Provided for non-commercial research and educational use only.
Not for reproduction or distribution or commercial use.
This article was originally published by IWA Publishing. IWA Publishing recognizes
the retention of the right by the author(s) to photocopy or make single electronic
copies of the paper for their own personal use, including for their own classroom use,
or the personal use of colleagues, provided the copies are not offered for sale and
are not distributed in a systematic way outside of their employing institution.
Please note that you are not permitted to post the IWA Publishing PDF version of
your paper on your own website or your institution’s website or repository.
Please direct any queries regarding use or permissions to email@example.com
Direct potable reclamation in Windhoek: a critical review
of the design philosophy of new Goreangab drinking
water reclamation plant
P. du Pisani and J. G. Menge
Direct drinking water reclamation from the Goreangab reclamation plant, has been a reality in
Windhoek, Namibia since 1968. Potable reclamation is a ﬁxed part of the water supply and waste
water has become an indispensible resource for the survival and continued growth of the city. The
multi barrier concepts that were applied 40 years ago have been reﬁned over many years.
Improvements in water treatment technology have made it possible to improve the reliability and the
drinking water quality of the reclamation treatment process. With the latest upgrade, which was
designed 14 years ago and commissioned in 2002, a speciﬁc design philosophy was followed. This
paper will assess whether the objectives of the design philosophy have been met in terms of removal
efﬁciencies and safety of drinking water, which contains at present 25% reclaimed water. The basis
and aims of the multi barrier design that was applied is discussed and with the aid of natural organic
matter (NOM) and microorganism removal, the reliability of the philosophy is tested and compared
with the goals set. Comparisons are drawn between the new plant and the previous plant and how
the new plant is able to accommodate changes in raw water quality. It can be concluded that the
water quality has been improved and the barrier principle does reduce the risk and improve the
P. du Pisani (corresponding author)
J. G. Menge
P O Box 59,
City of Windhoek,
Key words |direct drinking water reclamation, efﬂuent organic matter, multi barrier concept,
Windhoek is the capital city of Namibia. It lies in the central
highlands of the country which is classiﬁed as the most arid
country in sub Saharan Africa. Around 1847 the ﬁrst settle-
ment took place around the hot springs in what is today
Windhoek. As early as 1911, the water table had dropped
to the extent that the springs had stopped ﬂowing and the
ﬁrst borehole was drilled to supply water to the settlement.
The country, then called South West Africa, has no per-
ennial rivers, except on the very southern and very northern
borders, the closest of which is 750 km from Windhoek.
Population growth and agricultural activities put ground-
water reserves under severe stress. A small dam was built
in 1933, but due to the small catchment, erratic rainfall,
on average 360 mm per year and the high evaporation rate
of around 3,400 mm per year, this dam added little to the
water needs of Windhoek.
Reuse of treated sewage efﬂuent for the power station,
was ﬁrst considered in 1954, but not implemented. A
water crisis in 1957 led to the construction of the Goreangab
dam and treatment works on an ephemeral river just west of
the town, which was completed in 1959.
Growing demand in the early 1960s again forced the
consideration of treated sewage efﬂuent as a resource and
after extensive piloting, in 1968, the Goreangab water treat-
ment plant was converted to a two train operation,
respectively treating secondary sewage efﬂuent and dam
water. Both product waters were combined and introduced
directly into the drinking water supply. On 24 November
214 © IWA Publishing 2013 Water Science & Technology: Water Supply |13.2 |2013
1968, the Sunday Tribune declared: ‘Windhoek drinks
sewage water! It’s a puriﬁed world ﬁrst’. The total capacity
of this plant was 4.3 Ml per day which represented between
10 and 12% of maximum daily demand.
The process train employed had been the object of
almost eight years of pilot studies and although the phrase
had not been coined, was based on what is today referred
to as the multiple barrier principle. The process train was
upgraded four times between 1968 and 1992. This plant,
the old Goreangab reclamation plant (Old Plant) also
served as full scale test site for the new Goreangab recla-
mation plant (New Plant), which is the subject of this paper.
NEW GOREANGAB RECLAMATION PLANT
A study conducted in 1991 (Haarhoff ) into the treat-
ment capacity of the Old Plant concluded that it could be
extended to 14.4 Ml per day with minor changes. In 1992
a decision was taken to increase the capacity to 21 Ml per
day (Burmeister et al.;Haarhoff & van der Merwe
The New Plant would equally use secondary treated sewage
efﬂuent and Goreangab dam water as feed. After comparing
27 raw water parameters, consultants FMG Goreangab Joint
Venture in the process train report concluded that, the two
sources were sufﬁciently similar to treat both in the same
Water quality standards
As no speciﬁc guidelines or standards for potable recla-
mation were available, the following standards and
guidelines were considered to develop guidelines for the
•Guidelines for the Evaluation of Drinking water for
Human Consumption (1991) Dept of Water Affairs,
Namibia (Namibian Guidelines ).
•Potable Water Quality Criteria (Rand Water ).
•WHO Drinking Water Guidelines (WHO ).
•The National Drinking Water Standards and Health
Advisories USEPA (USEPA ).
•The European Community Guidelines for the use of
water for human consumption (80/778/EWG) (1980
and 1994 draft) (EC ).
