Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

THE PERFORMANCE EVALUATION OF MEMBRANE ULTRAFILTRATION

AND DIRECT UPFLOW RAPID SAND FILTRATION IN TREATING LOW


TURBIDITY RAW WATER FOR POTABLE USE

B Tenza, N Toolsee, R Rajagopaul, P Thompson, L Sibiya

Umgeni Water, PO Box 9, Pietermaritzburg, 3200; Tel: 031 2033025; Fax: 031 2617202;
Email: Naheen.Toolsee@umgeni.co.za ; Blessing.Tenza@umgeni.co.za

Umgeni water catchment systems have experienced a drop in raw water turbidity for the past
decade. The low raw water turbidity (< 5 NTU) posed operational and performance problems
for the conventional water treatment process at one of the Umgeni Waterworks. As a result
direct up flow filtration (DUF) and membrane ultrafiltration (UF) were explored on pilot plant
scale. The performance UF and DUF was compared with that of a conventional water
treatment evaluation unit (CWTEU).

UF produced better quality water, in terms of turbidity, than the CWTEU. The DUF final water
quality was lower than that of the CWTEU when an organic polymeric coagulant was used.
With alum as coagulant both the DUF and the CWTEU produced final water of similar higher
water quality.

In general, the DUF and the WTEU had similar filter run times. Under similar raw water
conditions, the filtration cycle for both was 24 hours, similar to what is expected from a full
scale rapid gravity filter. The run time for the UF pilot plant was significantly longer, typically
about 20 days before the onset of the ‘chemical in place, CIP, membrane cleaning.
Water recovery for both the DUF and the CWTEU were similar at between 95 and 96%, while
for the UF water recovery was relatively lower at 92%.

Polymeric organic coagulant requirement for the DUP was the same as that for the
conventional water treatment process. However, the DUP required significantly lower doses
of alum compared to the conventional water treatment plant. UF required no pretreatment.

Both the CWTEU and the UF pilot plants had similar energy requirements while the DUF pilot
had a significantly lower energy requirement relative to the other two processes.
INTRODUCTION
Umgeni water has experienced a significant reduction in raw water turbidity at most of its
waterworks for the past decade (Toolsee et al, 2012). At the Wiggins Waterworks site, raw
water turbidity, < 5 NTU were recorded. Operational and performance challenges were
experienced in some of the larger waterworks like Wiggins Waterworks that use conventional
water treatment processes. Conventional clarifiers are designed to treat relatively high
turbidity water; and the high suspended solids in the influent facilitate settling after
coagulation and flocculation (Foley, 1980).
As a short to medium term solution, bentonite is dosed to improve coagulation and
flocculation (Toolsee et al, 2012). This has the disadvantage of higher chemical costs and
sludge volumes.

For a more sustainable and long-term solution, other treatment options were investigated.
Desktop studies indicated that ultrafiltration and direct filtration were viable options for the
treatment of low turbidity waters. These two technologies were investigated and compared
with conventional water treatment processes on a pilot plant scale.

While direct filtration offers a low cost solution for low turbidity raw water, its application in
South Africa has been restricted since surface water turbidities can exceed 80 NTU, which is
considered high for this process (WRC Report No. 354/1/97).

Direct filtration methods (two stage system including coagulation and direct filtration) have
been previously explored in the developed countries. A recent development in this area has
been the evaluation of a three stage treatment system, consisting of coagulation, floc
conditioning and direct filtration. The system has already been applied in America for removal
of turbidity, colour, algae, as well as organisms (Reid and Loewenthal, 1989). Direct filtration
processes can be operated at rates up to three times as high compared to conventional water
treatment processes for comparable water quality (Westerhoff et al. 1980).

Membrane-based treatment systems have gained importance for drinking water treatment in
the developed countries (Maryna Peter-Varbanets, 2007).

During the last decade UF technology has become more efficient and the costs of
membranes have decreased significantly (Churchhouse, 2000). Ultrafiltration (UF) provides
an effective barrier for microorganisms, suspended particles and colloids and is increasingly
implemented for the treatment of drinking water. Membrane fouling was one of the major
limitations of membrane technology (Ramesh et al., 2007).

The pore size of ultrafiltration (UF) membranes is small enough to ensure high log-removal of
most microbiological pollutants such as Cryptosporidium, Giardia and total bacterial counts
(Hagen, 1998).
Locally (in South Africa), there was active research in low pressure ultrafiltration membrane
systems (Jacobs et al., 2000, Pillay et al, 2003, Pryor et al., 1998). Research findings showed
that ultrafiltration is an attractive and viable treatment technology for potable water treatment.

OBJECTIVE

The main objectives were the comparison of the performance of direct filtration and
ultrafiltration with that of a conventional water treatment process using the following criteria:
 Final water quality
 Water recovery
 Chemical use
 Energy costs

METHODOLOGY

Trials were conducted at Umgeni Water Wiggins Process Evaluation Facility (PEF) using an
Ultrafiltration (UF, Direct up flow filter (DUF) and a conventional water treatment (CWTEU)
pilot plant as a control. A summary of the process descriptions and operating parameters is
presented on Table 1.

