population picture has been obtained it was imperative that various other

estimates be established to improve the usefulness of the modelling.

6.3.1 Household Calculations

Using Census data and ReX and the base population estimates to determine the average number of

people per household in the base period. The trend of people per household was difficult to expand

and thus as a result the 2011 period people per household figure was used as a constant going forward.

Multiplying the Vector with the Matrix of population figures give us 3 different sets of possible household

figures.

0

1,000,000

2,000,000

3,000,000

4,000,000

5,000,000

6,000,000

7,000,000

8,000,000

9,000,000

10,000,000

19

96

19

98

20

00

20

02

20

04

20

06

20

08

20

10

20

12

20

14

20

16

20

18

20

20

20

22

20

24

20

26

20

28

20

30

20

32

20

34

20

36

20

38

20

40

20

42

20

44

20

46

20

48

20

50

20

52

20

54

20

56

20

58

20

60

Peo

ple

Tshwane Population

predicted_Population Low predicted_Population Medium predicted_Population High

42

6.3.2 Age Categories

The following age categories were established for use in the generation of the following tables to minimise the risk of

empty categories.

0 - 4 20 - 24 40 - 44 60 - 64

5 - 9 25 - 29 45 - 49 65 - 69

10 - 14 30 - 34 50 - 54 70 - 74

15 - 19 35 - 39 55 - 59 75 +

6.3.2.1 Ward Breakdowns by male and female

Using Census data and interpolating the missing figures between the three census estimates gives us

a base period of percentage growth in a ward and gender group. This was then extrapolated for the

period 2012 to 2060. The shares of each individual vector moved in a manner that after the

extrapolation all shares added up to 100 percent

Below is an excerpt of the growth share matrices.

79900001 79900002 … 79900105 79900001 79900002 … 79900105

1996 Male 0,5 0,44 … 0,41 Female 0,53 0,49 … 0,44

1997 Male 0,496 0,416 … 0,424 Female 0,524 0,466 … 0,452

2011 Male 0,43 0,34 … 0,46 Female 0,44 0,37 … 0,44

…

…

…

…

… …

…

…

…

… …

2030 Male 0,42 0,32 … 0,47 Female 0,43 0,35 … 0,44

…

…

…

…

… …

…

…

…

… …

2060 Male 0,40 0,30 … 0,48 Female 0,40 0,32 … 0,44

Using these weights alongside the estimated population figures the population per ward for the period

is calculated. The resulting figure was also calibrated equally across all observations per year to ensure

that the population figures match the estimated figures.

6.3.2.2 Pop by Gender, Age and population Group.

Further, using more disaggregated percentages, from census, and 5 year age gaps and following the

same process as outlined in 3.2 the population per age group per gender and race was calculated for

the period 1996 to 2060. This was done using each set of population estimates. This results in three

very large matrices of population figures, 286 columns by 65 rows. Below is a very condensed version

of one of the tables.

43

Year African/Black … African/Black … Coloured … Coloured … Indian/Asian … Indian/Asian … White … White

Male … Female … Male … Female … Male … Female … Male … Female

0 - 4 … 0 - 4 … 0 - 4 … 0 - 4 … 0 - 4 … 0 - 4 … 0 - 4 … 0 - 4

1996 63 652 … 65 471 … 1 628 … 1 628 … 1 265 … 1 083 … 17 271 … 16 543

…

…

…

…

…

…

…

…

…

…

…

…

…

…

…

…

2011 125 124 … 124 480 … 3 214 … 3 214 … 2 249 … 2 249 … 20 905 … 19 941

…

…

…

…

…

…

…

…

…

…

…

…

…

…

…

…

2030 211 865 … 209 774 … 5 445 … 5 445 … 3 752 … 3 837 … 32 301 … 30 778

…

…

…

…

…

…

…

…

…

…

…

…

…

…

…

…

2060 354 223 … 348 187 … 9 112 … 9 112 … 6 130 … 6 488 … 46 151 … 43 884

These tables provide us with population group totals for the City of Tshwane for the period estimated.

The results were also calibrated.

6.3.2.3 Ward population by age group, race group and gender

This is the most disaggregated table and contains population estimates across all 105 wards broken

down in age groups, by race and gender. The resulting matrices have 13 441 columns and 65 rows.

The table is too large to include in the document.

7 WATER DEMAND MODELLING

This section aimed at using collected data from the City of Tshwane on water demand and usage with

an economic and statistic view to determine the needs and supply of water to the city. The analysis

covers the period 2012 – 2055 in line with the fruition date of Tshwane Vision 2055.

This section will cover the approach used and the results of the modelling. The first step was to

determine the future demand of the City. This was obtained from the WRMP and set to a horizon of

2055. Then using existing bulk input data the bulk input between 2014 and 2055 was estimated using

nonlinear interpolation. This forms the baseline comparison for the demand analysis.

In order to calculate the growth in demand the user categories as provided by the Water and Sanitation

department was used, each category was assigned an appropriate forecast technique and the

forecasting was done on the number of occupied records as it directly relates to a business or

household without having to make further assumptions as to land use and specific development criteria.