•A guide for the planning, design and implementation
of a water reclamation scheme (Meiring & Partners
The process train report recognized that, in potable rec-
lamation from waste water, microbiological safety would be
the overriding concern. It was however also realized that
trying to set guidelines to guarantee the protection of
people’s health over a lifetime of exposure, would often be
The USEPA approach adopted, which considered either
a maximum contaminant level (MCL) (USEPA ), or a
treatment technique speciﬁcation according to the surface
water treatment rule (USEPA ). For microbiological
contaminants which are not easily measured, such as proto-
zoan cysts, the product of disinfectant residual
concentration and effective contact time in mg.min/l (Ct)
technique was adopted to ensure that the contaminant is
exposed to a certain disinfection concentration for a deﬁned
time to ensure a minimum of 3 log removal for protozoan
cysts and 4 log removal for viruses. The report further
suggested that, by achieving a higher Ct product, higher
log removals could be achieved. The approach was further
based on a guideline for total organic carbon (TOC) from
Water Factory 21, in California, where the TOC of the
injected water is less than 1 mg/l (Williams ) and their
standard was set at less than 2 mg/l based on health risk
model and risk criteria of 1 in 1 million. Breakthrough of
Giardia cysts and Cryptosporidium oocysts during the oper-
ation of the Old Plant between 1994 and 1996 prompted the
immediate upgrade of the Old Plant to include a ﬁlter-to-
waste and automatic ﬁlter backwash system. For the New
Plant, an additional partial and complete safety barrier
was added to include ozonation and ultra-ﬁltration (UF)
(Menge et al.). A partial safety barrier removes between
40 and 60% of a constituent and a complete barrier would
remove more than 99.9%.
Another speciﬁc quality concern identiﬁed, were the
high number of natural organic matter (NOM) (Jacquemet
215 P. du Pisani & J. G. Menge |A critical review of the design philosophy of a drinking water reclamation plant Water Science & Technology: Water Supply |13.2 |2013
et al.;Haarhoff et al.) and anthropogenic organic
compounds found in waste water. Because the treatment of
these would rely on high ozone dosages for oxidation of
organic compounds, coupled with breakpoint disinfection
with chlorine would cause the formation of trihalomethane
(THM), bromate and other byproducts. Bromate and THMs
are accepted as being potentially carcinogenic and of special
interest in the plant design. The removal of organic com-
pounds to minimize the formation of byproducts and to
prolong the lifetime of the activated carbon therefore featured
strongly in the process design (NIWR ;Haarhoff &
Menge ). Other contaminants in waste water that
required special attention were deﬁned as aromatic hydrocar-
bons such as benzene and toluene, as well as phenols and
pesticides. The plant was designed to reduce the organic
load, measured as dissolved organic carbon (DOC) with
enhanced coagulation (EC), dissolved air ﬂotation (DAF)
and rapid sand ﬁltration (RSF) and subsequent ozone, bio-
logical activated carbon (BAC) and activated carbon
adsorption (GAC) (Haarhoff et al.).
Ultimately, the proposed ﬁnal water quality guidelines
were deﬁned as shown in Table 1. This table only contains
the main operational test parameters. Other parameters of
concern such as heavy metals, aromatics or pesticides are
speciﬁed according to the Rand Water (Rand Water )
or USEPA (USEPA ) guidelines.
The multiple barrier concept
At the time of the process train report (WHO ;Haarhoff
et al.), the concept of multiple barriers was more or less
synonymous with margins of safety built into the process
train. If a certain contaminant would manage to pass
through a process speciﬁcally installed for such contami-
nant, there should be further downstream processes that
could eliminate or reduce such contaminant to within the
In the philosophy adopted for the New Plant, three
types of barriers were considered:
•Non treatment barriers, such as diversion of all industrial
•Treatment barriers, which comprise the actual process
steps built into the plant.
•Operational barriers such as powdered activated carbon
(PAC) dosing only used as and when needed.
In the elaboration of the process design, the following water
quality concerns would be addressed: (a) physical and organo-
leptic, (b) microbiological and biological, (c) organics and
disinfection byproducts, (d) macro elements, (e) stability.
Physical and organoleptic
As potable reclamation from sewage has to contend with the
so-called ‘yuk’factor and customer acceptance has to be
maintained 100% of the time, the aesthetic parameters are
extremely important. For these parameters, it was proposed
to provide at least two barriers namely EC, DAF and RSF as
the ﬁrst barrier and UF as the second barrier for suspended
particles. For the removal of colour, taste and odour, one
barrier which is ozone, BAC and GAC. See Table 2 and
Microbiological and biological
The main concerns that had been catered for in the process
design, were viruses and bacteria. Certain bacteria groups
were deﬁned that would serve as indicators. These were
total coliform, faecal coliforms (including Escherichia coli),
faecal streptococci, sulphite-reducing Clostridium and
somatic coliphage. Giardia and Cryptosporidium were also
considered as of major concern in waste water reclamation.
Even though not yet well deﬁned in water guidelines at the
time, four barriers were included for Giardia cysts. For Cryp-
tosporidium oocysts, two complete barriers and one partial
barrier were deﬁned. See Table 2 and Figure 1.