Raw water was from an impounded (Inanda dam) water source, characterized by relatively
low turbidity (< 5NTU) and average pH between 6.5 and 7. The raw water was also fairly
clean, with relatively low organic content (<4 mg/l TOC) and colour <10 hazen units. Inanda
dam water is delivered to the Wiggins Process Evaluation Facility, PEF, about 40 km from the
dam, through a network of tunnels and pipes.

Pilot Plant Operation


Each process was operated at its recommended operating parameters. Where applicable,
chemical doses were optimized based on final water turbidity conforming to UW potable water
internal standards (0.5NTU).
The pilot plants were operated in parallel and in continuous mode. In most cases dataloggers
were used to log data on a continuous basis. The CWTEU pilot plant and the UF pilot plant
were both equipped with a PLC and SCADA for control and datalogging respectively. On-line
analytical instruments were installed on the pilot plants to measure pH, turbidity and filter
headloss (pressure transducers).
Table 1: Summary of process description and operating parameters

Process Conventional Treatment Pilot plant Ultrafiltration Pilot Plant (UF) Direct Up flow filter (DUF)
Description
Process Disinfectant
Tank
Flow Raw Pre-treatment
Backwash
Pump To Permeate
Tank
Water Chemicals
Diagram Vacuum
From Permeate
Tank Coagulant Filtrate Tank
Vent Reject/ Drain
Tank

DUF Filter
Flocculation
Balancing

Pulsation Chamber
Coagulant Dosing

Vessel
Tank Pump UF
tank Raw Water
Filter Tank
Chemical storage
Back Tanks Backwash
Pulsator B/W
wash out Static Pump
Clarifier Drain
Compressed Feed Mixer
Air Pump Treated
Flash Drain Water
mixer

Air in
Back wash

Water in
Filtrate From Raw water drain

CIP Tank
Sludge Tank Air Blower
CIP Filter
Feed
Pump
CIP Pump

Process Pre-treatment, Coagulation, flocculation, Pre-treatment (pre-filter), filtration. Coagulation, floc conditioning, filtration
Train sedimentation, filtration

Flowrate, 4.5 4 4.5


m3/h
Clarifier/ Clarifier up flow rate, 2.5m/h Constant flux, 40 LMH Filtration rate,12m/h
filter rates Filtration rate, 3.8 – 4.3 m/h Variable TMP to max of 1.5 bar,
dead-end mode filtration
Treatment 4 mg/l Bentonite Direct UF of raw water: no chemicals
Chemical 2 mg/l Polymeric Organic UF of pretreated clarified water: 2 mg/l Polymeric organic
Coagulant Treatment chemicals as per
conventional treatment pilot plant
Filtration 30 hours before onset of backwash 20 - 40 days with intermittent backflush 24 Hours before onset of backwash
cycle
Filter Backwash with water 12 m/h for 5 min Back-flush for 60 s 80 LMH, 30 min Air scour, 55 m/h for two 8 min cycles
cleaning Air scour 55 m/h for 5 min cycle Air and water, 19 m/h water and 55 m/h air
Water only rinse, 17 m/h for 10 min Cleaning-in place (CIP) after 20 days for 15 min
Water only rinse, 21 m/h for 10 min
RESULTS

The results presented graphically represent the trends obtained from a number of repeat runs
and are typical of the general relationships observed. Turbidity was used as a measure of
water quality.

Final Water Turbidity

Figure 1 shows a comparison of the final water turbidity produced by the three processes. A
polymeric organic coagulant was used for the conventional water treatment plant and the
DUF. There was no pre-treatment for the UF system. Final water turbidities for the three
processes were compliant with SANS 241 turbidity limits for potable water (< 1NTU). While
the UF consistently produced the lowest final water turbidities (< 0.2NTU), the DUF produced
higher final water turbidities relative to the other two processes.

In general, the DUF and the conventional rapid gravity filter had similar filter run times. Under
similar raw water conditions, the filtration cycle for both was 24 hours, similar to what is
expected from a full-scale rapid gravity filter. The run time for the UF pilot plant was
significantly longer, typically about 20 days before the onset of the ‘chemical in place, CIP,
membrane cleaning event.

Figure 1: Comparison of Final Water Turbidity: Organic Polymeric as Coagulant


Figure 2 represents a similar set of trends as presented in Figure 1. The only difference was
that alum was used as coagulant for both the conventional water treatment plant and the DUP
in this case (Figure 2). Figure 2 also indicates that the performance of alum is significantly
better than that of the organic polymeric coagulant.