44

Table 7.1: Forecasting and Rationale

Category Forecasted using Rationale

Businesses GDP Data There is a direct link between business output

and growth and GDP

Clusters Population Data This represents cluster accommodation and thus

the more people there are the more accommodation is needed

Dummy Taken as Given Had a zero value assumed no change

Education Forecasted using internal average

growth

Given that this is a very complex relationship and in the control of Government mostly it

seemed prudent to assume these institutions will keep growing at its existing rate

Farms and Agricultural Holdings GDP Data Assumed that this includes business activities as the large AH and farm areas often lends itself to

that

Flats Population Data Direct link between people and accommodation

Government institutions Forecasted using internal average

growth



Industrial GDP Data There is a direct link between business output

and growth and GDP

Informal Population Data Direct link between people and accommodation

No Treas Forecasted using internal average

growth



Other Forecasted using internal average

growth

The trend was identified and assumed to continue, this was deemed prudent as it would

have required further assumptions

Parks Forecasted using internal average

growth



Relocate Taken as Given Had a zero value assumed no change

RES[ 1] Population Data Direct link between people and accommodation





RES[>2000] Population Data Direct link between people and accommodation

Unknown Forecasted using internal average

growth

The trend was identified and assumed to continue, this was deemed prudent as it would

have required further assumptions

45

Category Forecasted using Rationale

Large Consumers GDP Data

This section does not only include businesses, but other large consumers as well, due to the

complexity of this variable it was assumed as it includes businesses to grow at the business rate

With respect to the population projections the City of Tshwane population model was used to determine

the growth rate of households, these figures were slightly amended to account for service backlogs.

The backlogs were obtained from IHS and forecasted using the existing internal rate. The household

numbers is given below:

Table 7.2: Household figures for the period 2012 - 2055

Year HH Low HH Medium HH High

2012 857 567 963 621 979 158

2015 922 680 1 048 920 1 084 538

2020 1 022 381 1 182 262 1 251 349

2025 1 120 354 1 313 877 1 416 433

2030 1 216 332 1 443 497 1 579 522

2035 1 310 007 1 570 813 1 740 307

2040 1 401 022 1 695 470 1 898 433

2045 1 488 966 1 817 056 2 053 488

2050 1 573 364 1 935 096 2 204 997

2055 1 653 667 2 049 041 2 352 411

The growth rate of each household group was then used and applied to each relevant category resulting

in category forecasts for the population driven records.

The business forecast was obtained by using a GDP model that was estimated using the VAREX

method the results for annual growth percentages were obtained and applied to the business records.

A word of caution is that using this approach will likely show an accelerated view of business demand

growth, but in so doing it provides a more pessimistic view.

The forecast was needed from 2015 onwards as we already had demand data available for the earlier

years. Two scenarios were created one with the model’s projected growth, average of 3.8%) and

another where the city grows at the desired rate of 5.2%.

The combined forecast data then in conjunction with the kilolitre per day, which was obtained from the

City was used. Given the complexity of determining future demand, the most recent observation was

used as a constant for the forecast.

Table 7.3: Kilolitre per day utilized for each respective year

46

Year 2006 2007 2008 2009 2010 2011 2012 2013 2014- 2055

Unit of Measurement kl/day/meter

Businesses 1,760 1,990 1,696 1,721 1,812 1,716 1,714 1,813 1,972

Clusters 1,272 1,149 0,464 0,465 0,452 0,462 0,480 0,502 0,524

Dummy 0,000 0,000 0,000 0,000 0,000 0,000 7,000 0,000 0,000

Education 3,385 3,748 3,383 3,833 3,864 3,709 3,551 3,130 2,709

Farms and Agricultural Holdings 2,531 2,819 2,328 2,203 2,051 1,588 1,164 1,251 1,339

Flats 0,634 0,769 0,440 0,399 0,416 0,397 0,380 0,447 0,515

Government institutions 2,638 2,562 2,142 2,523 2,568 2,714 1,795 1,872 1,950

Industrial 3,004 2,785 2,319 2,263 2,192 2,119 2,199 2,130 2,061

Informal 0,282 0,280 0,354 0,421 0,275 0,435 0,295 0,405 0,515

No Treas 0,000 0,000 0,000 0,000 0,000 1,000 17,000 0,000 0,000

Other 2,243 2,288 1,868 1,958 2,010 1,940 13,046 7,161 1,277

Parks 2,177 3,229 3,309 3,353 2,950 2,869 2,815 2,783 2,751

Relocate 0,000 0,000 0,000 0,000 0,000 0,000 0,000 0,000 0,000

RES[ 1] 0,461 0,480 0,397 0,380 0,390 0,386 0,384 0,449 0,513

RES[ 500] 0,556 0,473 0,407 0,409 0,394 0,391 0,396 0,422 0,448

RES[ 1000] 1,059 1,125 0,890 0,867 0,785 0,779 0,761 0,760 0,759

RES[ 1500] 1,414 1,576 1,250 1,222 1,071 1,080 1,051 1,020 0,990

RES[ 2000] 1,786 1,981 1,565 1,548 1,344 1,340 1,328 1,286 1,244

RES[>2000] 2,127 2,385 1,833 1,829 1,641 1,432 1,552 1,547 1,542

Unknown 1,316 0,872 0,835 0,907 0,883 0,770 0,705 1 0,522

Large Consumers 10,077 9,621 5,138 5,400 7,345 6,452 6,666 5,585 4,504

The information allowed for the creation of 6 forecasting scenarios for water demand. Each of these

forecasts will be discussed in turn. Furthermore, with respect to water demand a 21% water loss

constant as obtained from the water data provided to see the impact of water loss in the demand

modelling.

7.1 Projections with normal GDP

The projections for demand are mapped against the total water requirements as obtained from WRP

consultants. The graph below shows the requirement results not taking into account the water losses.

Given his we see that the low projection never crosses the CoT Scenario as created by WRP, but

crosses the recon study as obtained from the DWA. Both the medium and high scenario’s cross the

WRP scenario in 2052 and 2043 respectively but does not breach the total requirements of the City of

Tshwane as modelled in the WRMP.

47

Figure 7.1: Mapping of CoT scenarios against WRP requirements

When taking into account the water losses we found that for both the low and medium term the water

demand exceeds the original water demand estimates for a portion of the period, but return to positive

by the end of the period. However, this is not the case with the high scenario as it remains in the

negative. This illustrated the importance of reducing water losses. However, by shocking the model

and reducing the water losses to 15%, brings all balances between demands to a positive end by 2055.

