50 Years of Direct Potable Reuse

The Windhoek Experience

Content

1. Context: Namibia and Windhoek

2. Early Water Supply to Windhoek

3. Reclamation in the Early Years (1968 – 1998)

4. Current Reclamation Practice (1999 to present)

5. Future for Windhoek

6. Conclusion

Namibia and WindhoekOrientation and Context

Country of 2 Deserts

Country of 2 Deserts

Surface Water – Southern Africa

Surface Water – Namibia

Windhoek:

1650m AMSL

Annual Rainfall – 360mm

Evaporation – 3400mm/a

80

0km

Water Supply to WindhoekThe Road to Direct Potable Reclamation

Water Supply History1800’

s

• Early settlement – hot natural springs. Names: /Ai//gams ("Fire water" - Nama) Otjomuise (“Place of steam" - Herero)

1911 • Development of well field

1933• Avis Dam – 2.4 Mm3

1957 • Water Crisis - Well field over abstracted by more than 50%

1959• Goreangab Dam – 3.6 Mm3

1960• Wellfield over abstracted by 100%

1968• Potable reuse implemented

Future

Period preceding Potable Reuse (1958 – 1961)

1957 Water Crisis

• Water rationing practiced during drought period – peak demand

Executive Commission Appointed

• Established by the Hon. Admin. General of South West Africa on 7 February 1958

• Members from Govt. Dept. of Water Affairs and City of Windhoek (Water Supply)

Commission Terms of Reference

• Determination of Windhoek Demand for foreseeable future

• Review of available sources and safe yield

• Investigate additional resources and report on possible solutions

• Report on Financial Implications of proposed solutions

Demand Projections

Actual Population Increase (1939-1957)

Period Pop. Inc. (%)

1939 - 1945 3.6 %

1945 - 1951 8.8 %

1951 – 1957 10.5 %

Post WW II“Population explosion”

Projected Demand Growth Rates (1958)

Period Estimated Growth (%)

1958 - 1960 12.5 %

1961 - 1962 10 %

1963 – 1969 8.5 %

1970 - onwards 7 % (on 1969)Review of Windhoek’s Water Supplies – September 1960, Annex. B, pg. 67

Available Water Sources (1960)Avis Dam (2.4 Mm3)

• Average abstraction attained over initial lifespan (1934 to 1959) – 0.375 Mm3/a

• Empty 4 out of 10 years on average (current status)

• “For all practical purposes therefore, Avis Dam can at 95% probability of a certain yield, provide nothing.”

Goreangab Dam (3.6 Mm3)

• New (1959) – Expected yield of 0.6 Mm3/a at 95% reliability

• Considered together with Avis dam to be able to provide 0.8 - 1.0 Mm3/a at 75% reliability

• Dam constructed downstream of the city (15km away) – prone to pollution (today not used for potable)

Windhoek Aquifer

• Safe Yield established at 1.6 Mm3/a (3.1 Mm3/a)

• Draw-off in 1960 – 3.22 Mm3

• Continued over-abstraction of the aquifer a serious concern (Executive Commission Report, 1958)

Safe Yield (All Sources) - 2.6 Mm3/a vs Demand (1960) – 3.8 Mm3, with High Growth Projections

Water Supply in 1960

35 000

Population

3.8 Mm3

Consumption

Avis Dam - Empty

Groundwater – 3.2 Mm3 (SY – 1.6 Mm3/a)

Goreangab Dam – 0.6 Mm3

Proposed Solutions (1958 Executive Commission)

• Implemented with Immediate Effect

• Large Scale Pipe replacement and increased meterage resulted in reduction of losses from 20% to 7.5% by 1961

Water Loss Control

• Not considered economically viableDiversion of Michaelis River to Avis

Dam

• Not considered economically viableDiversion of overflow from Hofnung

Dam to Avis Dam

• Confluence of Gammams, Aretaragas & Augeigas River – 2km downstream of Goreangab Dam

• Calculated safe yield used in conjunction with existing Goreangab Dam -1.6 Mm3/a

• More expensive than Swakop Dam Scheme per m3 of waterGammams III Dam Scheme

• Estimated safe yield of 7.14 Mm3/a (5.14 Mm3/a to Windhoek) – 1st State Supply Scheme to WHK

• Benefit of Supply to Okahandja (2.0 Mm3/a)

• Construction & Operational Cost – Distance of 70km and Elevation of 250mSwakop Dam Scheme

• Supply potential of 33.3% of all water produced

• CoW - Technically viable; Administration not convinced

• Most economical operated in conjunction with exiting Goreangab Scheme

Purified Sewage Effluent Reuse (DPR)