Organics and disinfection byproducts (DBPs)
At the design stage of the plant, DBPs were not yet well
deﬁned in international standards and guidelines. It was
however suspected that these products contributed to
cancer formation in laboratory tests. The anthropogenic
character of the raw water needs special attention. In the
Old Plant only two partial barriers existed, namely coagu-
lation, DAF and RSF followed by carbon adsorption.
In the New Plant four partial barriers would be used,
216 P. du Pisani & J. G. Menge |A critical review of the design philosophy of a drinking water reclamation plant Water Science & Technology: Water Supply |13.2 |2013
pre-ozonation, EC, DAF and RSF and subsequent ozone,
BAC and GAC. See Table 2 and Figure 1.
It was realized that organics subjected to oxidation by
ozone and chlorine, were the cause of DBP formation. During
the pilot phase it was not possible to test for bromate residuals.
Proper instrumentation for the determination of bromate was
not available in Southern Africa at that time and an ion chro-
matograph was only obtained in 2005. The process would
therefore focus on maximal reduction of organics and control
of the ozone dosage to prevent the formation of DBPs. A set
Table 1 |Final water quality
Units Target Maximum
Physical and organoleptic
Calcium carbonate precipitation potential mg/l CaCO
Chemical oxygen demand mg/l 10 15
Colour mg/l Pt 8 10
Dissolved organic carbon mg/l 3 5
Total dissolved solids mg/l 1,000 1,200
Turbidity NTU 0.1 0.2
abs/m 5.0 6.0
Aluminium Al mg/l N/A 0.15
Ammonia N mg/l N/A 0.10
Chloride Cl mg/l Not removed 250
Iron Fe mg/l 0.05 0.1
Manganese Mn mg/l 0.0025 0.005
Nitrate and nitrite N mg/l Not removed 10
Nitrite N mg/l Not removed 0.2
mg/l Not removed 200
Heterotrophic plate count per 1 ml 80 100
Total coliforms per 100 ml N/A 0
Faecal coliforms per 100 ml N/A 0
E. coli per 100 ml N/A 0
Coliphage per 100 ml N/A 0
Enteric viruses CPE per 10 l N/A 0 or 4 log Rem
Faecal streptococci per 100 ml N/A 0
Clostridium spores per 100 ml N/A 0
Clostridium viable cells per 100 ml N/A 0
Total trihalomethanes μg/l 20 40
Chlorophyll aμg/l N/A 1
Giardia per 100 l 0 or 6 log Rem 0 or 5 log Rem
Cryptosporidium per 100 l 0 or 6 log Rem 0 or 5 log Rem
Note: Other parameters will be adhered to as by Rand Water Guidelines (Rand Water 1994).
217 P. du Pisani & J. G. Menge |A critical review of the design philosophy of a drinking water reclamation plant Water Science & Technology: Water Supply |13.2 |2013
of operational guidelines was therefore prescribed to minimize
organics and thereby minimizing DBPs. These were:
•The use of EC, not only to remove turbidity, but to maxi-
mally precipitate organic carbon.
•Moving the chlorination step right to the end of the pro-
cess after maximal organics removal.
•Use of alternative oxidants where possible (permanga-
•Provide activated carbon steps for DBP adsorption.
The process train report recognized nitrogen (nitrate) as a
main inorganic concern. It was however recommended
that this parameter be best dealt with during secondary
sewage treatment through biological nitriﬁcation and
Table 2 |Barriers provided in design for different critical parameters. A comparison between the Old Plant and the New Plant is made. The different treatment steps are abbreviated.
(C: Complete barrier; P: Partial barrier)
Barrier 1 Barrier 2 Barrier 3 Barrier 4
Physical and organoleptic CD/DAF/RSF : C GAC : P
Microbiological: bacteria and viruses BPCL2 : C BPCL2 : C BPCL2 : C
Biological: Giardia, Cryptosporidium CD/DAF/RSF : C BPCL2 : P BPCL2 : P BPCL2 : P
Organics and DBPs CD/DAF/RSF : P GAC : P
Macro elements: Fe, Mn GAC : P
Stability CD (Lime, NaOH) : C
Physical and organoleptic CD/DAF/RSF: C UF: C GAC: P
Microbiological: bacteria and viruses POZ: P OZ: C UF: C BPCl2: C
Biological: Giardia, Cryptosporidium CD/DAF/RSF: C OZ: P UF: C BPCl2: P
Organics and DBPs POZ: P CD/DAF/RSF: P OZ: P BAC-GAC: P
Macro elements: Fe, Mn POZ: P CD/DAF/RSF: P OZ: P BAC-GAC: P
Stability CD (NaOH): C
Where MIX is mixture of dam and treated waste efﬂuent, CD is chemical dosing, POZ is pre-ozonation, DAF is dissolved air ﬂotation, RSF is dual media rapid sand ﬁltration, OZ is ozone contact,
BAC is biological activated carbon, GAC is granular activated carbon, UF is membrane ultra-ﬁltration, CT is contact chamber, PS is high lift pumps Treatment chemicals added: PAC is powder
activated carbon, O
is ozone, Fe is Ferric ion, HCl is hydrochloric acid, Poly is polymer, MnO
is permanganate, H
is peroxide, BPCL
is break point chlorination, NaOH is caustic soda.
Figure 1 |Diagram of New Plant. For a description of symbols see Table 2.