Showing the importance of curbing water losses in ensuring that water demand needs are met.

200

300

400

500

600

700

800

900

20

12

20

13

20

14

20

15

20

16

20

17

20

18

20

19

20

20

20

21

20

22

20

23

20

24

20

25

20

26

20

27

20

28

20

29

20

30

20

31

20

32

20

33

20

34

20

35

20

36

20

37

20

38

20

39

20

40

20

41

20

42

20

43

20

44

20

45

20

46

20

47

20

48

20

49

20

50

20

51

20

52

20

53

20

54

20

55

Wat

er

Re

qu

irem

en

ts (

mill

ion

m3

/a)

Total CoT (million m3/a) CoT Scenario Recon Study DD Low DD Medium DD High

48

Figure 7.2: Water Demand balance calculations

7.2 Projections with High GDP

When shocking the model and assuming that the economy will grow at 5.2% per annum, which is an

accelerated growth scenario it shows that the water demand for the economy will exceed the estimated

total requirements in the WRMP. The realistic scenario will breach the top requirements estimate by

2053, whilst the medium and high scenarios will breach the total requirements estimates in 2050 and

2046 respectively. This shows that if we are to shock the model with accelerated growth there is a

possibility that our water requirements given the current usage patterns might not be sufficient, this will

require that water demand management be put in place.

-200000

-150000

-100000

-50000

0

50000

100000

150000

200000

Wat

er D

eman

d B

alan

ce K

l/d

21% Diff Low 21% Diff Med 21% Diff High

-100000

-50000

0

50000

100000

150000

200000

250000

300000

350000

Wat

er

Dem

and

Bal

ance

Kl/

d


49

Figure 7.3: Water requirements mapped High Growth Scenarios

If one adds water losses to the mix. The negative difference between the original demand estimates

and the new estimates increase exponentially. No amount of water loss improvement can cancel out

the negative effect permanently as in all cases the new scenario exceeds the total requirements of the

City of Tshwane. The scenarios presented represent an accelerated high growth, thriving population,

placing extensive pressure on the city of Tshwane resource demand and shows that in such a situation

that if water losses are not kerbed or we reuse water or educate the people about responsible water

use we might face greater, yet avoidable increases in demand.

200

300

400

500

600

700

800

900

1000

1100

20

12

20

13

20

14

20

15

20

16

20

17

20

18

20

19

20

20

20

21

20

22

20

23

20

24

20

25

20

26

20

27

20

28

20

29

20

30

20

31

20

32

20

33

20

34

20

35

20

36

20

37

20

38

20

39

20

40

20

41

20

42

20

43

20

44

20

45

20

46

20

47

20

48

20

49

20

50

20

51

20

52

20

53

20

54

20

55

Wat

er

Re

qu

irem

en

ts (

mill

ion

m3

/a)

Total CoT (million m3/a) CoT Scenario Recon Study DD Low DD Medium DD High

50

Figure 7.4: High Growth Scenario Water Balance results

7.3 Infrastructure

Given the complexities of the resource system it was not possible to remodel the water system,

however, there is no reason to contradict this analysis. Therefore wat we set out to do is to determine

whether the infrastructure expansion given the current share of Rand Water, water supply,

approximately 72%, would be sufficient to meet demand. Taking the information as given in the WRMP

we calculated the water infrastructure capacity of the city. Currently this is approximately 328 million

cubic meters per year. With the expansions planned and excluding the Raw water augmentation plans.

On this infrastructure we mapped all demand scenarios, peak summer demand as well as the maximum

water requirement for the City of Tshwane.

Figure 7.5: Water system and supply mapping 2012 - 2055

-800000

-700000

-600000

-500000

-400000

-300000

-200000

-100000

0

20

06

20

08

20

10

20

12

20

14

20

16

20

18

20

20

20

22

20

24

20

26

20

28

20

30

20

32

20

34

20

36

20

38

20

40

20

42

20

44

20

46

20

48

20

50

20

52

20

54

Wat

er D

eman

d B

alan

ce K

l/d


0

200

400

600

800

1000

1200

1400

1600

2012 2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 2052 2054

Wat

er

Re

qu

ire

me

nts

(m

illio

n m

3/a

)

WTP System Rand Water DD Low

DD Medium DD High Total CoT (million m3/a)

DD Low SPF DD Medium SPF DD High SPF

51

From Figure 6.5 taking into account both water sources given the assumption that 72% of water comes

from Rand water. Mapping on it the realistic economic Growth, population variant scenarios we see

that the planned water infrastructure is more than sufficient to deal with the expected water demand,

even the total water requirements for the City falls under the total supply consistently. When mapping

the Peak summer demand using a factor of 1.5 of demand we find that for the high scenario there are

multiple breaches with respect to the total possible supply and demand. This implies that water demand

management might be needed. The medium scenario also breached the supply levels at 2053. The

implication that this shows that if the population growth accelerates due to possibly urbanisation it could

be possible that down the line we might have to further add capacity to the system to address demand

pressures on the system.

Figure 6.6 below shows a similar graph with the increased growth scenario. Again on the demand side

normally the system is able to meet the demands of residents and businesses. However, in peak

summer time the demand will need to be monitored as it exceeds supply. Significantly especially from

2046 onwards.

Figure 7.6: Water system mapping High Growth scenario 2012 - 2055

0

200

400

600

800

1000

1200

1400

1600

2012 2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 2052 2054

Wat

er

Re

qu

ire

me

nts

(m

illio

n m

3/a

)

WTP System Rand Water DD Low

DD Medium DD High Total CoT (million m3/a)

DD Low SPF DD Medium SPF DD High SPF

52

7.4 Water Losses Costs.

Water losses are an unavoidable occurrence, which as shown above if not adequately managed as

entered in the model as a negative shock could have serious consequences for water supply in the city.