“It is obvious that the present sources cannot cope with current and future requirements, and as will be seen, the boreholes arealready sorely overtaxed”- 1958 Executive Commission Findings

DPR as the most Feasible Option“It can be readily seen that the reuse of water from the Sewage Works is the key to the whole report,…Every effort should be made to ensure the supply from this source”- Review of Windhoek’s Water Supplies, Town Engineer, Sep. 1960

1958 Executive Commission

• Implement Water Loss Control

• Build Gammams III Dam Scheme (already at planning and design phase) – to be completed by 1962

• Plan and Build Swakop Dam Scheme – to be completed by 1965

CoW Review (1960)

• Re-use most economical

• Re-use sufficient until 1967

• Swakop Dam Scheme should be constructed prior to Gammams III Dam Scheme for economic reasons

• Pushed for higher yield from Aquifer

DPR (1968 to 1998)Establishment and early days

17

Establishment of DPR in Windhoek• Direct Potable Reclamation considered as early as 1954

• Initially planned to commence in 1961 – delays in implementation

• Early worked based on a case study in Chanute, Kansas, USA• Flow of Neosho river virtually ceased in 1956 after 5 yrs of drought• River dammed below town’s sewage outfall and water reused as potable for

5 months

• CoW teamed up with the Council for Scientific and Industrial Research (CSIR) in South Africa – Contract renewed in 1961 – More focussed approach

• Extensive pilot study conducted from 1962 to 1965

• Led to design of the full scale plant for 4 800 m3/d (1.6 Mm3/a)

• Construction started in 1967 and completed in 1968

• Production starts on 18 October 1968 just in time to avert another dry spell producing 15 - 25% of potable supply within first year

• Swakop Dam Scheme (S von Bach Dam) completed in 1970, WTP in 1971

• Gammams III Scheme was never built

Early Days of Windhoek DPR• Following implementation of Windhoek DPR – Stander Plant constructed in Pretoria, RSA

in 1970 as research and demonstration facility

• Continued research results in early upgrades to Windhoek plant until 1980s

• During the initial period monitoring done by 4 different independent laboratories

• Major improvement with introduction of ASP (1979) over biofilter (humus tank) effluent used previously – nutrient removal

• Biomonitoring using fish introduced in 1982

Public Acceptance

• Water shortage and lack of alternatives helped push through initial decision

• Next major drought (82/83) – residents criticize City Council for unused effluent downstream of reclamation facility

• Led to Council Decision (CR 432/8/84) for capacity upgrade to existing DPR from initial 4 800 m3/d to 7 600 m3/d

Windhoek DPR - History Upgrade Wastewater treatment (#1) Maturation ponds (Retention time) Upgrade Windhoek Direct Potable Reclamation (#2) Capacity

1962-1979 Biological filters PI 14-10 days 1968-1976 Goreangab Mark I 4.8 Ml/d

1976-1980 Goreangab Mark II 4.8 Ml/d

1979-1994 Activated sludge Nutrient Removal PII 10-6 days 1980-1986 Goreangab Mark III 4.8 Ml/d

1986-1994 Goreangab Mark IV 7.2 Ml/d

1994-2016 Activated sludge Nutrient Removal PIII 6-3 days 1994-2001 Goreangab Mark V 14.4 Ml/d

2002-2016 Goreangab Mark VI 21 Ml/d

#1 PI: Primary settling - Biofilters - Secondary settling - Maturation ponds (14 days) #1 PII: Primary settling - Five-stage Bardenpho nutrient removal activated sludge - Secondary settling - Maturation ponds (10 days) #1 PIII: Primary settling - UCT or modified Johannesburg nutrient removal activated sludge - Secondary settling - Maturation ponds (6 days)

#2 Mark I: [Carbon dioxide], [Alum], Algae Flotation, Foam Fractionation, [Alum + Lime], [Breakpoint chlorination], Settling, Rapid sand filtration, Activated carbon, [Chlorine], Blending

#2 Mark II: [Lime], Settling, Ammonia stripping, [Carbon dioxide], [Chlorine], [Alum + Lime], Settling, [Carbon dioxide], Rapid sand filtration, [Breakpoint chlorination], Activated carbon, [Chlorine], Blending

