Physics and Chemistry of the Earth 67–69 (2014) 226–235

Contents lists available at ScienceDirect

Physics and Chemistry of the Earth

journal homepage: www.elsevier.com/locate/pce

Characterisation of raw sewage and performance assessment of primary

settling tanks at Firle Sewage Treatment Works, Harare, Zimbabwe
Simon Takawira Muserere a,b,⇑, Zvikomborero Hoko a,1, Innocent Nhapi a,2
a
Civil Engineering Department, University of Zimbabwe, MP 167, Mt Pleasant, Harare, Zimbabwe
b
Harare Water Department, City of Harare, 2nd Floor, Old Mutual House Corner, Sam Nujoma and Speke Avenue, Harare, Zimbabwe

a r t i c l e i n f o a b s t r a c t

Article history: The need for more stringent effluent discharge standards as prescribed by the Environmental Manage-
Available online 25 October 2013 ment Act 20:27 to protect the environment can be sustainably achieved with the aid of Activated Sludge
Models. Thus, the researchers believe it is time to re-evaluate wastewater characteristics at Firle Sewage
Keywords: Treatment Works (STW) and make use of activated sludge simulators to address pollution challenges
Biological Nutrient Removal caused by the sewage plant. Therefore, this paper characterizes raw sewage and assesses settled and
Characterise unsettled sewage in order to evaluate the performance of the primary treatment system and the suitabil-
Firle Sewage Treatment Works
ity of the settled sewage for treatment by the subsequent Biological Nutrient Removal (BNR) system at
Settled sewage
Treatability of sewage
Firle STW. Parameters studied included COD, BOD, TKN, TP, NH3, TSS, pH and Alkalinity. Composite sam-
ples were collected over a 9-day campaign period (27 June to 6 July 2012), hourly grab samples over
24 hrs and composite samples on 6 March 2012 which were then analysed in the lab in accordance with
Standard Methods for the Examination of Water and Wastewater to support the City of Harare 2004–
2012 lab historical records. Concentrations for unsettled sewage in mg/L were COD (527 ± 32), BOD
(297 ± 83) TKN (19.0 ± 2.0), TP (18 ± 3), NH3 (24.0 ± 12.9), TSS (219 ± 57), while pH was 7.0 ± 0 and Alka-
linity 266 ± 36 mg/L. For settled sewage the corresponding values in mg/L were COD (522 ± 15), BOD
(324 ± 102), TKN (21.0 ± 3.0), TP (19.0 ± 2.0), NH3 (25.6 ± 11.2), TSS (250 ± 66), while pH was 7.0 ± 0
and Alkalinity 271 ± 17 mg/L. The plant design values for raw sewage are COD (650 mg/L), BOD
(200 mg/L), TKN (40 mg/L) and TP (11 mg/L). Thus, COD and nitrogen were within the plant design range
while BOD and TP were higher. Treatability of sewage in BNR systems is often inferred from the levels of
critical parameters and also the ratios of TKN/COD and COD/TP. The wastewater average settled COD/
BOD, COD/TP and TKN/COD ratio were 1.7 ± 0.5, 27.1 ± 3.1 and 0.04 ± 0.01 respectively and corresponding
unsettled ratios were 1.8 ± 0.5, 30.77 ± 6.8 and 0.04 ± 0 respectively. Thus, treatability by the 3-stage BNR
system appears highly feasible for nitrogen and is likely to be complex for phosphorous. Fractionation of
COD, TP and TN is recommended to appropriately advise further steps to optimise the plant operations.
Ó 2013 The Authors. Published by Elsevier Ltd. Open access under CC BY-NC-ND license.

1. Introduction develop national and local action plans aiming at safe and produc-
tive use of wastewater (Sato et al., 2013). According to
The availability of up to date information on wastewater gener- environmental experts, a fast growing population produces a great
ation, treatment, and use is important for policy makers, research- deal of waste around the globe, hence in order to ensure that the
ers, and practitioners, as well as public institutions, in order to world is a safer and cleaner place for future generations to live
in, there is an ever-increasing pressure to reduce waste and im-
prove waste treatment from both a moral and legal perspective.
A recent World Bank study found that for improved sanitation
there are almost 90% of the population in developed countries,
but only about 30% of the population in developing countries,
⇑ Corresponding author at: Civil Engineering Department, University of has access to improved sanitation. As water supplies fall and stress
Zimbabwe, MP 167, Mt Pleasant, Harare, Zimbabwe. Mobile: +263 773142217; rises in many areas, the potential resource of wastewater is being
fax: +263 4706221. widely recognized, says the study published in 2013 by Elsevier
E-mail addresses: [email protected], [email protected] Journal Agricultural Water Management. According to a United Na-
(S.T. Muserere), [email protected] (Z. Hoko), [email protected] (I. Nhapi).
1 tions (UN) study of 2013 among 48 Sub-Saharan African countries,
Mobile: +263 772338899.
2
Mobile: +263 772111487. complete information on wastewater generation, treatment, and

