The implementation and evaluation of biological nitrification as a possible treatment option for the small-scale drinking water supply of a rural Upper Austrian community was investigated. The drinking water supply of this community (average system input volume: 20 m3/d) is based on the use of deep anaerobic groundwater with a high ammonium content of geogenic origin (up to 5 mg/l) which must be treated to prevent the formation of nitrites in the drinking water supply system. This paper describes the implementation and operation of biological nitrification despite several constraints including space availability, location and financial and manpower resources. A pilot drinking water treatment plant, including biological nitrification implemented in sand filters, was designed and constructed for a maximum treatment capacity of 1.2 m3/h. Online monitoring of selected physicochemical parameters has provided continuous treatment performance data. Treatment performance of the plant was evaluated under standard operation as well as in the case of selected malfunction events.

INTRODUCTION

The presence of ammonium in water resources used for drinking water purposes can be problematic as ammonium can compromise disinfection efficiency, cause the failure of filters for the removal of manganese, cause taste and odour problems, and biochemically build up to nitrite and nitrate (WHO 2011). High nitrite and nitrate concentrations are not desired in drinking water because of the risk of methaemoglobinaemia, especially for infants and elderly people. Natural ammonium levels in groundwater and surface water are usually below 0.2 mg/l but high concentrations of up to 3 mg/l can be measured in anaerobic groundwater (WHO 2011). The WHO drinking water quality guideline values for nitrite and nitrate are 3 mg/l and 50 mg/l, respectively. There are no health-based guideline values for ammonium.

In Austria, ammonium is used as an indicator function parameter with a guideline value of 0.5 mg/l. However, concentrations up to 5 mg/l are tolerated where ammonium results from natural conditions. The maximum admissible values for nitrite and nitrate in drinking water are 0.1 mg/l and 50 mg/l, respectively (BMG 2015).

Many rural communities in the northern part of the Federal State of Upper-Austria, Austria use deep anaerobic groundwater as their sole source for drinking water supply. In this region, groundwater often shows low nitrate and nitrite concentrations but high ammonium content (up to 5 mg/l) of geogenic origin. Its use for drinking water purposes, without any treatment, is especially an issue when the water is supplied through a drinking water supply system including a storage tank. In a drinking water storage tank, the anaerobic groundwater will come in contact with air, which can, under certain conditions, trigger uncontrolled and incomplete biological nitrification processes directly within the water supply system. This results in lower ammonium but higher nitrate and nitrite concentrations in the supplied water. In some cases, the supplied water may not fulfil local drinking water quality requirements any more.

It is this particular problem that a community (named ‘A’ in this paper) of this region had to face in 2006. The drinking water supply of this community (193 inhabitants in 2013; average water consumption: 20 m3/d) relies on a 230 m deep groundwater source with an ammonium content reaching up to 5 mg/l. Prior to 2006, water was pumped from the borehole directly into the supply system. Following local population growth, a drinking water storage tank with two chambers of 30 m3 was built and put into operation in 2006 to cope with the increasing water demand. However, in the same year, nitrite concentrations of up to 0.56 mg/l were measured in water samples collected from the supply systems indicating that biological nitrification occurred uncontrolled and incompletely, directly within the supply system. As a first emergency measure, it was decided to use only one chamber of the storage tank and to overlay the stored water by a floating polyethylene (PE) cover to prevent oxygen dissolution into the water.

Uncontrolled and incomplete biological nitrification in drinking water supply systems is an issue for all communities but especially for small rural communities, which lack alternative water resources and financial and human resources. In the case of the A community, a feasibility study has shown that the most economical solution was to set-up a treatment plant. Connections to larger scale drinking water supply schemes of neighbouring communities were not possible due to the high costs required to build long distance water pipelines.

To deal with this issue faced by other communities in the region, a project supported by the Austrian Federal Ministry of Agriculture, Forestry, Environment and Water Management and the Federal State of Upper-Austria was carried out by the University of Natural Resources and Life Sciences, Vienna to evaluate the practicability of biological nitrification for the drinking water treatment of small-scale water supply systems. The project was carried out from May 2013 to March 2014 and included the planning, implementation, operation and optimisation of a pilot drinking water treatment plant for the A community.