However, hypothetically wat does this mean for revenue if we are able to curb those losses completely,

not very realistic, or partially. This allows for us to see the possible revenue implications of the losses.

The value of the losses are rated at each category’s maximum price 2014/15 Financial year and the

water losses are split between three categories as shown in the table below.

Table 7.4: Key input Variables

Business Government Households

Share 7% 1% 92%

Cost 14,39 14,39 20,91

This then multiplied by the water losses per annum provides us with a breakdown of potential revenue

of these losses in this framework. It can be seen that although the income view of the losses come to

approximately R 11 million in 2055 with an approximate R 3 million saving if water losses are reduced

by 6%, it is a relatively small amount in the grander City of Tshwane operations network, but it has to

be noted that even though it is small ultimately a set of small interventions that improve systems that

save the city money can lead to a significant improvement in city cost and operation efficiencies.

Table 7.5: Water losses results

UAW 21% UAW 15% Saving

2006 R 3 977 674 R 3 977 674 R 0

2015 R 3 849 802 R 2 749 858 R 1 099 943

2020 R 4 403 616 R 3 145 440 R 1 258 176

2030 R 5 813 483 R 4 152 488 R 1 660 995

2040 R 7 626 336 R 5 447 383 R 2 178 953

2050 R 9 842 175 R 7 030 125 R 2 812 050

2055 R 11 101 213 R 7 929 438 R 3 171 775

53

8 RISKS THAT MAY INFLUENCE THE CITY OF TSHWANE WATER RESOURCES MASTER

PLAN (WRMP)

8.1 Population increase

The City of Tshwane has 1900 potential future development areas (FDA)s that will be developed over

45 – 50 year period. These developments will add a net contribution of 1 458 Ml/d to the future annual

average daily demand (AADD). The current demand is 987 Ml/d and the estimated future demand is 2

591 Ml/d. Projected water supply levels can only suffice an AADD of +/- 2% which is a projected

population increase. However, population projections are often lower than the actual increase as

highlighted by the 2011 census that recorded a 3.1% growth over the period 2001 to 2011 when the

projected growth was 2.4%.

• The current population of the City of Tshwane is marginally above 2.9 million. By 2030, 11 million

more South Africans will inhabit cities which means a potential growth for the city. Current in-

migration trends reveal that the bulk of rural-urban migrants in southern Africa will desire to

reside within the Gauteng province.

• The City needs to prepare for the changes that will be imposed by population increase and

assess current capacity into dealing with future changes. Tshwane has been dubbed the fastest

growing

Using the City of Tshwane population Model it is estimated that the population for the period is outlined

below. The estimates aimed to provide variability in the planning of the city to account for various growth

trajectories.

Table 8.1: City of Tshwane population figures, across three scenarios

Low Medium High

2006 2 522 018 2 522 018 2 522 018

2015 3 154 853 3 555 180 3 668 131

2020 3 506 428 4 013 439 4 232 526

2025 3 858 004 4 471 698 4 796 921

2030 4 209 579 4 929 957 5 361 316

2035 4 561 154 5 388 216 5 925 711

2040 4 912 729 5 846 475 6 490 105

2045 5 264 304 6 304 734 7 054 500

2050 5 615 880 6 762 993 7 618 895

2055 5 967 455 7 221 252 8 183 290

54

8.2 Climate change

General agreement exists that Climate Change will be a major uncertainty on the long term planning of

future water resources. Global climate models project drastic changes in the weather patterns that will

result in extreme wet and dry spells.

Projections have revealed massive droughts over the next 100 years as claimed by climatologists

(Douglas, 2006). According to Bear Springs Bossom (2012), climate change will change life on earth

and will have an impact on every nation and the global biology. According to the Global Water Institute

(2013), more than 2.8 billion people in 48 countries are faced with future water scarcity by 2025. The

projections suggest that numbers will reach 7 billion by the middle of this century and water withdrawals

will have increased by 50% mainly in low-income nations as well as countries and regions with absolute

water scarcity.

As lives of the general public improve in terms of sanitation, the stress on the world’s water supplies is

escalating and two thirds of the world population is at risk of being stained by water scarcity (Ibid).

Cities will experience most of the scarcity since half of the world population is currently residing in cities

and 60% of the total population is projected to have been urbanised by 2030.

Projected future water scarcity is a global phenomenon, on a larger scale this has been attested to

climate change and the ongoing environmental degradation as well as the destruction of fresh water

bodies. The major factors that will most likely cause a water scarcity in the city of Tshwane and the

surrounding areas encompass population growth, pollution of water sources and Climate change.

There are projections and models that highlight future water scarcity in the city; this scenario will be a

consequence of numerous factors that encompass nature as well as human actions and population

dynamics.

An increase in population number will have a direct stress on water supply. This is most likely to

compromise sanitation which intern will put pressure on the health system. Anthropogenic factors that

are currently affecting climate are a result of unsustainable planning and energy production (UNDP,

2007). The effects of these actions have an influence of the availability of water in the City of Tshwane.

The rural population is more dependent on the agricultural sector for their livelihood and has limited

resources to cope with and adapt to climate change. Rural to urban migration will then be an alternative

as the village was unable to provide livelihood to the people and the living conditions might be

intolerable.

55

According to a study commissioned in 2015 by The World Resources Institute (WRI) South Africa,

Botswana and Namibia sit squarely within a region that is already vulnerable to climate change. Water

supplies are limited, and risk from floods and droughts is high. Projected temperature increases in

southern Africa are likely to exceed the global average, along with overall drying and increased rainfall

variability. On the water demand side, according to the said study, a 40 to 70 percent—or greater—

increase is expected, further exacerbating the region’s concerns.