http://dx.doi.org/10.1016/j.pce.2013.10.004
1474-7065 Ó 2013 The Authors. Published by Elsevier Ltd. Open access under CC BY-NC-ND license.
 S.T. Muserere et al. / Physics and Chemistry of the Earth 67–69 (2014) 226–235 227

use is available from only three countries which are Senegal, Sey- this study characterizes unsettled and settled sewage concentra-
chelles and South Africa. The study argues that even this informa- tions at Firle STW in order to assess the efficiency of the primary
tion is old, as the data from Seychelles and South Africa relate to treatment system and subsequent treatability of the settled sew-
2000–2003. The countries with partial data available are 11 with age by the 3-stage BNR system at the plant.
the remaining 32 countries including Zimbabwe having no data
on wastewater. The study found out that most wastewater goes 2. Study area description
untreated in Sub-Saharan Africa, where water pollution triggers
the spread of waterborne diseases such as diarrhoea and cholera. 2.1. Population and socio-economic issues
The characterization of wastewater in terms of quantity and
quality is principally a function of the wastewater origin i.e. indus- Population of Zimbabwe is approximately 12.97 million (Zim-
trial or domestic. The Federal Clean Water Act defines pollutant as babwe National Statistics Agency, 2012) with 16.9% leaving in
meaning ‘‘dredged spoil, solid waste, incinerator residue, sewage, greater Harare (City of Harare, Chitungwiza, Ruwa, Norton and Ep-
garbage, sewage sludge, munitions, chemical wastes, biological worth). It is reported that in 2011 only 850,000 people out of the
materials, radioactive materials, heat, wrecked or discarded 12.97 million people were formally employed, John Robertson an
equipment, rock, sand, cellar dirt, and industrial, municipal, and economic analyst claims that the number of formally employed
agricultural waste discharged into water’’. Despite this broad, Zimbabweans in 2011 equals that of 1970. Thus, it is critical to pro-
all-encompassing definition, wastewater can typically be charac- vide sustainable wastewater management system for Harare, given
terized as either domestic, industrial, or some combination of both the economic environment in the country.
as suggested by Corbitt (1999). According to Environmental
Protection Agency the first step in addressing wastewater prob- 2.2. Sewer infrastructure in Harare
lems is to develop a source reduction program and selecting an
appropriate pre-treatment system, this can only be possible by The sewer network of Harare consists mainly of reinforced con-
analysing the physical and chemical characteristics of the process crete pipes for diameters greater than 450 mm and PVC for
wastewater stream in question. The measured levels of pollutants 400 mm diameters and below, and steel pipes for river crossings.
are subsequently compared with the discharge regulations limits The total length of sewer network is approximately 4,500 km with
applicable to each area. 75% of Harare on sewer reticulation and the rest on septic tanks
Environmental friendly wastewater disposal methods can be according to Harare Sewage Master Plan report by Société Anon-
through surface, subsurface or dilution means, and is much de- yme Française d’ Etude de Gestion et d’ Entreprises (Safege)
pended on the degree of treatment. Appropriate wastewater treat- (2003) a French consortium. This scenario has not changed much
ment is necessary to prevent contamination of receiving waters to since then because of economic decline over the past ten years.
levels which will interfere with their intended use. Intended water The sewage treatment plants in Harare are overloaded with total
use varies from water supply, recreation, or any other required design capacity of 219,500 m3/d while current inflows average
purpose such as irrigation. Thus, wastewater treatment consists 287,000 m3/d. Firle STW catchment mainly comprises the wa-
of applying a series of processes to improve its quality suitable tershed of the Mukuvisi River to the South and South East of down-
for disposal. Usually wastewater treatment will involve collecting town Harare (Fig. 1).
the wastewater in a central isolated location, the Wastewater
Treatment Plant, and then process it to acceptable standard. Gener- 2.3. Firle works Unit 4
ally, large volumes of wastewater are involved, such that treatment
processes are carried out on continuously flowing wastewaters The study site was Firle Unit 4 which has course and fine
rather than batches. Wastewater treatment can then be catego- screens that are manually racked, with two mechanically operated
rized by the nature of the treatment process operation being used detritor grit removal system linked to a division box splitting flow
such as, physical, chemical or biological. to Units 1, 2, 3 and 4. A 1200 mm diameter gravity line at a slope of
According to previous research findings Harare’s current water 1 in 790 connects this division box to a central distribution box for
management challenges are caused by discharging poorly treated Unit 4’s four Dortmund Primary Settling Tanks (PSTs). The PSTs are
wastewater into Lake Chivero, the city’s raw water source accord- linked by a 275 mm diameter 6 m deep reinforced concrete pipe-
ing to (Nhapi, 2004). Primarily the water and wastewater chal- line gravitating raw sludge to a raw sludge pump station, which
lenges in Harare are a symptom of lack of adequate skills and then pumps sludge to anaerobic digesters. Effluent from the four
technical knowhow to manage the high tech systems for water PSTs gravitates through a 400 mm diameter gravity line into a
and wastewater supply infrastructure. It is believed that Firle channel that directs tank effluent into the fermentation basin via
STW, the largest sewage treatment plant in Harare, with a design a box flume equipped with a digital flow recorder. The four Dort-
capacity of 144,000 m3/day and approximately 10 km upstream mund PSTs have a total surface area of 530.9 m2 and designed for
of the lake is the major single polluter of Lake Chivero according 45% COD reduction efficiency according to the designer Stewart
to (Japan International Cooperation Agency, 1996). Harare Water Scott International (1984). In the settling tank the solids (sludge)
Quality Laboratory wastewater effluent reports of 2012 revealed and liquid (effluent) portions of the incoming raw sewage are sep-
that Firle STW normally operates in the high environmental hazard arated. The effluent overflows a circular weir and gravitates to the
range in terms of Environmental Management Act Chapter 20:27 BNR unit (Fig. 2).
effluent discharge standards.
Designs of wastewater plants are normally based on assump- 3. Materials and methods
tions derived from historical records and at design stage it is as-
sumed that the wastewater treatment processes will work as per 3.1. Data collection and verification
design. Resultantly wastewater plants are seldom optimised since
in practice wastewater characteristics change with changing tech- Data was gathered from the City of Harare Laboratory records
nology while operations normally remain static. Plant performance for the period 2004–2012. To supplement the historical data, anal-
assessment involves adjusting primary settling tank performance yses were also carried out over a 9-day period from 27 June to 6
(i.e. solids removal) to correctly reflect primary effluent TSS and July 2012. Samples were taken from a division box upstream of
BOD5 concentration variations suggests Dold et al. (2011). Thus, PSTs (S2) and in the collection chamber downstream of PSTs (S3).
228 S.T. Muserere et al. / Physics and Chemistry of the Earth 67–69 (2014) 226–235