This paper describes the operation of the pilot drinking water treatment plant throughout and after the project (until April 2015) and highlights treatment performances during standard operation, following backwashing events and an interruption of the aeration process. The available literature provides information about the operation of larger drinking water treatment plants with biological nitrification (e.g. Kihn et al. 2002; Van der Aa et al. 2002; Lee et al. 2014), but information is scarce for small-scale plants.

THE PILOT DRINKING WATER TREATMENT PLANT

A schematic of the pilot plant is provided in Figure 1. Treatment is based on four successive steps: (1) aeration (oxidator), (2) biological nitrification (filter F1-1 and F1-2), (3) rapid sand filtration (filter F2) and (4) UV-disinfection.
Figure 1

Schematic of the pilot treatment plant.

Figure 1

Schematic of the pilot treatment plant.

The pilot plant was designed for a maximum treatment capacity of 1.2 m3/h. Due to space constraints, the treatment units were installed inside a 10-foot container.

The oxidator has a volume and a main cross-section of 15 l and 0.01 m2, respectively. It is made of polypropylene and equipped with air escape valves to release the air surplus. A 15 cm long stainless steel air diffuser located at the bottom of the oxidator enables the diffusion of air bubbles. The average pore size of the air diffuser is 20 μm. Air is pumped through an oil free compressor into an 80 l pressure tank, filtered through a 0.01 μm particle filter and then injected through the diffuser. Raw water is aerated as water flows top-down through the oxidator.

Following aeration, the water undergoes biological nitrification in two identical filters (F1-1 and F1-2) run in parallel. These filters have PE vessels with a volume of 436 l and an inner diameter of 0.6 m. The filters are filled with two types of quartz sand (average particle sizes of 3.15–5.6 and 1–2 mm with respective volumes of 40 and 300 l). Rapid sand filtration occurs within F2. The volume and inner diameter of its PE pressure vessel are 250 l and 47 cm, respectfully. F2 is also filled with two types of quartz sand (average particle sizes of 3.15–5.6 and 1–2 mm with equal volumes of 100 l). None of the sand filters were inoculated with nitrifying biomass.

The last treatment step is UV-disinfection. The UV unit operates with one 55 W low pressure mercury vapour lamp and is designed for a maximum treatment capacity of 1.8 m3/h. The unit is certified by the Austrian Association for Gas and Water for drinking water treatment. The piping system is made of PE-X pipes authorized for drinking water purposes and with an inner diameter of 17.7 mm. A pressure sustaining valve located on the outflow of the UV unit enables regulation of the plant inner pressure.

MATERIAL AND METHODS

Water samples were collected and analysed either on-site or in a laboratory. The physicochemical and microbiological parameters investigated are given in Table 1. Ammonium, nitrite, and nitrate were either measured onsite with cuvette tests or in the laboratory using photometry and ion chromatography. UV/VIS spectrometer probes as well as optical sensors were used for online measurements of NO3-N and dissolved oxygen (Figures 5 and 6). The sensors were set to form a unique measuring track enabling successive analyses of water quality after each treatment step.