Water Stress

Water stress measures total annual water withdrawals (municipal, industrial, and agricultural) expressed as a percentage of the total annual available blue water. Higher values indicate more competition among users. Score Value [0-1) Low (<10%) [1-2) Low to medium (10-20%) [2-3) Medium to high (20-40%) [3-4) High (40-80%) [4-5] Extremely high (>80%) This implies that Botswana and Namibia

used between 10% and 20% water during

2010. The WRI anticipates that this figure

will rise to 40% and 80% in 2040.

South Africa

The WRI predicts that South Africa will

consume 20% to 40% water by 2020,

except for agricultural sector which will be

on a 40% to 80% level.

All Sectors Industrial Domestic Agricultural

2020 2.98 2.57 2.49 3.11

2030 3.04 2.77 2.69 3.14

2040 3.19 2.98 2.90 3.29

http://www.wri.org/sites/default/files/uploads/water_stress_projections_large.jpg

56

8.3 ACID MINE DRAINAGE (AMD)

The process of AMD starts in most cases when mines cease activities or are abandoned, especially

when pumps that were pumping water out to the surface are turned off, the rebound of the water table

can lead to contaminated ground water being discharged (Johnson and Hallberg, 2005). Water draining

from abandoned mines is often highly acidic; posing a major risk to the receiving environment due to

the presence of heavy metals such as iron, aluminum and manganese (ibid). In addition, disposal of

mine waste affects not only the environment on the surface but also the ground water in the long run

through infiltration processes of heavy metals that it contains. This contamination of ground water

through AMD is putting pressure on available fresh water by contaminating it with heavy metals. Bell

and Donnelly (2006) estimate that ground water represent about 90% of fresh water worldwide.

Turton (2009) stressed that South Africa is a water-constrained country, with a long history of mining

industry that use to be the economic drive of the country. Consequently, the mining industry is the

biggest contributor to environmental pollution. To date there are an estimated 6000 abandoned mines

around the country and 400km2 of mine tailings dams (Witwatersrand Goldfield) with the potential to

generate AMD and contaminate the source of water (ibid). Moreover, according to the South African

law, unclaimed mines belong to the government and the government should assume all responsibility,

including for environmental risk linked to mines residue hazard (Lifferink, 2009).

The most prominent effect of mining activities that can negatively affect the quality of land is acid mine

drainage, which is generated through mining activities and after mine closure. AMD formation takes

longer to generate but its effects can persist even after mines have been shut down. AMD effects results

in the pollution of the environment including water and soil.

57

Figure 8.1: AMD formed owing to interaction between water and mine residue on the surface tailing reclamation operation in the Western Basin. This site drains into pit operations that are directly connected to mine void

Source: Report to the inter-ministerial committee on AMD, 2010

Figure 8.2: acid drainage flowing into streams near Krugersdorp Source: Rachel, 2011

AMD mismanagement will result in the spills of acidic water into some of the country’s water systems,

such as the Vaal and Limpopo rivers that are important water resources for the country. Moreover,

excessive stress is put on the country’s economy and water-scarce environment, with the potential to

weaken the agricultural and industrial sectors.

According to the Development Bank of Southern Africa (2009), South Africa’s water resources are

limited and support a dynamic growing economy and provision of services; and so there are areas in

which water is already, or could soon become, a constraint to economic and social development due

to acid mine drainage that is costly to manage. Therefore, it is of great importance that water resources

that are in the vicinity of the mining areas be protected against the threat posed by AMD. The challenges

of acid mine drainage management and the danger of its effects on the environment and human health

is a major concern for governments, mining companies, scientists, ecologists and environmentalists to

mention but a few.

The United States Environmental Protection Agency (USEPA) stresses the effects of AMD on the

environment can be rated as second to global warming (Mandres et al.2009). It is thus a matter of

urgency that the government, in collaboration with all parties concerned with and interested in the issue

of AMD, establishes a working partnership for the protection of the environment against the negative

58

effects of AMD around Gauteng region and the watershed. In this regard, the National Environmental

Management Act 107 of 1998 (NEMA) states that the government must prevent pollution and the

National Water Act 36 of 2008 regulates the protection of water resources. Also, the Department of

Water Affairs and Forestry (DWAF) proposes a holistic system approach to prevent the pollution of the

environment by recovering the cost of water quality management from the polluter (Sakoane, 2005).

The greatest challenge pertaining to AMD is the fact that no one claims ownership of some of the (gold)

mines that are responsible for generating AMD in the water-shed, (thus there is no one to account for

the pollution). Therefore, the current situation of AMD requires that all spheres of government unite

their forces and work together and/or in conjunction with each other through ‘cooperative governance’

as stipulated by the South African Constitution for a better management of the situation. It is unfortunate

that government institutions are working in silos (Tracy et al. 2009). This means that there is no co-

ordination of government intervention as well as effective institutional co-operation regarding the

management of environmental disasters caused by mining activities. Whereas a successful AMD

management will requires an efficient intergovernmental action to prevent negative impacts of AMD on

the environment, social-economic and human health.

8.3.1 AMD Generation

The oxidation of sulphur-rich mine wastes interactions with water and oxygen results in the release of

Acid Mining Drainage. When the ore containing high amounts of pyrite and other sulphide minerals is

processed, the rock and tailings produced on the surface are exposed to water and oxygen. Adding to

that, the excavation practice exposes sulphide in the walls of opencast and underground operations,

and perturbs the host rock and hydrological regime around mined-out areas, allowing the entrance of

water and oxygen, (Report to the inter-ministerial committee on Acid Mine drainage, 2010).