Fig. 1. Location map of Firle Sewage Treatment Works.

Fig. 2. Schematic diagram of Firle Unit 4 and sampling points: (S1. . ..S9 denotes sampling points).

The sampling points were selected such that sewage is completely taken were inspected to check if the samples were representative
mixed at these points. of the processes (especially well-mixing at sampling points), if all
The raw data collected from the City of Harare laboratory re- other recommended procedures were being followed (such as
cords were processed, filtered, cleaned, and corrected before being sample preservation and processing according to standard meth-
used. To ensure quality compliance, sampling procedures were re- ods), and if analyses were repeatable as suggested by Dold et al.
examined and established. The positions where the samples were (2011).
 S.T. Muserere et al. / Physics and Chemistry of the Earth 67–69 (2014) 226–235 229

Table 1 sampling program, proper handling and preservation of samples,

Typical municipal wastewater characterisation adopted from (Metcalf and Eddy Inc., tracking chain of custody, sample ID procedures, field quality
2003).
assurance and proper analysis as suggested by Tjandraatmadia
Parameter High Medium Low et al. (2009). A one day flow weighted composite and hourly grab
COD 1200 750 500 sampling campaign on 6 March 2012 and a 9-day flow weighted
BOD 560 350 230 sampling campaign starting from 27 June to 6 July 2012 were car-
TKN 100 60 30 ried out to support the historical data. To ensure quality control,
NH3 75 45 20
TP 25 15 6
split sampling was done. The split sample was thoroughly mixed
TSS 600 400 250 before splitting it with one portion used as a control measure. Sam-
pH 8.0 7.5 7.0 ples were taken to the lab within an hour after taking the last sam-
Alkalinity eqv/m3 7 4 1 ple fraction and were preserved by keeping them in a freezer at
COD/BOD 2.5–3.5 2.0–2.5 1.5–2.0
temperatures below 4 °C in accordance with Standard Methods
COD/TP 45–60 35–45 20–35
for the Examination of Water and Wastewater by American Public
Health Association, American Water Works Association and Water
Environment Federation (APHA-AWWA-WEF, 2005). Each sample
3.2. Selection of parameters was re-analysed to check for repeatability of results.
The samples were collected by Firle STW samplers with re-
Common parameters used to assess the constituents found in quired expertise. A calibrated beaker tied to a 2 m long steel rod
wastewater include the following TS, TSS, TFS, TVS, VSS, ISS, FSS, was used to collect samples at the mid-depth of each sampling
TDS, VDS, FDS, NH4+, Org N, TN, TKN, TP, NO2 , NO3 , Org P, pH, point. The beaker was rinsed with acidified water first then three
COD, BOD5, TOC, NO3 , NO2 , Temperature, Alkalinity and various times with sample water before final sample collection. The sam-
gases as suggested by Metcalf and Eddy Inc (2003). In plant pling containers were filled to capacity and tightly closed.
models, influent flow rate, TKN, COD and TP concentrations are
essential data according to Dold et al. (2011). Information on influ-
ent ammonia, BOD, TSS and VSS is useful for supporting influent 3.4. Testing methods
wastewater characteristic fractions and for determining influent
inorganic solids which impact on sludge production (Water Envi- To select an appropriate laboratory, initially 6 L containers were
ronmental Research Foundation, 2003). At a minimum, daily aver- used which were then split into three 2 L samples and send to
age flow and flow-weighted composite concentrations are required three different laboratories, Zimlab (a SANAS accredited private
(Dold et al., 2011). Therefore, the following parameters were se- laboratory), Standards Association of Zimbabwe (SAZ) laboratory
lected for analyses: COD, BOD5, TKN, TP, NH3, TSS, pH and and City of Harare (COH) laboratory. Results of the three laborato-
alkalinity. ries were analysed and upon noticing missing COD and BOD results
on SAZ and COH laboratories records further enquiries revealed
this was due to broken down lab equipment. Zimlab equipment
3.3. Sampling method was fully operational and SANAS accredited hence was selected
for further sample analyses. The researchers participated in sam-
The six criteria for quality data collection were followed i.e. col- pling and analysis of the samples. Sample analyses were in accor-
lecting representative samples, formulating the objective of the dance with (APHA-AWWA-WEF, 2005). COD was analysed in