Table 1

Physicochemical and microbiological parameters investigated

ParametersMethods
NH4-N [mg/l] Photometry, sodium nitroprusside (DIN 38406 T5 (E5))/ 
NH4-N [mg/l] Hach LCK304 or 303 
NO2-N [mg/l] Ion chromatography (EN ISO 10304 T1 T2) 
NO2-N [mg/l] Hach LCK341 or 342 
NO3-N [mg/l] Ion chromatography (EN ISO 10304 T1 T2) 
NO3-N [mg/l] Hach LCK339 
pH Potentiometry, glass electrode (DIN 38404 T5 (C5)) 
EC [μS/cm] Conductometry (DIN EN 27888 (C8)) 
Temperature [°C] Special thermometer (DIN 38404 T4 (C4) 2) 
Carbonate Hardness (CH) [°dH] Titration pH 4.3/8.2 (DIN 38409 T7 (H7)) 
Phosphate [mg/l] Photometry, ammonium molybdate (DIN EN ISO 6878 (D11) 6) 
Fe dissolved [μg/l] ICP-MS (DIN EN ISO 17294-1 (E29)) 
Mn dissolved [μg/l] ICP-MS (DIN EN ISO 17294-1 (E29)) 
Turbidity [NTU] Nephelometry/formazin-Lsg. (DIN EN ISO 7027 (C2)) 
TOC [mg/l] UV-oxidation NDIR detection (DIN EN 1484 (H3)) 
DOC [mg/l] UV-oxidation NDIR detection (DIN EN 1484 (H3)) 
E. coli and coliform bacteria MPN, 100 ml, Colilert®-18(C-18), Quanti-Tray™ (IDEXX) 
E. coli and coliform bacteria Membrane filtration, 100/250 ml, Chromocult-Agar enhanced selectivity (CC) 
Enterococcus Membrane filtration, 100/250 ml, Slanetz Bartle-Agar (SB) (EN ISO 7899-2) 
P. aeruginosa Membrane filtration, 100/250 ml, Cetrimid-agar (Cet) 24/48 h (ÖNORM EN 12780) 
Colony count 22 °C and 37 °C EN ISO 6222 
ParametersMethods
NH4-N [mg/l] Photometry, sodium nitroprusside (DIN 38406 T5 (E5))/ 
NH4-N [mg/l] Hach LCK304 or 303 
NO2-N [mg/l] Ion chromatography (EN ISO 10304 T1 T2) 
NO2-N [mg/l] Hach LCK341 or 342 
NO3-N [mg/l] Ion chromatography (EN ISO 10304 T1 T2) 
NO3-N [mg/l] Hach LCK339 
pH Potentiometry, glass electrode (DIN 38404 T5 (C5)) 
EC [μS/cm] Conductometry (DIN EN 27888 (C8)) 
Temperature [°C] Special thermometer (DIN 38404 T4 (C4) 2) 
Carbonate Hardness (CH) [°dH] Titration pH 4.3/8.2 (DIN 38409 T7 (H7)) 
Phosphate [mg/l] Photometry, ammonium molybdate (DIN EN ISO 6878 (D11) 6) 
Fe dissolved [μg/l] ICP-MS (DIN EN ISO 17294-1 (E29)) 
Mn dissolved [μg/l] ICP-MS (DIN EN ISO 17294-1 (E29)) 
Turbidity [NTU] Nephelometry/formazin-Lsg. (DIN EN ISO 7027 (C2)) 
TOC [mg/l] UV-oxidation NDIR detection (DIN EN 1484 (H3)) 
DOC [mg/l] UV-oxidation NDIR detection (DIN EN 1484 (H3)) 
E. coli and coliform bacteria MPN, 100 ml, Colilert®-18(C-18), Quanti-Tray™ (IDEXX) 
E. coli and coliform bacteria Membrane filtration, 100/250 ml, Chromocult-Agar enhanced selectivity (CC) 
Enterococcus Membrane filtration, 100/250 ml, Slanetz Bartle-Agar (SB) (EN ISO 7899-2) 
P. aeruginosa Membrane filtration, 100/250 ml, Cetrimid-agar (Cet) 24/48 h (ÖNORM EN 12780) 
Colony count 22 °C and 37 °C EN ISO 6222 

Operational data including treated water volumes, pressures and flow rates were recorded based on manual readings of water meters, flow meters, and manometers located up- and downstream of each treatment unit. The aeration rate was measured by a flow meter located upstream of the air diffuser.

During the statistical analyses of physicochemical results, values below limits of detection (LOD) were handled as values equal to LOD/2. This applies for NO3-N (LOD: 0.23 mg/l for the cuvette tests otherwise, 0.1 mg/l), NO2-N (LOD: 0.015 mg/l for the cuvette tests, otherwise 0.03 mg/l), dissolved Fe (LOD: 5 μg/l), and dissolved Mn (LOD: 0.5 μg/l).

Data were analysed with the open source, BSD-licensed python library Pandas and figures were made by using the matplotlib based Python visualisation library Seaborn.