Figure 2: Comparison of Final Water Turbidity: Alum as Coagulant

Water Recovery

Each process was further evaluated in terms of operational performance (Table 3). Water
recovery was determined for each process, taking into account water losses that occur during
normal operation.
For conventional water treatment processes, filter backwash water make up the largest
component of water ‘losses’ with clarifier desludging contributing the balance. The scenario is
similar for the DUP with a floc conditioner. As indicated in Table 3, water recovery for both
these processes is similar between 95 and 96%.
For the UF water recovery is relatively lower at 92%, with the membrane back flushing event
contributing to about 90% of the water ‘loss’.
Chemical Consumption
Table 3, indicates that conventional treatment processes are likely to use more pretreatment
chemicals, especially coagulants and coagulant aids (bentonite) to produce large settleable
flocs required for efficient solids separation and removal.

Polymeric organic coagulant requirement for the DUP was the same as that for the
conventional water treatment process. This was contrary to expectations that direct filtration
required pin sized flocs and would thus require a lower coagulant dose. However, the DUP
required significantly lower doses of alum compared to the conventional water treatment
plant.

There was no pretreatment chemical requirement for UF treatment as raw water was filtered
directly through the UF membranes.

Energy Requirements

Due to the similar capacities of the three pilot plants and level of automation including the use
of PLC and SCADA that was common to all of them, total power consumption was used as a
comparison. Table 3 indicates that both the conventional water treatment and the UF pilot
plants had similar energy requirements while the DUF pilot had a significantly lower energy
requirement relative to the other two processes.

Table 2: Summary of Operational Performance

Performance Conventional Ultrafiltration Direct up filter (DUF)


Measurement treatment process membrane (UF)
Water Recovery 95% water recovery 92% water recovery 96% water recovery
Water loss: 2% Water loss: 7% Water loss: 3%
backwash water, ~3% membrane back flushing, backwash water, ~1%
clarifier desludging ~1% membrane draining floc conditioner drains

Chemical 4 mg/l Bentonite Cleaning chemicals 2 mg/l Polymeric


Requirements 2 mg/l Polymeric organic (2000 mg Sodium organic coagulant
coagulant hypochlorite/ l of cleaning
solution, 1000 mg
Sodium hydroxide/ l of
cleaning solution)
Total Power 0.65 0.68 0.23
consumption,
kWh/m3
CONCLUSIONS

The following conclusions can be drawn from the process performance comparison of the two
technologies evaluated for potable water treatment applicable to low raw water turbidities.

 All three technologies are capable of producing final turbidity which is within a SANS 241-
1:2011 standard.
 UF produced better quality final water in terms of turbidity.
 UF membrane had longer filtration runs between CIPs
 Energy requirements for DUF were lower compared to conventional treatment process
and the UF.
 It can be concluded that both the DUF and UF technologies are viable alternatives to
conventional water treatment processes in the treatment of low turbidity water.

ACKOWLEDGEMENTS

The authors wish to thank Umgeni Water for funding the research and permission for
presenting this paper. Dow Water & Process solutions is also acknowledged and thanked for
technical support and the loan of the UF pilot plant.
REFERENCES

Pillay, V.L., Buckley, C.A., 2003. Evaluation and Optimization of a Crossflow Microfilter for the
Production of Potable Water in Rural and urban Areas. WRC Report 662/1/03.

Pryor, M.J., Jacobs, E.P., Botes, J.P., Pillay, V.L., 1998. A low pressure ultrafiltration
membrane system for potable water supply to developing communities in South Africa.
Desalination 119, 103–111.

Jacobs, E.P., Pillay, V.L., Pryor, M., Swart, P., 2000. Water Supply to Rural and Peri-urban
Communities Using Membrane Technologies. WRC Report 764/1/00.

Logsdon, G S (1978) Direct Filtration: Past, Present and Future. Civil Engineering - ASCE.

Culp, R L (1977) Direct Filtration. Journal of the American Water Works Association

CD Swartz, CJ van der Walt and PJ van der Merwe (1990). Evaluation of direct series
filtration for the treatment of South African Surface waters. WRC Report No. 354/1/97

Odira, M A A (1985) Up flow Filters in Flocculation and Direct Filtration of Waters of High
Turbidity. PhD Dissertation, Tampere University of Technology, Finland.

Foley, P. D. (1980) Experience with direct filtration at Ontario's Lake Huron Treatment Plant.
Jour. AWWA 72(3)162.

Churchhouse, S., (2000) Membrane Bioreactors hit the big time - from lab to full scale
application In Membrantechnik, 3.

Peter-Verbanets M., Hammes F., Vital M., Pronk W. (2010). Stabilization of flux during dead-
end ultra-low pressure ultrafiltration. Water research, 44(12), 3607- 3616.

Peter-Verbanets M., Pronk, W., Point-of-use membrane systems: place in the world of water
supply. Techneau WP 2.5.2, (D.2.5.2.), August 2006.

Ramesh, A.,Lee,D.J.,Lai,J.Y.,2007.Membranebiofoulingby extracellular polymeric substances


or soluble microbial products from membrane bioreactor sludge

You might also like