218 P. du Pisani & J. G. Menge |A critical review of the design philosophy of a drinking water reclamation plant Water Science & Technology: Water Supply |13.2 |2013
de-nitriﬁcation. This risk could be controlled through an
operational barrier however, the blending of reclaimed
water with water from natural sources.
Iron and manganese were two other inorganic par-
ameters of special concern. Special measures were
incorporated in the New Plant to reduce their concen-
tration. Four partial barriers would be used to reduce
these, pre-ozonation, EC, DAF and RSF and subsequent
ozone, BAC, GAC. See Table 2 and Figure 1.
The plant design did not provide for total dissolved
solids (TDS) removal, because reverse osmosis (RO) was
not considered as a treatment option at the time. The tech-
nology was regarded as too expensive and there was no
proven track record with the type of water to be treated.
TDS would be reduced by blending the treated water with
natural water which is low in TDS.
Only one barrier, dosing of caustic soda for stability was
added before the ﬁnal disinfection. See Table 2 and Figure 1.
Final process design
The ultimate design of the new Goreangab water recla-
mation plant, relying on the above proposals and which
culminated from a Design and Construct Turnkey tender
procedure, were accepted as shown in Figure 1. For a
comparison between the two plants the diagram of the
Old Plant is provided in Figure 2.
The New Plant was completed in August 2002 and the oper-
ation and maintenance of the plant was contracted out to
WINGOC, a consortium consisting of Veolia Water,
Berlin Water International and VA Tech Wabag. The agree-
ment is based on water quality achieved, with strict penalties
for exceeding any target values and a prohibition on supply
of water exceeding any absolute values.
Quality standards were furthermore deﬁned as intermedi-
ate values to be achieved after every process step, and ﬁnal
water standards to be achieved in the ﬁnal product water (du
Pisani ). For this purpose, a rigorous sampling regime
was prescribed as well as online instrumentation connected
to a SCADA system and refrigerated composite samplers
that monitor water quality throughout the process at critical
control points (CCP). As a result of the above, extensive quality
data are available, which is analysed and reported hereunder,
to critically review the efﬁciency of the multiple barrier system
adopted for New Plant. From about one million test results of
numerous parametersanalysed for the whole Windhoek water
cycle, over a period of 16 years, the following operational par-
ameters are selected to demonstrate the improvement in water
quality between the Old Plant and New Plant. Table 3 shows
Figure 2 |Diagram of Old Plant. For a description of symbols see Table 2.
219 P. du Pisani & J. G. Menge |A critical review of the design philosophy of a drinking water reclamation plant Water Science & Technology: Water Supply |13.2 |2013
the raw water mixture of wastewater treated efﬂuent and dam
water that was treated in the Old and New Plants. Table 4 and
Figures 3–7show the ﬁnal reclaimed water that was to be
blended with a conventional water source. The 50
probability data for the period 1995 to 2001 for
the Old Plant and 2002 to 2010 for the New Plant respectively
are shown. Table 5 presents the macro elements for the raw
untreated and treated ﬁnal water of the New Plant. Table 6,
presented at the end of the paper, shows the water quality of
different water sources in Windhoek’s distribution system.
The reclaimed water is compared with the surface water and
three different borehole groups, individually and grouped
together, with the primary objective to determine if the
advanced wastewater treatment system can reliably reduce
contaminants of public health concern to levels such as that
the health risk posed by any proposed potable use of the trea-
ted waterare no greater than those associated with the present
water supply, as recommended by the Western Consortium for
Public Health ().
Numerous desktop, pilot and full scale studies have been
conducted on the Old Plant between 1991 and 1998 as
well as on the New Plant between 2002 and 2010. Many
reports have been published about the performance of
the plant, the removal of Giardia and Cryptosporidium,
viruses and other substances (Menge et al. ,).
It will not be possible to deal with all data and infor-
mation in this publication. For the purpose of this paper,
only ﬁve operational parameters (turbidity, chemical
oxygen demand (COD), DOC, ultraviolet absorbance
) and heterotrophic plate count (HPC)) are used
to discuss the improvement of water quality by adding
additional barriers to reach the water quality standards
stet in the design report. The 95% probability data show
that the new plant has achieved the target values for all
the parameters, except COD. The aesthetic ﬁnal water
quality improved by a factor of 10, microbial risk was
Table 3 |Raw water quality data of the Old Plant and New Plant. The amount of data (n) are shown at the bottom of the table
Old Plant New Plant Old Plant New Plant Old Plant New Plant Old Plant New Plant Old Plant New Plant
50%P 12.28 2.17 34.31 26.00 11.26 9.18 25 18 3,133 4,800
85%P 28.50 6.20 39.96 33.00 13.63 10.75 28 20 14,335 19,095
95%P 41.00 15.48 43.80 37.00 14.44 12.00 31 23 42,235 39,150
99%P 93.27 49.15 49.28 41.00 16.14 13.17 34 26 71,170 78,765
(n) 318 368 307 368 215 368 317 370 223 348
Table 4 |Performance data of the Old Plant and New Plant are compared with highlight the vast improvement accomplished by additional treatment barriers. The Performance Target and