The following nomenclature was used to describe the results: RW (raw water), RWO2 (oxidator effluent), F1-1 (F1-1 effluent), F1-2 (F1-2 effluent), F2 (F2 effluent), UV (UV effluent).

RESULTS

Operation of the pilot drinking water treatment plant

Figure 2 shows the evolution of treatment flow rates, pressures and aeration rates between July 2013 and May 2015. Operation of the pilot drinking water treatment plant started in June 2013. Throughout the whole monitoring period, treatment flow rates between 0.6 and 1 m3/h were applied. The effluent of the pilot plant was not used for drinking water supply but directed toward the communal waste water system. The variations of operational values between July 2013 and March 2014 are explained by several configuration changes of the pilot plant. The vertical lines displayed on Figure 2 indicate three main changes: installation and commissioning of the second nitrification filter F1-2 (29/10/2013), installation and commissioning of the oxidator and pressure sustaining valve on the outlet of the UV unit (14/3/2014), and repair of F1-1 (2/7/2014).
Figure 2

Operation of the pilot drinking water treatment plant between July 2013 and May 2015.

Figure 2

Operation of the pilot drinking water treatment plant between July 2013 and May 2015.

The pilot plant as described previously in this paper was only in operation as such starting from March 14th, 2014. Prior to that date, there was no oxidator and air was directly injected into the raw water inlet under the assumption that turbulent flow in the water pipelines would enable the dissolution of air into the water. However, complete dissolution upstream of the nitrification filters could not be achieved and therefore an oxidator was later installed. During this period, pressures up to 2.2 bar in the plant inlet are assumed to have been caused by the accumulation of air inside the filters. From March 2014, the aeration rate was increased from 5 to 9 l/min to guarantee complete nitrification and a minimum O2 concentration of 2 mg/l in the plant effluent. Operation of the oxidator has shown optimisation potential as unknown quantities of air are constantly exhausted through the oxidator air escape valves. The pressure sustaining valve located on the outlet of the UV-unit was set to 1 bar in order to provide sufficient pressure inside the plant to reach the dissolved O2 level (at least 18 mg/l) required to achieve complete nitrification. Since mid-July 2014, each filter was backwashed once a month (F1-1 and F1-2 on the same day; F2 alone two weeks later). In July 2014, a defect on the outlet pipe of F1-1 was identified and fixed. This defect would cause some of the water flowing toward F1-1 to by-pass the filtration material and flow directly toward F2. The defect was suspected long before its identification onsite due to the poor treatment performance of F1-1 (see further below).

Treatment performance of the pilot plant (physicochemical quality)

Table 2 provides statistics on the physicochemical quality of the raw water and F2 effluent during stable operation. These results encompass the analyses carried out in 2014, 2015 and the results of selected analyses carried out in 2013. Results of the raw water quality analyses indicate favourable and stable conditions for biological nitrification especially with regard to water temperature, pH, phosphate content, and carbonate hardness (CH). Complete nitrification was achieved after treatment with a full conversion of ammonium into nitrate and decreases for pH, electrical conductivity (EC), CH and phosphate. The treatment also enables a slight reduction of turbidity and the removal of iron and manganese. Total organic carbon (TOC) and dissolved organic carbon (DOC) contents remained constant around 0.3 mg/l. Dissolved oxygen values given in Table 2 for the raw water are assumed to be higher than real values due to unwanted dissolution of oxygen during sampling. Online measurements carried out in 2013 directly within the system have indicated more realistic stable values below 1 mg/l.

Table 2

Physicochemical quality of raw water (RW) and the plant effluent (F2) during standard operation