(a) (b)

(c)

Fig. 3. Settled and unsettled COD for Firle Sewage Treatment Plant: (a) Hourly grab samples on 6 March 2012 and (b) daily composite samples for the period 27 June to 6 July
2012. (c) Raw sewage COD concentrations for the period 2004–2006: Firle Sewage Treatment Plant.
230 S.T. Muserere et al. / Physics and Chemistry of the Earth 67–69 (2014) 226–235

(a) (b)

(c)

Fig. 4. Settled and unsettled BOD5 concentrations for Firle Sewage Treatment Plant: (a) Hourly grab samples on 6 March 2012 and (b) daily composite samples for the period
27 June to 6 July 2012. (c) Raw sewage BOD5 concentration for the period 2004–2012: Firle Sewage Treatment Plant.

(a) (b)

(c)

Fig. 5. Settled and unsettled TP concentration at Firle Sewage Treatment Plant: (a) Hourly grab samples on 6 March 2012 and (b) daily composite samples 27 June to 6 July
2012. (c) Raw sewage TP concentration for the period 2004–2011: Firle Sewage Treatment Plant.

accordance with APHA (5220). BOD was analysed according to 3.5. Data analysis
APHA (5210), TKN in accordance with APHA (4500-N), Phosphates
were determined according to APHA (4500-P), and TSS in accor- Normality of data collected was tested using Statistical Package
dance with APHA (2540), Ammonia in accordance with APHA for Statistical Sciences (SPSS) software. The data did not follow nor-
(4500 – NH3), pH in accordance with APHA (4500 – H+), Alkalinity mal distribution hence non parametric method was used to ana-
(2320B). lyse the data. Data transformation was not performed since the
The compositions of typical municipal wastewater (Table 1) as data was from historical records and the City of Harare laboratory
suggested by (Metcalf and Eddy Inc, 2003) were used to determine staff reconfirmed the results as correct. The means for each year
strength of parameters. were used to represent mean concentration for that year and were
 S.T. Muserere et al. / Physics and Chemistry of the Earth 67–69 (2014) 226–235 231

(a) (b)

(c)

Fig. 6. Settled and unsettled TSS concentrations for Firle Sewage Treatment Plant: (a) hourly grab samples on 6 March 2012 and (b) daily composite samples from 27 June to 6
July 2012. (c) Raw sewage TSS concentrations for the period 2004–2012: Firle Sewage Treatment Plant.

(a) (b)

(c)

Fig. 7. Settled and unsettled TA concentrations for Firle Sewage Treatment Plant: (a) Hourly grab samples on 6 March 2012 and (b) daily composite samples for the period 27
June to 6 July 2012. (c) Raw sewage TA concentrations for the period 2004–2012: Firle Sewage Treatment Plant.

used in analysing City of Harare historical results (secondary data). of four samples were analysed per month, giving a maximum of 48
However, average concentrations and standard deviations were samples per year per parameter. SPSS was self adjusting hence the
calculated for Zimlab analysed results (primary data). Standard N-value was not necessarily the same from 1 year to the next;
deviations were representing ±in the research results. City of Har- therefore N = 48 was maximum and N = 9 was the minimum on
are has five sewage treatment plants and does not have the capac- all the parameters analysed. Significance tests on concentration
ity to analyse samples for each plant on a daily basis. The change with state, settled and unsettled, were performed to assess
laboratory has a rotational plant assessment system, hence a total PSTs performance. Simply using the average concentrations and
232 S.T. Muserere et al. / Physics and Chemistry of the Earth 67–69 (2014) 226–235

iod 27 June to 6 July 2012 were plotted against date for TKN
(Fig. 8), NH3 (Fig. 9) and pH (Fig. 10). Settled and unsettled treat-
ability ratios were plotted against time for hourly grab sampling
results and against date for daily composite sampling results
respectively (Figs. 11–13).