Maximum Allowable limit and the amount of data (n) are shown at the bottom of the table
Old Plant New Plant Old Plant New Plant Old Plant New Plant Old Plant New Plant Old Plant New Plant
50%P 0.65 0.06 12.0 5.70 3.82 1.60 5.44 1.25 1.00 0.00
85%P 1.10 0.07 17.3 10.4 5.78 2.31 7.38 1.85 5.00 0.40
95%P 1.62 0.08 20.3 14.8 6.79 2.60 8.41 2.23 16.0 1.15
99%P 2.52 0.13 22.6 19.5 8.37 3.00 11.11 2.83 134 5.83
Target 0.10 0.10 10.0 10.0 3.00 3.00 5.00 5.00 80 80
Max 0.20 0.20 15.0 15.0 5.00 5.00 6.00 6.00 100 100
(n) 314 363 301 368 212 367 312 369 308 366
220 P. du Pisani & J. G. Menge |A critical review of the design philosophy of a drinking water reclamation plant Water Science & Technology: Water Supply |13.2 |2013
reduced with 1 log by comparing the ﬁnal water quality
data of Old Plant and New Plant. In addition, the organic
composite parameters DOC, COD and UV
considerable reduction and improvement in concentration
Figure 4 (DOC), Figure 5 (COD) and Figure 6 (UV
are a proof that the additional barriers, namely pre-ozona-
tion, EC, ozonation and BAC followed by GAC in the
New Plant compared with normal coagulation, and carbon
adsorption without any oxidation in the Old Plant vastly
enhanced the organic removal. COD is the parameter that
did not follow this pattern. There could be two reasons for
this. Firstly, the COD method is not accurate lower than
12 mg/l. Secondly, the COD seems to contain a dissolved
fraction of inert material (most probably organic and inor-
ganic material) which is not removed by the treatment
steps described above. This is also evident when one com-
pares the percentage hypothetical fractions of DOC and
COD during the operational period (values based on 95
percentile) of the New Plant, where the biodegradable frac-
tions are 27 versus 10%, the non-biodegradable-adsorbable
fractions are 33 versus 22% and the non-biodegradable‐
non-adsorbable fractions are 40 versus 68%. This
Figure 3 |Turbidity Old Plant and New Plant.
Figure 5 |COD Old Plant and New Plant.
Figure 4 |DOC Old Plant and New Plant.
Figure 6 |UV254 Old Plant and New Plant.
Figure 7 |HPC Old Plant and New Plant.
221 P. du Pisani & J. G. Menge |A critical review of the design philosophy of a drinking water reclamation plant Water Science & Technology: Water Supply |13.2 |2013
observation was already reported during the pilot studies
(Miecznic ). The 95
percentile of COD was within
the maximum allowable target of 15 mg/l. PAC was rec-
ommended as an additional operational safety barrier in
the event that another barrier should fail. During an ozone
failure PAC was dosed. Increased fouling of the UF mem-
branes resulted in an increase in backwash cycles. PAC
dosing was discontinued and the ﬂow through the plant
was reduced to the point that the backwash cycles could
be managed. No tests were run to see if the addition of
powder activated carbon would have had a positive inﬂu-
ence on the COD removal. It is clear that DOC and UV
are more sensitive methods to measure the presence of
organic carbon than the COD method. Further investi-
gations into the fractionation of DOC have revealed that
the different DOC fractions react differently to different treat-
ment steps (Jacquemet et al.). The COD will be phased
out as surrogate parameter for drinking water treatment.
As mentioned earlier, direct potable reclamation was also
pioneered in Southern Africa since the 1950s. Each treatment
step was rigorously researched and tested under extreme con-
ditions. Full scale operation added further conﬁdence. An
epidemiological study was conducted (Isaäcson et al.)
over a period of ten years and concluded that no correlation
could be found between the consumption of reclaimed water,
natural water and disease patterns recorded. Diseases were
rather related to cultural differences in lifestyles. Although
proposed, such a study was not conducted prior or during
the operation of the New Plant, for the following reasons:
the sample group would be too small and diverse, the control
would be too small, enormous movement of people from one
township to another, questionable medical records and
ﬁnally the cost would be too high for an effort that would
likely have a similar conclusion as the ﬁrst study. The focus
was rather to concentrate on rigorous monitoring, imple-
menting ISO and HACCP certiﬁcation and conducting a
risk assessment (Ander & Forss ) for the whole treatment
process, following a ranking approach as suggested in Water
Safety Plans (WHO ). Vigilance, training and motivation
of the operational staff play a major part of the operational
success of the reclamation scheme in Windhoek. Continued
research is part of the agreement with the private operator
of the plant. A steering committee suggests new research pro-
jects and reviews all the research work on an annual basis. In
this way students are exposed to the ﬁeld of water reuse
(Menge et al.). A special monitoring programme for
the distribution is followed, which is described in detail by
Iiputa et al..
In addition to routine monitoring, a separate health
research programme is conducted in parallel, which covers
virus testing for cytopathogenic effect and polymerase
chain reaction toxicity (waterﬂea lethality and urease
enzyme), as well as Ames Salmonella mutagenicity testing.
These tests are negative for the water sources providing
Windhoek with treated drinking water. A special pro-
gramme to trace the removal of endocrine disrupting
compounds (EDCs) and medical substances has until now,
not shown that these compounds are present in the ﬁnal
water of the New Plant.