 StatisticNH4-N [mg/l]NO2-N [mg/l]NO3-N [mg/l]pHEC [μS/cm]T [°C]CH [°dH]P [mg/l]Fe [μg/l]Mn [μg/l]Turb. [NTU]O2 [mg/l]
RW Min 3.39 0.003 0.05 8.2 323 12.5 10.5 0.06 5.7 5.30 0.19 0.7 
Median 3.49 0.008 0.12 8.4 358 16.5 10.6 0.07 22.1 6.25 0.23 1.5 
Max 4.20 0.017 0.23 8.8 384 18.9 10.6 0.08 25.8 15.40 0.29 2.7 
Count 16 15 15 13 13 13 11 
F2 Min 0.01 0.002 3.20 6.9 306 15.2 9.1 0.05 2.5 0.25 0.11 0.2 
Median 0.02 0.015 3.50 7.2 339 17.0 9.1 0.05 2.5 0.50 0.13 3.2 
Max 0.11 0.025 3.81 7.5 342 19.3 9.3 0.05 11.2 0.50 0.13 7.1 
Count 29 29 29 25 12 24 23 
 StatisticNH4-N [mg/l]NO2-N [mg/l]NO3-N [mg/l]pHEC [μS/cm]T [°C]CH [°dH]P [mg/l]Fe [μg/l]Mn [μg/l]Turb. [NTU]O2 [mg/l]
RW Min 3.39 0.003 0.05 8.2 323 12.5 10.5 0.06 5.7 5.30 0.19 0.7 
Median 3.49 0.008 0.12 8.4 358 16.5 10.6 0.07 22.1 6.25 0.23 1.5 
Max 4.20 0.017 0.23 8.8 384 18.9 10.6 0.08 25.8 15.40 0.29 2.7 
Count 16 15 15 13 13 13 11 
F2 Min 0.01 0.002 3.20 6.9 306 15.2 9.1 0.05 2.5 0.25 0.11 0.2 
Median 0.02 0.015 3.50 7.2 339 17.0 9.1 0.05 2.5 0.50 0.13 3.2 
Max 0.11 0.025 3.81 7.5 342 19.3 9.3 0.05 11.2 0.50 0.13 7.1 
Count 29 29 29 25 12 24 23 

Figure 3 shows the time evolution of NH4-N, NO2-N and NO3-N concentrations at different stages of the treatment process as well as interruption periods (e.g. no aeration or no flow; depicted as shaded areas). It took approximatively 14 weeks, without any inoculation of the filters with external nitrifying biomass, to measure complete nitrification in F2 effluent (mid. September 2013). High NO3-N concentrations were already measured in August but together with NO2-N concentrations above 0.2 mg/l. It is very likely that the F2 start-up could have been lowered with an optimized aeration process and stable operating conditions. Only 14 days were necessary for F1-2 to achieve complete nitrification (also see Figure 6). F1-2 was commissioned on the 29th of October 2013 but a defect of the aeration process enabled effective aeration only after the 28th of November 2013. The faster start-up of F1-2 is very likely due to the presence of nitrifying biomass in the upstream pipes which was not the case for F2. The results for F1-1 are not relevant because of the by-pass defect previously described.
Figure 3

Treatment performance (NH4-N, NO2-N and NO3-N concentrations).

Figure 3

Treatment performance (NH4-N, NO2-N and NO3-N concentrations).

The ammonium removal rate (ARR) of the filters was calculated as defined by Lee et al. (2014): ARR=Q*(Cin–Cout)/(A*Δz) with Q: flow rate [m3/h]; Cin and Cout: NH4-N inlet and outlet concentrations [g/m3], respectively; A: cross sectional area of the filter [m2]; and Δz: active depth of the filter bed [m]. Based on RW- and F2 measurements, ARR was calculated in the range of 1.5–4 g NH4-N/h.m3 depending on flow rate. However, these values are higher than real ARR because nitrification already starts in the oxidator as revealed by three measurements carried out during the project (Table 3). The ammonium removal capacity (ARC), i.e. the maximum possible ARR of the filters could not be estimated because of insufficient pump capacity in the plant inlet. Future works are required to characterize the system ARC. The ARR calculated appears to be in line with the results obtained by other researchers at plants treating groundwater: 3.2 (Štembal et al. 2004), 3.4 (Lee et al. 2014), between 1.5 and 5 (de Vet et al. 2011), up to 6.9 g NH4–N/h.m3 (Tatari et al. 2013).

