I. INTRODUCTION
Industrial or municipal
wastewater treatment in a biological treatment facility produces liquid
effluents which are practically devoid of organic contaminants, however, they
still contain:
A remaining small organic load (typically
BOD5<50 mg/l, COD<150 mg/l)
A portion of all recalcitrant organic
compounds
Suspended solids
Significant number of microbes (some of them
pathogens).
In order to remove all the
above from the effluent stream, tertiary treatment units are often employed
which include among others like:
Addition of chemicals that adjust the pH in
the range where they act as coagulants.
Addition of coagulants and poly-electrolytes
for the removal of small agglomerates and colloidal particles
Filtering through multi-layer sand systems
and
Removal of pathogens through the addition of
Cl2 (or of hypochlorous oxide), ozone or through UV irradiation.
The main drawbacks of the
above classical systems for tertiary treatment are:
A constant and highly reliable operation of
the biological wastewater unit is assumed.
If for any reason there is a small carry
over of activated sludge into the treated water effluent, something rather
common in a biological treatment unit, the tertiary treatment unit becomes
essentially non-operational and it returns back to normal operation after a lot
of effort and expense.
As the coagulation / filtering process
affects only the suspended solids and colloids, the effluent still contains
dissolved organic contaminants that need to be removed. This is typically
accomplished using activated carbon filters (GAC), the operation of which
requires a significant expense as they need to be changed on a regular basis.
Finally, experienced and dedicated personnel
are required to operate the system.
Tertiary treatment based on
a membrane/microsieve filtration with a high retention compared to common sand
or multi-layer filters method has none of the above drawbacks and at the same
time it produces very clean treated water that can be reused readily.
Membrane biological
reactors (MBRs), containing of a bio-reactor with integrated micro- or
ultrafiltration membranes, have been used in many applications to treat both
municipal and industrial wastewater, and to reuse treated wastewater.
The concept of MBR systems
consists of utilizing a bio-reactor (e.g., activated sludge system) and a
microfiltration as one unit process for wastewater treatment thereby replacing,
and in some cases supplementing, the solids separation function of secondary
clarification and effluent filtration. Consequently, MBR systems provide the
following advantages:
1. Higher volumetric loading rates and shorter hydraulic retention
times (HRTs);
2. Longer sludge retention times (SRTs) resulting in less sludge
production;
3. High quality effluent in terms of low turbidity, bacteria, total
suspended solid (TSS), and biochemical oxygen demand (BOD5); and
4. Small footprint required, compared to that of the conventional
activated sludge system.
In addition, by replacing
solids separation by gravity settling in secondary clarifiers, the MBR systems
avoid issues of filamentous sludge bulking and other flocculants settling and
clarification problems. Further on the aeration tank SRT is no longer
controlled by solids loading limitations of secondary clarifier.
In
II. EXPERIMENTAL SET-UP
The membrane module has
been submerged in an existing SBR treatment plant with a height of 2.85 m and a
total footprint of 1.183 m. The volume accumulates to 9.7 m which has been
calculated for a connection parameter of 6 PE. During both test phases an
inflow of 5 PE which results in around 750 l/d has been calculated.
One of the main targets has
been to determine to what extent the degree of maintenance for a membrane
module in decentralized wastewater treatment can be reduced. In two different
modes of operation the membrane module has been submerged first in the
so-called polishing chamber followed by a second phase with a submerged
membrane module in the biological reaction chamber. Membranes have not been
back flushed nor chemical cleaned within the different modes, but exchanged at
the start-up of each phase.
The used membranes had a
molecular cut-off of about 150 kDa.
Figure 1. View
of the small WWTP in Oberhausen, Germany
First phase: Membrane module submerged in the polishing
chamber
Via the suction side of the
module with a membrane surface area of 3.2 m the permeate has been pumped to
the former sampling shaft by an impeller pump. From there the treated and
filtered wastewater has been lifted to the infiltration lines by a submersible
pump. During this mode of operation the filter module has been operated
independent of the operation cycle of the Sequencing batch reactor.
At the beginning of the
first research phase in August 2005 the flow rate corresponded to the design
parameters, so reaching a defined level in the biological reaction chamber
caused the shut down of the pump by a level control. Thus, run dry and
consequently an irreversible damage of the membranes had been avoided. The
filtration mode has been operated clocking which results in a material relaxation
during the filtration break.
Within the first test phase
specific permeate fluxes at an average of 8 l/hm and maximum 15 l/mh has
been achieved during the summer and early autumn months while the water
temperatures have been around 20 C. Decreasing of the temperatures
resulted in a proportional decrease of the specific fluxes (cf. Fig. 2).
Figure 2. Permeate
flux and temperature vs. first research phase - operation of the membrane
module submerged in the polishing chamber
Temperatures during winter
time caused a resulting flux below 5 l/mh. Rising of the temperatures during
spring time 2006 resulted again in an increased permeate flux. In the first and
second test periods different parameters have been analyzed on a regular basis.