4.1. Wastewater characterisation

Wastewater characterisation was based on average settled and

unsettled parameters concentrations (Table 2) and the concentra-
tions variation over time were also analysed using graphs
(Figs. 3–10).

Fig. 8. Settled and unsettled TKN concentrations for the period 27 June to 6 July
2012: Firle Sewage Treatment Plant. 4.1.1. COD classification
The COD concentration at Firle STW was 527 ± 32 mg/L and
522 ± 15 mg/L for unsettled and settled sewage respectively. Thus
COD concentration was in the low range for both settled and unset-
tled sewage according to Metcalf and Eddy Inc (2003) (Table 1).
The Firle STW COD concentration decreased from 2004 to
2006 and then increased from 2006 to 2010 when another de-
cline was experienced (Fig. 3c). This trend was also noted in
BOD5 (Fig. 4c) and TSS (Fig. 6c). The changes in quantity and
quality of wastewater such as the one experienced at Firle
STW is determined by many factors as suggested by (Henze
and Comeau, 2008), these include the behaviour and lifestyle
of inhabitants, technical and judiciary framework by which the
community is surrounded. At household level and industrial pro-
duction solid and liquid wastes are generated with significant
possibilities for changing the amounts and composition of the
waste streams generated. The concentration of pollutant is a
combination of pollutant load and the amount of water mixed
Fig. 9. Hourly grab settled and unsettled Ammonia concentration on 6 March 2012: with the pollutant, thus the daily or yearly polluting load such
Firle Sewage Treatment Plant. as the measurements being analysed at Firle STW form a good
basis for an evaluation of the composition of sewage as suggest
by Metcalf and Eddy Inc (2003). It therefore, follows that the
daily and yearly concentrations data collected for Firle STW gives
a good evaluation of its wastewater characteristics. According to
Henze and Comeau (2008) composition of wastewater can vary
with time on a given location partly due to variations in the dis-
charged amounts of substances. The major reasons for these
variations are water consumption and underground infiltration
during wastewater transportation. Thus, the COD concentrations
variations noted at Firle STW during the period under study are
normal for municipal sewers.

4.1.2. BOD5 classification

The settled and unsettled BOD5 concentrations were 324 ± 102
and 297 ± 83 mg/L respectively. The BOD5 concentrations were
Fig. 10. Settled and unsettled sewage pH values for the period 27 June to 6 July
2012: Firle Sewage Treatment Plant. therefore, in the low to medium category according to Metcalf
and Eddy Inc (2003) (Table 1).
Biochemical Oxygen Demand (BOD) is an important water pol-
lutant to be studied for Firle STW Unit 4 since it employs biological
standard deviation could not expose extent of concentration treatment processes. The pollutant degrade biochemically and af-
change given the unexpected low removal efficiencies of Primary fect water quality according to the location as well as the strength
Settling Tanks. of the discharge. The BOD changed from one year to the next
(Fig. 4c). It becomes critical to analyse the concentration of this
parameter considering that Firle STW is located 10 km upstream
4. Results and discussions of Lake Chivero. The BOD5 concentrations measured exceeded the
design level of 200 mg/L as recommended by Stewart Scott Inter-
Concentrations were plotted against time based on 6 March national (1984) the designer of the plant. Therefore, it is important
2012 hourly grab sampling results (Figs. 3a–10a), concentrations to examine carefully the concentration levels of influent and efflu-
against date for the 27 June to 6 July 2012 flow weighted compos- ent BOD to establish potential water quality impacts in order to
ite sampling results (Figs. 3b–10b) and concentration against year regulate these discharges as suggested by Brill et al. (2010). How-
for historical results (City of Harare laboratory records) (Figs. 3c– ever, given the COD concentrations measured at Firle STW the
10c). Daily flow weighted composite sampling results for the per- BOD5 concentration may be favourable to the plant since they
 S.T. Muserere et al. / Physics and Chemistry of the Earth 67–69 (2014) 226–235 233

(a) (b)

Fig. 11. Unsettled and settled COD/BOD ratio for Firle Sewage Treatment Plant. (a) Ratio on 6 March 2012 and (b) for period 27 June to 6 July 2012.