The ﬁnal water quality impact of the New Plant on the
distribution system is twofold. The parameters mentioned ear-
lier have a positive inﬂuence on the distribution system and
Table 5 |Macro elements which are important in water supply which are not removed by the New Plant and bromate in ﬁnal water and distribution system (values with* are raw design
values and NS ¼not speciﬁed)
TKN (mg/l-N) TKN (mg/l-N) NH
(mg/l-N) TDS (mg/l) TDS (mg/l) BrO
50%P 1.90 0.25 0.35 0.05 10.00 10.00 737 858 0.08 0.02
85%P 3.30 0.51 0.98 0.15 16.00 16.00 858 973 .011 0.02
95%P 4.39 1.19 1.80 0.15 20.45 21.00 931 1,039 0.12 0.03
99%P 7.79 2.30 3.01 0.26 27.45 26.88 1,005 1,126 0.13 0.03
Target 3.31* 1.56 2.30* NS 7.71* NS 563* 1,000 0.01 0.01
Max NS 1.95 NS 0.10 NS 10.00 NS 1,200 0.025 0.025
(n) 344 343 363 361 352 353 369 368 17 40
222 P. du Pisani & J. G. Menge |A critical review of the design philosophy of a drinking water reclamation plant Water Science & Technology: Water Supply |13.2 |2013
Table 6 |Comparison of the water quality in the distribution system with that of underground, surface sources and reclaimed water with the Namibian Guideline Group A quality over the period January 2006 to June 2007 (all
values are based the 95th% percentile)
Water Reservoirs Consumer
Aesthetic pH –7.74 7.93 8.01 7.90 8.02 8.4 7.94 8.01 6,0-9,0
Conductivity mS/m 25WC 130 68 165 108 29 160 75 67 150
TDS calc mg/l 871 457 1106 722 194 1072 503 449 1000
Turbidity NTU 5.27 4.64 8.08 4.88 1.1 0.082 1.2 1.5 1.0
Total alkalinity mg/l CaCO
513 237 246 334 111 237 283 228 –
Total hardness mg/l CaCO
350 252 281 281 110 242 229 226 300
Calcium hardness mg/l CaCO
220 112 93 158 71 153 145 140 375
Magnesium hardness mg/l CaCO
125 69 85 87 29 87 85 83 290
Chlorophyll A μg/l –––– 1.02 0.95 ––
Temperature WC 28 25 58.5 34.1 27 25.8 27 27 –
HPC per 1 ml 10000 152 10000 5076 10000 1232 295 100
Total coliform per 100 ml 10 045 13.8 0 0000
Faecal coliform per 100 ml –––– –0–0
E. coli Tryptone per 100 ml 57 048 35 -0 0 1.9 0
Faecal streptococci per 100 ml –––– –0–––
Pseudomonas per 100 ml –––– –0–––
Clostridium spores cfu 100 ml –––– –0–––
Clostridium viable cfu 100 ml –––– –0–––
Som. coliphage 100 ml PFU/100 ml –––– –0–––
Protozoa Giardia –––– 00 0 ––
Cryptosporidium –––– 00 0 ––
DOC mg/l 1.39 3.15 1.5 2.29 4.77 2.64 4.2 4.5 –
COD dis mg/l –––– –17.2 –––
UV 254 dis abs/cm 0.0691 0.0649 0.0582 0.0618 0.0759 0.0197 0.0707 0.0662 –
Free chlorine mg/l 2.2 2.2 –1.95 1.26 1.72 2.2 2.2 0,1–
Total chlorine mg/l 2.2 2.2 –2.07 1.68 –2.2 2.2 0,1–
Total trihalomethane μg/l 30 147 18 38 96 26 107 107 –
Chloroform μg/l 10 73 6 16 54 3 57 57 –
Dichloromonobromoform μg/l 9 43 7 13 26 4 27 27 –
Monochlorodibromoform μg/l 9 21 4 6 11 9 16 16 –
Bromoform μg/l 2 10 1 3 5 10 7 7 -
precipitation pot. mg/l CaCO
–––– –– – ––
223 P. du Pisani & J. G. Menge |A critical review of the design philosophy of a drinking water reclamation plant Water Science & Technology: Water Supply |13.2 |2013
improve the quality. The DOC of the surface water that is
blended with the potable reclaimed water is between 2.5
after a good rainy season and 8.0 mg/l during times of
drought. The high nitrates released from the wastewater treat-
ment plant, entering the New Plant, are diluted by the natural
water and raise the level in the distribution system slightly.
The 95% probability is 5.3 mg/l-N (guideline value 10). The
TDSs are raised considerably due to the higher percentage
of recycling, especially during low rainfall periods, but have
not exceeded the maximum guideline of 800 mg/l. During
the design stage the WHO bromate guideline was
0.025 mg/l (WHO ), which was then changed to
0.01 mg/l (WHO ). Due to high bromide levels in the
wastewater, it is very difﬁcult to reach the 0.01 mg/l guideline
target, without compromising the purpose of the ozonation
step. For this reason partial treatment with RO is considered,
which will reduce the concentration well within guideline
levels. RO is considered and pilot testing has started in the
second half of 2011. The operational results are currently
being evaluated to calculate the operating cost. The quality
results are well within expected limits. It can be argued that
in an inland situation, with extreme water shortages, that
RO would not be a preferred option, as the water losses
can amount to between 10 and 15% and the removal of
brine would also be a cost factor to be considered.