Table 3

Ammonium, nitrite and nitrate measurements of raw water (RW) and the oxidator effluent

Date NH4-N [mg/l]NO2-N [mg/l]NO3-N [mg/l]Flow rate [m3/h]
26/11/2014 RW 3.56 0.017 0.01 0.46 
Oxidator 2.99 0.251 0.46 0.46 
17/02/2015 RW 3.7 0.015 0.012 0.41 
Oxidator 3.14 0.286 0.49 0.41 
23/04/2015 RW 3.2 0.003 0.05 0.36 
Oxidator 2.7 0.240 0.30 0.36 
Date NH4-N [mg/l]NO2-N [mg/l]NO3-N [mg/l]Flow rate [m3/h]
26/11/2014 RW 3.56 0.017 0.01 0.46 
Oxidator 2.99 0.251 0.46 0.46 
17/02/2015 RW 3.7 0.015 0.012 0.41 
Oxidator 3.14 0.286 0.49 0.41 
23/04/2015 RW 3.2 0.003 0.05 0.36 
Oxidator 2.7 0.240 0.30 0.36 

The treatment plant should be optimized by adding a neutralization step before UV disinfection because the calcite dissolution capacity (CDC) increases during nitrification. CDC calculation, according to the German standard DIN 38404-10 and using the PHREEQC-based program developed by de Moel et al. (2013), indicates CDC values of −0.6 mg/l and 50 mg/l, respectively in the inflow and outflow of the pilot treatment plant. According to the Austrian drinking water standards, drinking water should not be corrosive, which in practice requires a CDC below 5 mg/l (depending on pipe material and water quality, calculation of other corrosion parameters might be required). The raw water is very soft with Ca2+ and Mg2+ concentrations of 7 mg/l and 2 mg/l, respectively and the pH decrease occurring during the nitrification process leads to an increase of aggressive CO2 content and therefore to high CDC values in the outflow of F2. To prevent water corrosiveness, a neutralization step to adjust pH (CO2 stripping or limestone filtration) should be implemented before disinfection. This additional step might not have been needed if dry filtration followed by rapid sand filtration had been implemented instead of pressure filtration, as the process enables the release of excessive CO2. However, due to the high ammonium concentration in the raw water, dry filtration would have required at least two filtration stages and thus more space, which was not available for this project.

Treatment performance of the pilot plant (microbiological quality)

Figure 4 shows the microbial treatment performance. In 2013, values for colony counts (22 °C) and coliform bacteria above the thresholds defined in the local drinking water legislation were reported at each measuring point of the system including in the raw water but not in the effluent of the UV unit. The high values are suspected to be the consequence of the community installing a new and stronger pump in their borehole in June 2013. It took almost 6 months for no coliform bacteria to be detected in the raw water and a year for the colony counts at 22 °C. The presence of Escherichia coli, coliform bacteria and enterococci could not be detected in the samples collected in 2014 and 2015. Values for colony counts (37 °C) were in most cases below 20 CFU/100 ml except during the sampling in February 2015 (possibly due to contamination during sampling). Flow cytometry analyses showed an increase of the total cell counts during the treatment process from an average of 8,544 cells/ml in the raw water to 101,982 cells/ml in the effluent of F2 (averages based on the results of five analyses). Out of 41 analyses, Pseudomonas aeruginosa presence was detected three times: two times in F2 effluent (15 CFU/250 ml in June 2014 and 5 CFU/100 ml in April 2015). Pseudomonas aeruginosa was not detected in the water samples collected from the outlet of the UV unit.
Figure 4

Treatment performance (coliform bacteria, colony counts 22 °C/37 °C, Pseudomonas aeruginosa, total cell count).

Figure 4

Treatment performance (coliform bacteria, colony counts 22 °C/37 °C, Pseudomonas aeruginosa, total cell count).

Impact of backwashing on filter performance

Filters were backwashed on a monthly basis during a three-step cycle using air and water: 3 min with air (50 m3/m2.h), followed by 3 min with a mixture of air and water (15 m3/m2.h) and finally with 250 l of water (15 m3/m2.h). Figure 5 shows the impact of backwashing on nitrate nitrogen concentration in the filter outflows for two backwashing events. High nitrate nitrogen concentrations in the treatment plant outflow were reached within 24 h after backwashing all filters on the same day (first case) and maximum nitrate nitrogen concentrations were measured in the plant outflow again after 3 days. In the second case, only F1-1 was backwashed, leading to a drop of NO3-N concentrations from 3.7 to 2.7 mg in the effluent of F1-1. Maximum concentrations were also reached 3 days later.
Figure 5

Evolution of nitrate nitrogen concentrations in the filter outflows before and after filter backwashing.