(a) (b)

Fig. 12. Unsettled and settled COD/TP concentrations for Firle Sewage Treatment Plant. (a) Ratio on 6 March 2012 and (b) ratio for period 27 June to 6 July 2012.

recommended that TP concentrations be reduced at Harare sewage

treatment plants among other sources. In addition to pollution
challenges, there is a range of potential health impacts arising from
water-borne diseases and water-related disease vectors caused by
poor wastewater management upstream of a water reservoir such
as river blindness (onchocerciasis), bilharzia (schistosomiasis), and
guinea worm disease (dracunculiasis) as reported by Thornton
et al. (1996). Lake Chivero is a system under extreme pressure both
as a source of water and as a means of waste removal for the city of
Harare, two diametrically opposing functions according to Mhlan-
ga in a recent study on limnology of lakes and rivers.

Fig. 13. Settled and unsettled TKN/COD for the period 27 June to 6 July 2012: Firle 4.1.4. TSS classification
Sewage Treatment Plant.
TSS settled and unsettled concentrations were 250 ± 66 mg/L
provide the much needed readily biodegradable COD for phospho- and 219 ± 57 mg/L respectively. The TSS concentrations were cate-
rous removal. gorised as low strength (Table 1).

4.1.3. TP classification 4.1.5. TA classification

Unsettled TP concentration was 18.0 ± 3 mg/L and settled TP Total Alkalinity settled and unsettled concentrations were
was 19.0 ± 2 mg/L, which was classified as medium to high 271 ± 17 and 266 ± 36 mg/L respectively. The TA concentrations
strength (Table 1). The TP concentration (Fig. 5) was generally were categorised as low strength (Table 1).
higher than the design value 11 mg/L according to the design by
Stewart Scott International (1984).
4.1.6. TKN classification
The concentrations of TP measured at Firle STW were high com-
The unsettled TKN concentration was 19 ± 2 mg/L and settled
pared to plant design assumptions of 11 mg/L. According to Henze
concentration was 21 ± 3 mg/L. Thus, the TKN concentrations were
and Comeau (2008) most of the nutrients (phosphorous) are nor-
in the low strength category according to Mamais et al. (1993).
mally soluble and cannot be removed by settling, filtration, floccu-
The TKN concentrations measured are within design assump-
lation and any other means of solid–liquid separation. The total
tions of the plant and typical of municipal sewer with low indus-
phosphorous in municipal wastewater consists of 70–90% soluble
trial contribution such as Firle STW. Total nitrogen and its
orthophosphates and 30–10% organically bound phosphorous
components are important in designing the nitrification and deni-
which is in soluble or particulate form, a small fraction of unbiode-
trification processes. The nitrification efficiency of an activated
gradable phosphorous (Ekama and Marais, 1984). The organically
sludge process decreases under cyclic loading conditions (Loosdr-
bound phosphorous is converted to orthophosphate in the acti-
echt, 2008).
vated sludge and then used by bacterium (Metcalf and Eddy Inc,
2003f). Given the high concentrations of TP at Firle STW and the
complexities of TP treatment it is important to fractionate the 4.1.7. Ammonia classification
parameter to appropriately recommend best methods to manage The unsettled Ammonia concentration was 24.0 ± 12.9 mg/L and
it. Previous researches carried out downstream of Firle STW have settled concentration was 25.6 ± 11.2 mg/L. Thus, the Ammonia
234 S.T. Muserere et al. / Physics and Chemistry of the Earth 67–69 (2014) 226–235

Table 2
Summary of results: settled and unsettled average parameters concentrations, percentage removal efficiencies and classification of parameters: Firle Sewage Treatment Plant.

Parameter Settled Unsettled Percentage Remarks Classification

Average daily composite sample Average daily composite sample removal
results ± Sdv results ± Sdv
COD 522 ± 15 527 ± 32 1 Reduction in Low strength
concentration
BOD5 324 ± 102 297 ± 83 9 Increase in Medium strength
concentration
TKN 21.0 ± 3.0 19.0 ± 2.0 10 Increase in Low strength
concentration
TP 19.0 ± 2 18 ± 3 6 Increase in Medium to high
concentration strength
NH3 25.6 ± 11.2 24 ± 12.9 7 Increase in Low strength
concentration
TSS 250 ± 66 219 ± 57 14 Increase in Low strength
concentration
pH 7.0 ± 0 7.0 ± 0 0 Increase in Low strength
concentration
TA 271 ± 17 266 ± 36 2 Increase in Low strength
concentration
COD/BOD 1.7 ± 0.5 1.8 ± 0.5 7 Decrease in ratio Low
COD/TP 27.1 ± 3.1 30.7 ± 6.8 12 Decrease in ratio Low
TKN/COD 0.04 ± 0.01 0.04 0 No change Low

Table 3
Significance test for concentration change from unsettled to settled sewage for the period 27 June to 6 July 2012: Firle Sewage Treatment Plant: test statistics.b