The reclamation scheme has operated reliably and has
saved Windhoek numerous times from severe water restric-
tions and in this way potable reclamation in Windhoek has
contributed to a stable economy despite severe droughts and
Were design expectations met?
The ﬁnal water quality targets set out in the design were
achieved by adding:
1. UF as an additional barrier for turbidity, microorganism
and protozoa removal.
2. Ozonation as treatment step to reduce organic precursors
for THM and effectively deal with EDCs and to reduce
the NOM content of the water.
Table 6 |Continued
Water Reservoirs Consumer
TKN mg/l as N –––– –4.37 –––
-N) mg/l as N 0.27 0.16 0.27 0.215 0.27 0.09 0.1595 0.27 1.0
Ortho phosphate (P) mg/l 0.9 0.91 0.9 0.9 0.9 0.9 0.9 0.9 –
-N) mg/l as N 1 4.76 1.79 3.07 1 22 5.84 5.72 10
-N) mg/l as N 0.5 0.345 0.5 0.42 0.5 0l5 0.5 0.5 –
Inorganics K mg/l 24.7 10.74 36.9 20.8 9.3 39.57 16.7 15.8 200
Na mg/l 101 57 158 93 18.5 266 96 97 100
CI mg/l 42 66 71 61 23.6 277 118 110 250
mg/l 265 CO 496 224 20 187 101 85 200
Fmg/l2.10 0.36 4.37 1.79 0.33 0.61 0.45 0.48 1,5
Br mg/l 0.20 0.10 0.30 0.11 0.127 0.14 0.15 0.11 1.0
Mn mg/l –––0.38 0.02 0.01 0.03 0.03 0.05
Fe mg/l 0.55 0.62 0.35 0.54 0.06 0.02 0.08 0.1 0.1
224 P. du Pisani & J. G. Menge |A critical review of the design philosophy of a drinking water reclamation plant Water Science & Technology: Water Supply |13.2 |2013
From the data analysed and reported, it is concluded,
that the multiple barrier system as incorporated in the
New Plant has indeed achieved the water quality as set
out in the design criteria of the plant.
What would be done differently if we were to design a
During the design phase in 1996 it was obvious that mem-
branes would be preferred as a primary treatment step to
remove suspended matter. At that time, the cost of mem-
branes was too high and there was no track record
available on membrane life over an extended period in waste-
water treatment. The ozone, BAC and GAC combination
proved in numerous international studies to be effective in
reducing organic compounds, especially EDCs and pharma-
ceuticals. A possible combination of treatment steps could be:
1. Membrane technology, which would replace a clariﬁer at
the biological treatment plant, chemical addition, ﬂoccu-
lation, DAF and RSF.
2. An oxidation and biological adsorption (BAC) and
adsorption (GAC) step would be maintained, to provide
an additional safety barrier against ﬂuctuations in waste-
water efﬂuent DOC and N concentrations.
3. Partial or full RO (depending on the situation) would be
considered as a polishing step to reduce DBPs, the salt
concentration and remove the refractive or non-biode-
gradable non-absorbable organic portion of DOC to
lower than 0.1 mg/l.
4. Ultraviolet as ﬁnal disinfection with the addition of a
chlorine residual of 0.4 mg/l to protect the water in the
distribution system in a warm climate with temperatures
between 15 and 32 WC.
Does direct potable reclamation have a future?
Windhoek’s residents have been supplied with directly
reclaimed water for nearly 43 years. The additional treatment
barriers discussed, have added an additional margin of safety,
shown by the data. With a proposed treatment train of a mem-
brane-ozone-BAC-GAC-RO-UV the bacteria, virus and
protozoa should be completely removed, even at very high
ﬂuctuating concentrations. Organic substances, measured
as DOC concentration would be below 0.01 mg/l. Variations
in raw water ammonia concentrations would also be
The Windhoek experience has proven that direct
potable reclamation indeed has a future in areas where
alternatives are not available.
The contribution of experts and students locally and abroad
are hereby acknowledged. A sincere appreciation to all who
contributed so diligently to collecting and analysing data,
compiling information and operating the different plants.
Ander, H. & Forss, M. Microbiological Risk Assessment of the
Water Reclamation Plant in Windhoek, Namibia. Master of
Science Thesis. Department of Civil Engineering, Chalmers
University of Technology. Göteborg, Sweden.
Burmeister, Van Niekerk Consulting Engineers Report on the
Upgrading and Extensions of Goreangab Reclamation Works,
Submitted to the City of Windhoek in October 1992, pp. 65–
67. City of Windhoek, Namibia.
du Pisani, P. L. Direct reclamation of potable water at
Windhoek’s Goreangab reclamation plant.Desalination 188
EC The European Community Guidelines for the use of
water for human consumption (80/778/EWG). Council
Directive of 15 July 1980.
Haarhoff, J. Report of a site investigation conducted at the
Goreangab water treatment plant from 1991–07–08 to 1991–
07–16. Report submitted to the City Engineer, City of
Haarhoff, J. & Menge, J. Proposals for Goreangab pilot study.
Report submitted to the City Engineer, City of Windhoek.
Haarhoff, J. & van der Merwe, B. Twenty-ﬁve years of
Wastewater Reclamation in Windhoek, Namibia.Water
Science and Technology 33,25–35.