Some environmental impacts relating to AMD according to Jennings et al, (2008) encompass:

1. Soil pollution from Mining waste: A 1991 CSIR report estimated that mining and discards (mines

dumps or tailings) constitutes about 75% of the total waste generated within the country. However, at

the writing of this report, there were no new data available to measure whether the situation has

improved or not. The MPRDA Act 28 of 2002 defined tailing as a “product derived from or incidental to

a mining operation and which is stockpiled, stored or accumulated for potential re-use, or which is

disposed of, by the holder of mining rights, mining permit or production rights”. The disposal of mine

dumps or tailing must be done in accordance with the environmental law set forth. However, some

59

tailing in the West Rand (which falls within the catchment basin) are disposed without following the

necessary environmental protection as regulated by the law (personal observation during site visit).

2. Atmospheric pollution from contaminated tailings: the current regulation emphasises the fact

that, uncovered mine dumps presents a danger to the air and atmosphere. Therefore, it requires that

all mine dumps around the country be covered to prevent airborne pollution and to control dust and

fumes released into the atmosphere. Again, the covering of mine dumps in and around the country and

in the West Rand Goldfield is not applied according to the law.

3. Water Pollution from AMD: the presence of metals in mine dumps presents a great danger to the

environment and human health as it can contaminate ground water by seeping underground which is

an important source of potable water in South Africa. AMD is characterized by high levels of heavy

metals in water, as such it is responsible for physical, chemical and biological degradation of water

streams habitat (Jennings et al. 2008). This means that once AMD is formed and decanted into the

environment, it becomes available to the biological organism of the receiving environment. Therefore,

when it reaches streams of water, fish and other animals are directly exposed to different metals

contained in AMD and H+ ions and this may weaken their respiration and acute toxicity may result after

consumption (ibid.).

According to Coetzee et al. (2006), the impact of mining activities, in particular gold mining, on South

African water resources can be divided into two major groups: - The impact on the availability of water

in the areas (qualitative aspect); - The impact on the quality of water. South Africa is a water scarce

country; therefore efficient and sustainable planning is required to protect the country’s water

resources; such as streams of water and ground water (aquifer). The relevance of this statement is that

AMD contains a number of chemicals that can cause acidity, high concentration of dissolvent metals

and deposition of precipitated metals oxides like iron hydroxide that put too much stress on receiving

aquatic ecosystem (Mcnight and Feder 1984; Kelly 1988 cited by Niyogi et al. 2001: 506).

Based on the aforementioned effects of AMD, it is clear that there are many threats to water resources

such as groundwater and streams of water in the West Rand due to effects of AMD. The consequences

of this situation will be affecting not just West Rand District Municipality but the entire Gauteng region

by contaminating the streams and ground water if government institutions fail to take responsible and

effective actions to deal with the issue of AMD in the West Rand goldfield area. Therefore, cooperative

planning will be appropriate because it will help all government departments to exchange their views

and found a common solution to the problem of AMD.

60

However, AMD is not the only thing that stresses groundwater and water resources in South Africa, as

noted by Ryan et al. (2010) that urbanization, agricultural and industrial activities continues to affect

groundwater negatively as well. More so, Witthuser and Holland (2008), argue that urban expansion,

industrial development, and mining activities have put too much pressure on the dolomitic aquifers

around Gauteng and the country as a whole.

8.4 Ground and surface water pollution

Ground water is a source of drinking water for many communities of the City of Tshwane and the

country as a whole. Ground water is prone to chemical pollutants that can enter aquifers through various

processes such as seepage, infiltration and percolation depending on the permeability and porosity of

the rock. Impervious surfaces can protect ground water depending on the depth of the water table.

Ground water contamination is almost always a consequence of anthropogenic activities. Ground water

vulnerability is directly related to population density and intensive land use. Virtually any activity that

leads to the release of waste or chemicals to the environment has a potential of leading to ground water

pollution. It is difficult and expensive to purify contaminated ground water.

Pollution prevention and remediation entails understanding the interrelation between surface and

ground water. Surface and ground water are interconnected and can be fully understood and

intelligently managed only when that fact is acknowledged. Water supply wells and boreholes near a

source of contamination are bound to be contaminated at a rate that is dependent on the underlying

rock porosity; any water body or river close by will also runs a risk of being polluted by the ground water.

A contaminant may diffuse within an aquifer in the same manner as ground water movement; however,

this is highly dependent on its physical, chemical, and biological properties. Due to their physical and

chemical properties, some contaminants do not follow ground water flow which is determined by the

nature of terrain. To some degree, it is possible to predict the flow of those substances that move along

with ground water within an aquifer, these flow in the direction of the topography from recharge to

discharge areas. On this account, potential contamination sources within the drainage basins that

supply the City of Tshwane need to be closely monitored.

8.4.1 SOURCES OF GROUND WATER POLLUTION

Natural Sources

Found naturally in rocks or soils: - iron, manganese, arsenic, chlorides, fluorides, sulfates

or radionuclides.

61

Septic Systems

One of the main causes of ground water contamination in urban areas is effluent spillage

due to aging infrastructure and poor waste management. Sources of this form of

contamination are: - septic tanks, cesspools, and privies.

Improper Disposal of Hazardous Waste

Hazardous waste should always be disposed of in a safe manner so as to avoid getting

it washed downstream by overland flow or getting in contact with ground water.

Hazardous waste should be disposed of by a licensed handler or through municipality

induced program of handling hazardous waste.

Landfills

There are thousands of solid waste landfills throughout the City of Tshwane and the

catchment area that feeds into its water supply. It is also common for chemicals that

should be disposed of in hazardous waste landfills to end up in municipal landfills. Once

in the landfill, chemicals can leach into the ground water.

Sewers and Other Pipelines

Poor maintenance of infrastructure such as sewerage disposal can also lead to ground

water contamination. Sewage consists of organic matter, inorganic salts, heavy metals,

bacteria, viruses, and nitrogen. Old pipelines carrying industrial chemicals are also known

to leak from time to time especially when materials transported are corrosive.