Parameter COD BOD5 TP NH3 TSS pH TA COD/BOD COD/TP

Mann–Whitney U 21.000 32.000 28.500 31.000 28.000 29.000 36.000 31.000 30.000
Wilcoxon W 66.000 77.000 73.500 76.000 73.000 74.000 81.000 76.000 75.000
Z 1.726 .751 1.060 .839 1.104 1.024 .401 .839 .927
Asymp. sig. (2-tailed) .084 .452 .289 .401 .269 .306 .688 .402 .354
Exact sig. [2⁄(1-tailed Sig.)] .094a .489a .297a .436a .297a .340a .730a .436a .387a
b
Test statistics: a. Not corrected for ties.

concentrations were in the low strength category according to removal efficiencies calculated all parameters were below the
(Mamais et al., 1993). design values by Stewart Scott International (1984) and litera-
ture recommendations. The differences between settled and
unsettled concentrations were very low and varied with time;
4.1.8. pH classification
hence statistical analysis was performed to determine the signif-
The unsettled sewage pH was 7.0 ± 0 and settled sewage was
icant changes in concentrations. The statistical results were tab-
7.0 ± 0. Thus the pH level was in the low category according to
ulated (Table 3).
Metcalf and Eddy Inc (2003) (Table 1).
All the parameters concentrations did not change significantly,
When considering wastewater treatment processes pH control
with the significant values above 0.05 (Table 3) suggesting the pri-
is essential for activity of particular organism such as autotrophic
mary settling tanks were not performing well.
nitrifying bacteria in aeration basin, this is very important for Firle
STW that employs a BNR system (Water Environmental Research
Foundation, 2003). There is a fundamental relationship between
4.3. Assessment of treatability of settled sewage with subsequent
pH, acidity and alkalinity which plays a major role in biological
processes
wastewater treatment, hence the need to measure and control
pH for best results (Ekama and Wentzel, 2008). Thus, the low pH
To assess treatability of the sewage the treatability ratios COD/
measured at Firle STW requires a close attention to ensure plant
BOD, COD/TP and TKN/COD were calculated. The average settled
buffering capacity.
and unsettled COD/BOD, COD/TP and TKN/COD ratios were
The settled and unsettled parameter concentrations were tabu-
1.7 ± 0.5 and 1.8 ± 0.5; 27.1 ± 3.1 and 30.7 ± 6.8; 0.04 ± 0.01 and
lated in order to analyse the removal efficiencies of the primary
0.04 respectively. The treatability ratios were then plotted against
settling tanks and evaluate the overall wastewater classification
time for 6 March 2012 hourly sampling results, COD/BOD (Fig. 11a)
(Table 2).
and COD/TP (Fig. 12a) and against date for the period 27 June to 6
The wastewater parameters concentrations were generally in
July 2012 COD/BOD (Fig. 11b), COD/TP (Fig. 12b) and TKN/COD
the low range except for TP which was in the medium to high
(Fig. 13).
and BOD in the medium categories according to Metcalf and Eddy
The ratios between the various components in wastewater have
Inc (2003) (Table 1). The sewage at Firle STW can therefore, be cat-
significant influence on the selection and functioning of wastewa-
egorised as low strength sewage.
ter treatment processes (Henze and Comeau, 2008). The low Firle
STW COD/BOD ratio suggests a highly biodegradable sewage, how-
4.2. Primary settling tanks performance assessment ever, the ratio insignificantly improved across the PSTs (the ratio
responded to COD reduction across the PSTs). On the other hand,
The Firle STW PSTs removal efficiencies were calculated based the COD/TP ratio was very low below the recommended minimum
on settled and unsettled concentrations (Table 2). From the of 50 suggesting low Volatile Fatty Acids (VFAs) which is not
 S.T. Muserere et al. / Physics and Chemistry of the Earth 67–69 (2014) 226–235 235