The COD concentration in permeate had been on a nearly constant level of around
30 mg/l after termination of the start up phase. NO3-N concentration
in permeate had been between 17 mg/l and 22 mg/l in the same period
of time. Determination of Total Coliforms has been as expected on a constant
low level of at most 10 CFU. No Total Coliforms had been detectable in more
than 65 % of all analyzed probes.
After some small problems
during the start-up phase the test equipment runs very reliable without any
problems until shut down of the first phase in spring 2006.
Second phase:
Operation of the membrane module submerged in the Sequencing Batch Reactor
(SBR)
The filter module has been dismounted
from the polishing chamber in May 2006, reassembled with new membranes and
submerged into the biological reaction chamber. Right from the beginning the
filtration area of the module has been increased to 4.73 m. Stabilized
operation of the plant in the MBR-mode initiated in autumn 2006. Starting from
a permeate flux comparable to the one of phase 1 the flux decreased during the
second test period according to the decreasing water temperatures (cf. Fig. 3).
Further reasons could have been the higher MLSS concentration as well as
occurrence of Extracellular Polymeric Substances (EPS) - polysaccharides and
proteins produced by bacteria resulting from substrate absence, stress or as a
product of filamentous bacteria.
Figure 3. Permeate flux and temperature vs. second
research phase operation of the membrane module submerged in the SBR
Due to the occurrence of
EPS there is the risk that common filtration membranes can be irreversible
blocked which will result in a dramatic flux decline. During this time COD
concentration of the permeate raised from 25 mg/l to a maximum of 130 mg/l but
remained below the discharge limit of 150 mg/l. Concentration of NO3-N
remained at a low level of around 5 mg/l. Total Coliforms were also hardly
determined like in the first test period.
After shutting down the
trials in the beginning of 2007 bacause of the operational problems it has been
obvious that the cleaning of the membrane surface by the aeration equipment had
to be optimized due to fouling on the membrane surface.
III. OUTLOOK: MICROSIEVE FILTRATION FOR MBR
APPLICATION
Since, microfiltration with
conventional polymeric membranes has some disadvantages regarding to gradual
flux decrease and membrane fouling, the development of a novel generation of
filter in the sector of microfiltration, so-called microsieves, were reinforced
in recent years [5].
The microsieves are
composed of metal, e.g. stainless steel, and feature custom-made pore geometry
with a multitude of more than 100 billions microscopic holes (Fig. 4). They are
characterized by high filtrate capacity (10 to 40 m/hm with water), high
selectivity, great robustness (500 N/mm), as well as easy cleaning, and
sterilization.
Figure 3. Microsieve
made of stainless steel with pore sizes of 2 m [2]
A newly developed,
laser-supported micro-welding method allows welding the micro filters into
gas-tight compact filter modules for technical application (Fig. 5).
Figure 4. Laser-welded
microsieve flat module made of stainless steel [2]
Since, due to the
manufacturing method, microsieves feature a very narrow pore size distribution
and have hardly any defects they guarantee high filtration reliability with
complete retention of all compounds smaller than the pore size. Furthermore,
the smooth microsieve surface inhibits the adhesion of filter cakes. Thus,
microsieves maintain high filtration efficiency, particularly in combination
with an antifouling strategy like shearing effects of bubbles during aeration.
In future, for these benefits, microsieve filtration can contribute to the aim
of low-maintenance hygienisation of WWTPs outlet. Advancements comparing to the
traditional filtration methods are expected. Recently, some filtration
experiments with artificial fluids and water are performed in order to
characterize the behavior of microsieves. Further investigations are planned
with microsieves in the field of wastewater treatment [3, 4].
IV. CONCLUSIONS
Operation of the submerged
membrane module without further cleaning strategies is as an alternative for
decentralized wastewater treatment. There are
advantages in operation in the specific case of submerging the membrane module
in the polishing chamber. As a matter of fact this solution is not always been
given because of load and specific treatment volume. By optimizing the distance
between membrane plates and thus better cleaning due to the arising air
bubbles, longer lifetime cycles of the membranes/microsieves might be
achievable which result in a more stabilized operation.
In future, metallic
microsieves can become a very efficient alternative filtration technique in MBR
applications compared to conventional polymeric micro filters.
Acknowledgements
The authors would like to
thank Mr. and Mrs. Schwarz and the whole family on whose real estate the trials
had been undertaken and who always supported the research team also during
difficult experiments. Furthermore we would like to thank UPONOR Klärtechnik GmbH in
The German Federal Ministry
of Education and Research is funding the development of microsieves under the
Reference No. FKZ 01 RI 05049.
REFERENCES
1.
Austermann-Haun U., 2004. Abwasserentsorgung in ländlich strukturierten
Gebieten. Lecture at the cooperation forum decentralised waste water treatment,
2.
3.
4.
5. Van Rijn C. J. M., 2004.
Nano and micro engineered membrane technology. Elsevier,