Haarhoff, J., van der Walt, C. J. & van der Merwe, B. Process
design considerations for the Windhoek water reclamation
plant. Proceedings of the Water Institute of Southern Africa
(WISA), held in Cape Town, May 1998, unpublished report.
Haarhoff, J., Kubare, M., Mamba, B., Krause, R., Nkambule, T.,
Matsebula, B. & Menge, J. NOM Characterization and
Removal at Six Southern African Water Treatment Plants.
Proceedings at the UNESCO-IHE (Institute for Water
Education) TU-DELFT, High Quality Drinking Water
Conference 2009. TU Delft The Netherlands, June 2009.
225 P. du Pisani & J. G. Menge |A critical review of the design philosophy of a drinking water reclamation plant Water Science & Technology: Water Supply |13.2 |2013
Iiputa, G. I., Nikodemus, K. & Menge, J. Strategic water
quality monitoring for drinking water safety in Windhoek.
Proceedings of the Water Institute of Southern Africa (WISA),
held in Sun City, May 2008, unpublished report.
Isaäcson, M., Sayed, A. R. & Hattingh, W. Studies on Health
Aspects of Water Reclamation during 1974 to 1983 in
Windhoek, South West Africa/Namibia. Report WRC 38/1/87
to the Water Research Commission, Pretoria, South Africa.
Jacquemet, V., Gherman, E. C., Konig, E. & Theron-Beukes, T.
Organic Matter Evolution in the Treatment Process
of the New Goreangab Water Reclamation Plant at
Windhoek, Namibia. 6th International Water Association
(IWA) Specialist Conference on Wastewater Reclamation and
Reuse for Sustainability –Guiding the Growth of Water
Reuse, Antwerp, Belgium, October 2007.
Meiring, P. G. J. Partners (Consulting Engineers). A Guide for
the Planning, Design and Implementation of a Water
Reclamation Scheme. Prepared for the Water Research
Commission, Pretoria, South Africa.
Menge, J. G., Du Pisani, P. & van der Merwe, B. Water Quality
Control in a Third World Country: Challenges to ensure good
quality reclaimed water in Windhoek, Namibia. Proceedings
at the 6th International Water Association (IWA) Specialist
Conference on Wastewater Reclamation and Reuse for
Sustainability under the Theme Guiding the Growth of Water
Reuse. Antwerp, Belgium, October 2007.
Menge, J. G., Haarhoff, J., König, E., Mertens, R. & van der
Merwe, B. Occurrence and removal of Giardia and
Cryptosporidium at the Goreangab Reclamation Plant. Water
Science and Technology: Water Supply 1(1), 97–106.
Miecznic, B. Quantiﬁcation of biodegradation adsorption in
ozonation/GAC ﬁlters for domestic wastewater reuse.
Research Report for the Municipality of Windhoek. City
Engineer, Water Services. Diploma Thesis Environmental
Engineering, Technical University of Berlin, Berlin,
Namibian Guidelines Guidelines for the Evaluation of
Drinking Water for Human Consumption with Regard to
Chemical, Physical and Bacteriological Quality. Department
of Water Affairs, Republic of Namibia.
National Institute of Water Research (NIWR), Council for
Scientiﬁc and Industrial Research (CSIR) In collaboration
with the Water Research Commission (WRC) Chapter
Eight –Ozonation. In: Manual for Water Renovation and
Reclamation, 2nd Edition. Technical Guide K42, UDC
826.315. Pretoria, Republic of South Africa.
Rand Water Potable Water Quality Guidelines. Scientiﬁc
Services, Rand Water, Johannesburg, Republic of South
USEPA Surface Water Treatment Rule (SWTR) (54 FR
27486, June 29, 1989. USEPA, Washington, DC).
USEPA Drinking water regulations and health advisories,
Ofﬁce of Water, EPA 822-B-96–002, Washington, DC.
Western Consortium for Public Health Health Effects Study –
City of San Diego: Total Resource Recovery Project. AQUA
III San Pasqual Health Effects Study. Final Summary Report.
Prepared by Western Consortium for Public Health.
Williams, R. Wastewater Reuse in Residential and Commercial
Situations. CMPS&F, Brisbane. Churchill Fellowship.
WHO, World Health Organization Health Effects relating to
Direct and Indirect reuse of Wastewater for Human
Consumption. Report of an International Working Meeting,
Amsterdam, The Netherlands, 13th–16th January 1975.
Technical Paper No 7. World Health Organisation
International Reference Centre for Community Water
Supply. The Hague, The Netherlands.
WHO, World Health Organization Guidelines for Drinking
Water Quality, Vol. 1, Recommendations, 2nd Edition. World
Health Organization, Geneva.
WHO, World Health Organization Bromate in Drinking-
Water. Background Document for Preparation of WHO
Guidelines for Drinking-water Quality. World Health
Organization (WHO/SDE/WSH/03.04/78), Geneva.
WHO, World Health Organization Water Safety Plan: step-
by-step Risk Management for Drinking Water Suppliers.
World Health Organization, Geneva.
First received 7 February 2012; accepted in revised form 28 August 2012
226 P. du Pisani & J. G. Menge |A critical review of the design philosophy of a drinking water reclamation plant Water Science & Technology: Water Supply |13.2 |2013