Other sources encompass pesticide and fertilizer use as well as drainage wells. Effects of ground water

contamination include Contamination of ground water can result in poor drinking water quality, loss of

water supply, degraded surface water systems, high cleanup costs, high costs for alternative water

supplies, and/or potential health problems.

8.5 SPRINGS AND BOREHOLES

The City obtains a significant portion of their water supply from boreholes and springs, which is blended

with Rand Water and water from Rietvlei Dam within the bulk distribution system. The two springs at

Fountain Valley produce approximately 37.5 Ml/day. The five boreholes at Kentron, Valhalla and

Erasmia produce an additional 10 Ml/d.

62

Table 8.2: CoT Springs and Borehole Water Sources

WATER SOURCES COMPARTMENTS

West fountain valley spring, ZP13 & ZP16 West Fountain sub-compartment

East fountain valley spring East fountain sub-compartment

Valhalla and Erasmia boreholes Erasmia compartment

Kentron borehole West Doornkloof sub-compartment

The identification and management of protection zones around water sources is important for the

protection of water quantity and quality. Protection is based on controlling land use within existing

current capture zones, and on controlling activities that may alter the size and shape of capture zones.

8.5.1 Capture Zones

The identification of capture zones entails zoning areas around boreholes and springs from which they

collect water. Capture zones are designed to protect water sources by controlling land uses and

pollution generating activities within a specific distance of the source from which groundwater is

collected within a certain number of days. Hence capture zones can be defined as being the area from

63

which water reaches a source within 30, 50 or 100 day travel times etc. The extent of these zones is

therefore controlled by abstraction rates, transmissivity and hydraulic gradient, and geological

conditions.

8.5.2 Aquifer Protection based on Capture Zones

Protecting the aquifer against impacts from contaminants of a more persistent nature that may enter

the capture zone of a water source, or that may exceed the attenuation capacity of the aquifer.

These can include contaminants originating from waste sites, fuel spills, mine wastes etc. Management

of these activities requires regulating the total load of contaminants that enter the capture zones of

water sources. This requires that long term steady state capture zones be identified.

In general, businesses that engage in any high risk land use practices should be avoided in this

zone. These include:

Disposal of solid or hazardous waste.

Animal feed lots.

The outside storage of herbicides, pesticides, fertilizers, or fungicides.

Industrial uses which discharge processed waters on site.

Chemical or bacteriological laboratories.

Metal polishing, finishing, and plating establishments.

Commercial wood finishing, preserving, painting, and furniture stripping establishments.

Wastes from commercial printing, photocopying and photographic processing establishments.

Motor vehicle service and repair shops, scrapyards, motor vehicle salvage operations, car

washes, as well as any similar use which might potentially effect groundwater quality.

Trucking and bus terminals.

Leather tanning and finishing.

Electrical components manufacturing or assembly.

New installation of underground storage tanks of liquid petroleum and/or products of any kind.

Storage of liquid petroleum products except for storage in a freestanding container within a

building.

Storage of petroleum, and/or any other regulated substances in underground storage tanks.

Any other use, which involves, as principle activity, the manufacture, storage, use, treatment,

transportation, or disposal or toxic or hazardous material.

9 Age and Maintenance of existing infrastructure

Infrastructure Asset Management can be defined as an integrated process of decision-making, planning

and control over the acquisition, use, safeguarding and disposal of assets to maximise their service

delivery potential and benefits, and to minimise their related risks and costs over their entire life. It thus

64

includes operation of infrastructure assets, and also planned maintenance and repair, refurbishment

and renewal, and provision for replacement of the infrastructure.

Asset replacement is one of the critical success factors identified in the City’s Water Conservation /

Water Demand Management Strategy to reduce its unaccounted for water. Older pipes are more

vulnerable to bursting, which has shown an increasing trend over the last years, which are placing a

large challenge on maintenance resources and inconvenience to consumers.

Network water leaks

Year 2011/12 2012/13 2013/14 January to

August 2015

Number 40,243 44,084 51,908 47 719

Annual Increase (%) + 9.5% + 17.7%

The average monthly leaks reported in 2015 for the whole of Tshwane is as follows:

Average nr of leaks

per month

Response within

48 hours 48 hours 3-7 days 8-30 days 3-30 days

3977 43.4% 1726 1087 1164 2251

(Due to problems with IT connectivity the statistics for Region 2, 5 and 7 are incomplete/not included)

The number of water network leaks has increased by 29% over the last 2 years, due to many pipes

reaching the end of their economic useful life. A pipeline replacement model was developed to

scientifically prioritise pipeline replacement. In order to delete the backlog in replacement a 5 year

program are proposed which will require R80m and R40m annually for the water and sewer system

replacements respectively.

Due to the deduction of the Water and Sanitation 2014/15 Capex by approximately 50% from the

previous year, backlog and extension programs had to be rescheduled. The water replacement

program is hereby highlighted as an important service delivery program and should be prioritised should

funding becomes available.

65

9.1 Leak repair reaction time

Year % Leak Repair in 48

hours

2010 78%

2011 74%

2012 73%

2013 65%

2014 48%

2015 38%

Tshwane’s leak repair reaction time shown a constant decline since 2010. Time to repair leakages

almost doubled over the last 5 years.

Reasons for decreased performance in leak repair activities include lack of resources: personnel,

budget, vehicles, materials, equipment, tools, wrong routing of complaints (customer care issue). A

major contributing factor for the constant decline since 2010 is the regionalisation model. The location

of the depots increased travel time of maintenance teams, inadequate stock levels due to Supply Chain

Management policies as well as the composition of ineffective response teams.

9.1.1 Water pipe system results

The Water and Sanitation Division’s IMQS (Integrated Management Query Station) reported that the suburbs identified on map below have a zero year remaining life. These areas reported the most water leakages and the systems have to be replaced immediately.