Figure 5

Evolution of nitrate nitrogen concentrations in the filter outflows before and after filter backwashing.

Performance of the treatment plant following an interruption of the aeration process

The aeration process was interrupted for 37 days before being applied again on November 28th (Figure 6). Despite this interruption, a rapid recovery of biological activity was observed and nitrate nitrogen concentrations of 3 mg/l could be measured in the outflow of the treatment plant less than 24 hours following the recommissioning of the aeration process. Figure 6 also shows lower NO3-N concentrations in the outflow of F1-(1), which are likely the result of the defect in the filter outlet identified in July 2014. On the 6th of December 2013, the O2-sensor was moved and installed on a bypass pipe directly between the air injection point on the plant inlet and F1-(1) and F1-(2) in order to better monitor and regulate the aeration process (RWO2). The fluctuations of O2-values are supposedly due to the presence of air bubbles in the water; a problem that was solved with the implementation of the oxidator in March 2014. The steady decrease of O2-values is due to the progressive accumulation of biofilm on the dissolved oxygen sensor.
Figure 6

Performance of the treatment plant following a 37 day interruption of the aeration process.

Figure 6

Performance of the treatment plant following a 37 day interruption of the aeration process.

CONCLUSIONS

Biological nitrification was successfully implemented at the project site and results show good treatment performance under standard operating conditions and also following interruptions. However, with respect to this project, a neutralization step should be implemented before disinfection to reduce water corrosiveness resulting from the nitrification process. Filter start-ups were estimated to be between 2 and 14 weeks depending on the filters. Filter inoculation was not necessary. Biological nitrification is a viable option for rural communities considering operational performance and low maintenance. The potential for design optimisation remains, especially with regard to the aeration process. Further work is also required to estimate the plant ARC.

REFERENCES

REFERENCES
BMG
2015
Österreichiches Lebensmittelbuch; IV. Auflage; Codex – Kapitel, B1, Trinkwasser. Bundesministerium für Gesundheit, Wien
.
de Moel
P. J.
van der Helm
A. W. C.
van Rijn
M.
van Dijk
J. C.
van der Meer
W. G. J.
2013
Assessment of calculation methods for calcium carbonate saturation in drinking water for DIN 38404-10 compliance
.
Drinking Water Engineering and Science
6
,
115
124
.
de Vet
W. W. J. M.
Kleerebezem
R.
van der Wielen
P. W. J. J.
Rietveld
L. C.
van Loosdrecht
M. C. M.
2011
Assessment of nitrification in groundwater filters for drinking water production by qPCR and activity measurement
.
Water Research
45
,
4008
4018
.
Kihn
A.
Andersson
A.
Laurent
P.
Servais
P.
Prevost
M.
2002
Impact of filtration material on nitrification in biological filters used in drinking water production
.
Journal of Water Supply: Research and Technology – AQUA
51
,
35
46
.
Lee
C. O.
Boe-Hansen
R.
Musovic
S.
Smets
B.
Albrechtsen
H.-J.
Binning
P.
2014
Effects of dynamic operating conditions on nitrification in biological rapid sand filters for drinking water treatment
.
Water Research
64
,
226
236
.
Štembal
T.
Markić
M.
Briški
F.
Sipos
L.
2004
Rapid start-up of biofilters for removal of ammonium, iron and manganese from ground water
.
Journal of Water Supply: Research and Technology – AQUA
53
(
7
),
509
518
.
Van der Aa
L. T. J.
Kors
L. J.
Wind
A. P. M.
Hofman
J. A. M. H.
Rietveld
L. C.
2002
Nitrification in rapid sand filter: phosphate limitation at low temperature
.
Water Science and Technology: Water Supply
2
(
1
),
37
46
.
WHO
2011
Guidelines for Drinking-Water Quality
, 4th edn.
World Health Organization, Geneva
,
Switzerland
.