#2 Mark III: [Chlorine], [Alum + Lime], Settling, [Breakpoint chlorination], [Alum + Lime], Settling, Rapid sand filtration, [Chlorine], Activated carbon, [Chlorine], Blending #2 Mark IV: [Alum + Lime], Dissolved air flotation, [Chlorine], [Alum + Lime], Settling, Rapid sand filtration, [Breakpoint chlorination], Activated carbon, [Chlorine], Blending #2 Mark Va: [Ferric], Dissolved air flotation, Rapid sand filtration, Activated carbon, [Break point chlorination], [Stabilisation: Lime], [Chlorine], Blending #2 Mark Vb: [Ferric], Dissolved air flotation, RSF(ftw) - RSF(ftw)+GAC, Activated carbon, [Break point chlorination], [Stabilisation: NaOH], [Chlorine], Blending #2 Mark VI: PAC, [Pre Ozonation], [Ferric], Dissolved air flotation, [MnO4 + NaOH], Rapid sand filtration (ftw), [Ozonation, H2O2], BAC+GAC, Ultra filtration, [Chlorine], [Stabilisation: NaOH], Blending

DPR in the Region

Evidence of no viable alternative

Only domestic effluent

Design based on adequate research and testing

Minimum of 48h storage before distribution

Max. 20% in final water blend – to be kept as constant as possible

Epidemiological studies should be done before and after introduction of DPR

Intensive public campaigns before introduction

“In view of the limited amount of water available and the growing demand, water conservation and reuse are obvious necessities in the South African Water Management. It is widely foreseen that reclamation of water from effluents and the direct introduction into domestic supply will become inevitable in the next decade or two.”- IAWPR Conference, 1982

Proposed Directives for DPR (1982)

Current Windhoek DPR Technology and Practice

Research, Development & Implementation

Pilot Testing – 400 days of data over 3 years

Primary deviations: Micro/Ultra Filtration &

Ozone + GAC

Establish Treatment objectives based on 6

International WQ Standards in addition to research & experience

Pioneering a quantitative framework for the Multiple Barrier

Concept

Design included full automation of the plant

Plant constructed from 1999 to 2001 – Testing

and inauguration in 2002

Process Train – changes Mark V to Mark VI

Process Train – changes Mark V to Mark VI

Goreangab Water Reclamation Plant

Principles of Windhoek DPR• Segregation of all industrial effluent in a separate drainage area and dedicated

treatment

Only WW of Domestic

origin

• Monitoring from the point of discharge up to the final consumer off-take –Compliance across unit treatment processes

Whole Cycle Monitoring

• Efficient removal of biodegradable constituents, including nutrients during upstream WWT process

Effective Biological Treatment

• Multiple unit treatment processes aimed at removal of targeted contaminants

Multiple Barrier Treatment

• Dilution of final effluent with water from natural sources in 33:66 ratio

Blending of final effluent

• Continuous efforts to educate the public and to keep communication open; thereby maintaining public confidence

Education and Communication

Multiple Barrier Concept

• 40% - 60% Removal Efficiency

Partial Barrier

• 99.9% Removal Efficiency

Complete Barrier

Future for WindhoekImportance and role of DPR in future Water Supply

Augmentation Alternatives

CoW Water Supply Strategy (2015)

Optimise

• WDM – Efficient use

• Water Loss Reduction

Reuse

• Potable

• Non-Potable

Sustainability

• Sources

• Climate Change

• Investment & Operations

ConclusionFinal thoughts

Need as a Driver to DPR

•Driving development of DPR as viable (economic) alternative

•Providing the impetus for implementation and public acceptance – threat of running out of water

Water Shortage 50s/60s

•Directly results in Council approving increased capacity and upgrading of plant (on public demand?)

Drought 1982/83

•Emergency upgrades to the plant to double capacity as mitigation to supply shortage

•Driver to the development and planning of a new and improved DPR facility

Drought 1996/97

•???

Drought 2014 – 2016

Importance of DPR in Windhoek supply

16%

4%

4%

76%

Windhoek Water SupplyBefore 2014-16 Drought

Reuse (Potable) Reuse (Non-potable) Aquifer Surface Water

29%

8%60%

3%

Windhoek Water SupplyDrought Mitigation

Reuse (Potable) Reuse (Non-potable) Aquifer Surface Water

WDM = 25% Reduction

Closing RemarksDirect Potable Reclamation• Is not easy to implement primarily due to Public Acceptance related issues

• Produces drinking water of a high quality

• Is a practical and economical alternative in arid regions to augment potable water supply and has contributed to a stable local economy during severe droughts and water shortages in Windhoek

• Requires continuous monitoring & research to improve systems, influenced by

• Changes in the catchment

• New guidelines and standards released

• Improvements in analytical ability

• Requires ongoing commitment on all levels, political will, comprehensive planning and training for continued success

“Water should not be judged by its history, but by its quality” – Dr Lukas van Vuuren