66

Figure 9.1: Systems age analysis

In summary, the replacement cost required to address the water leakages is tabled per region.

Table 9.1: Summary of replacement cost by regions

Region Replacement Cost

Region 1 R150 Mil

Region 2 R50 Mil

Region 3 R799 Mil

Region 4 R246 Mil

Region 5 R28 Mil

Region 6 R304 Mil

Region 7 R0 Mil

Water and Sanitation Department require R316 million per year over 5 years to finance the

replacement. The budget allocated for 2014/2015 for the department is R15 million.

10 Nonpayment of services

The nonpayment of services have drastic consequences on current and future service delivery

programmes within the City of Tshwane. A case in point is the current financial position of Eskom.

67

Eskom says that debt collection from municipalities and small power users remains a concern, with

payment in Gauteng a particular concern. The total municipal arrears debt “remains high” at R4-billion

as at 30 September 2014.

According to electricity expert Chris Yelland, the elimination of non-technical electricity losses in South

Africa – i.e. electricity theft and non-payment – would avoid the need for the Stage 1 (1000MW), Stage

2 (2000MW) and Stage 3 (3000MW) load shedding. (http://businesstech.co.za/news/general/76008/eskoms-massive-

soweto-headache/)

The same principle applies to Tshwane. If the non-payment of water can be reduced, the City will save

a significant amount of water as well as income. It is estimated that the level of nonrefundable water

within the City of Tshwane during 2014/2015 was 24.7%. The City can expect that the current level of

nonpayment will increase in the same ratio as the increase in population numbers. This calls for robust

action measures to ensure that Tshwane Vision 2055, Strategic Objective 1: Provide sustainable

services infrastructure and human settlement management will be achieved.

Figure 10.1: Non Revenue Water for the period 2007 - 2012

According to the Water and Sanitation Department the City lost approximately 83 900 000 kiloliters of

water per year at an annual cost of R461 million. Ironically, the said department require R316 million

http://businesstech.co.za/news/general/76008/eskoms-massive-soweto-headache/

http://businesstech.co.za/news/general/76008/eskoms-massive-soweto-headache/

68

per year to replace outdated infrastructure. By investing in the replacement of dated infrastructure the

City will significantly save money.

11 Reporting of water leakages.

Often there is a long lead time between the reporting of a water leak and the repair of a water leak.

This shows that there are intrinsic inefficiencies in the system. These inefficiencies are costing the city

money and affects resident satisfaction. In the figure below we see that the average lead time for the

repaid of leaks has steadily been taking longer and longer. Furthermore from Figure 10.2 below it can

also be seen that the average growth rate in complaints for the period 2012 – 2015 is 10.9%, but we

also observe that less than 50% of water leaks are attended to within 48 hours.

Figure 11.1: City wide leakages repair times

Figure 11.2: Average leaks per month for the period 2012 – 2015

69

Figure 11.3: Technical staff per 100 000 people

From Figure 10.3 above we see that the City faces a severe shortage of technical staff in the maintenance of water

systems, this could be viewed as a major constraint in being able to effectively provide service delivery to the residents

of the City of Tshwane.

12 Conclusion

The world’s demand for water is likely to surge in the next few decades. Rapidly growing populations

will drive increased consumption by people, farms and companies. More people will move to cities,

further straining supplies. An emerging middle class could clamor for more water-intensive food

production and electricity generation.

Whatever the drivers, extremely high water stress creates an environment in which companies, farms

and residents are highly dependent on limited amounts of water and vulnerable to the slightest change

in supply. Such situations severely threaten national water security and economic growth. National and

local governments must bring forward strong national climate action plans and support a strong

international climate agreement in Paris this November. Governments must also respond with

management and conservation practices that will help protect essential sustainable water resources

for years to come.

Water supplies can't be drought-proofed, but the vulnerability of the supplies to drought can be reduced.

There's a lot of different options. You can build new dams, you can enlarge existing dams, you can

build desalination plants, you can introduce water sharing strategies, you can transfer water from – you

know, via pipes and so on, you can introduce demand management - so restrictions and so on,

recycling water. There's a whole heap of different water saving or water augmentation options that you

can do.

70

What we need to do the next step is, okay given we now we have a better understanding of what the

risk of drought occurring is, what are the cost and benefits of those various options and particularly,

which combinations of those options.

13 Recommendations

Infrastructure developments are the arteries that connect a city to its peripheral neighbours. Anticipating the needs of a city for the future is the key to becoming a city. Planning is necessary to ensure delivery of a sound economic and social infrastructure to ensure liveability, quality of life and vibrancy in the city. Poor infrastructure maintenance and lack of infrastructure development to counter the increasing water

demand are some of the major courses of these leakages and shortages, however, illegal connection

are also regarded as losses since this water is unaccounted for. South Africa is currently sitting at 37%

of unaccounted water losses.

The city needs to look at its staff compliments and age analysis to effectively curb water losses through

the systematic replacement of aging infrastructure.

Further analysis needs to be conducted into the implications and cost of water losses and how the city

can effectively curb water losses.

The city needs to look into alternative methods such as water harvesting and water reuse as well as

determine the characteristics of each user category in the city to more effectively determine the per

user behaviour trend and the elasticity of those users to methods such as demand management.

Scenario Modelling needs to be done to take into account the possible shocks in the system such as

acid mine drainage and climate change.

The identification and management of protection zones around water sources (fountains and

boreholes) is important for the protection of water quantity and quality. Protection is based on controlling

land use within existing current capture zones, and on controlling activities that may alter the size and

shape of capture zones. Service delivery departments have to be mindful of the impact of new

developments on protection zones when commenting on land use applications.

City of Tshwane: Draft Water Scarcity Report · 3 DEFINING WATER SCARCITY The Food and Agriculture Organisation of the United Nations defines water scarcity when demand for freshwater

Documents