suitable for biological phosphorous removal as suggested by (Eka- was presented with constructive comments from delegates assist-
ma and Wentzel, 2008). The COD/TP ratio decreased across PSTs ing in finalising the paper.
which is not favourable for biological phosphorous removal, the
decrease was a response to TP increase across the PSTs. The TKN/
COD ratio was below 0.09 which is favourable for nitrogen re- References
moval. According to (Ekama and Wentzel, 2008) as the TKN/COD
ratio decreases below 0.09 the size of the anoxic reactor decreases, APHA-AWWA-WEF, 2005. Standard Methods for the Examination of Water and
Wastewater twenty first edition. American Publishing Health Association,
hence the lower the ratio the better the biological nitrogen re-
American Water Works Association and Water Environment Federation, New
moval. The denitrification potential are approximately propor- York.
tional to the influent COD concentration. The nitrification Brill Jr., D.E., Sudhir, J.W.E., Barbara, R.K., Lence, J., 2010. Water quality impacts of
biochemical oxygen demand under transferable discharge permit programs.
capacity is approximately proportional to the influent TKN. Conse-
Earth Sciences 20, 445–455.
quently the TKN/COD ratio of the influent is a good parameter in Corbitt, R.A., 1999. Standard Handbook of Environmental Engineering. McGraw-Hill
terms of which to assess process nitrification–denitrification Companies, Inc., London.
behaviour because it gives approximately the ratio of the nitrate Dold, P.L., Jones, R.M., Takacs, I., 2011. Practical guidance for WWTP model
calibration and associated data gathering requirements. EnviroSim Associates
generated to the dentrification potential (Ekama and Marais, Ltd., Ontario.
1984). Ekama, G.A., Marais, G.V.R., 1984. Nature of municipal wastewaters. Theory, design,
and operation of nutrient removal activated sludge processes. Water Research
Commission, Pretoria.
5. Conclusions Ekama, G.A., Wentzel, M.C., 2008. Organic material removal. In: Hence, M., van
Loosdrecht, M.C.M., Ekama, G.A., Brdjanovic, D. (Eds.), Biological Wastewater
Treatment: Principles, Design and Modelling. London, IWA Publishing, London.
Firle STW raw sewage is generally in the low to medium range Henze, M., Comeau, Y., 2008. Wastewater characterisation. In: Hence, M., van
for most parameters it can be concluded that the sewage is rela- Loosdrecht, M.C.M., Ekama, G.A., Brdjanovic, D. (Eds.), Biological Wastewater
Treatment: Principles, Design and Modelling. IWA Publishing, London.
tively easy to treat by biological processes. The PSTs are not per- Japan International Cooperation Agency, 1996. The study on water pollution control
forming as per design specifications hence are not optimally in the upper manyame river basin Main report. Ministry of Local Government,
performing. From the treatability ratios it can be concluded that Rural and Urban Development of the Republic of Zimbabwe, Harare
(unpublished).
treatability of the sewage by Firle Works Unit 4 BNR is highly fea-
Mamais, D., Jenkins, D., Pitt, P., 1993. A rapid physical–chemical method for the
sible for nitrogen and is likely to be complex for phosphorous. determination of readily biodegradable soluble COD in municipal wastewater.
Water Research 27.
Metcalf and Eddy Inc, 2003. Metcalf and Eddy: Wastewater Engineering: Treatment
6. Recommendations and Reuse. McGraw-Hill Companies, Inc., London.
Nhapi, I., 2004. Options for wastewater management in Harare, Zimbabwe,
Dissertation Submitted in Fulfilment of the requirements of the Academic
Fractionation of COD, TP and TN is recommended to clearly de- Board of Wageningen University and the Academic Board of the UNESCO-IHE
fine characteristics of the sewage and appropriately advise further Institute for Water Education, Delft, The Netherlands.
steps to optimise plant processes. Safege, 2003. Harare Sewerage Master Plan Final Report. Harare, City of Harare.
Sato, T., Qadirb, M., Yamamotoe, S., Endoe, T., Zahoora, A., 2013. Global, regional,
and country level need for data on wastewater generation, treatment, and use.
Acknowledgements Agricultural Waste Management 130, 1–13.
Stewart Scott International, 1984. Firle Sewage Works Unit 4 MAS Plant Operating
and Maintenance Instructions (unpublished).
This paper presents part of the research results of an MPhil Thornton, J., Steel, A., Rast, W. 1996. Water Quality Assessments – A Guide to Use of
study by Simon Takawira Muserere at the University of Zimbabwe. Biota, Sediments and Water in Environmental Monitoring, second ed., South
Africa.
The Department of Civil Engineering at the University of Zimbabwe Tjandraatmadia, G., Pollard, C., Gozukara, Y., Sheedy, C., 2009. Characterisation of
are acknowledged for guidance to the production of this paper. The priority contaminants in residential wastewater, CSIRO Australia.
authors acknowledge Harare municipality for assisting in data col- van Loosdrecht, M.C.M., Ekama, G.A., Wentzel, M.C., Brdjanovic, D., Hoolijmans,
C.M., 2008. Modelling Activated Sludge Processes. In: Henze, M., van
lection and for granting permission to carry out the study. The Loosdrecht, M.C.M., Ekama, G.A., Brdjanovic, D. (Eds.), Biological Wastewater
authors thank Mrs. E. Mukanganya of the Surveyor General Depart- Treatment: Principles, Modelling and Design. IWA Publishing, London.
ment for assisting in the drawing of the map of the location of the Water Environmental Research Foundation, 2003. Methods for wastewater
characterisation in activated sludge modelling. IWA. Water Environmental
study area. Finally the authors are grateful to Zimlab a SANAS
Federation, New York.
accredited laboratory, who assisted with sample analyses. The Zimbabwe Statistical Agency, 2012. Preliminary Report Census 2012. Harare
authors acknowledge 13th WaterNet Symposium were this paper Zimbabwe, Government of Zimbabwe.