182 Solid/Liquid Separation: Scale-up of Industrial Equipment

systems in particular will heat up significantly with processing time. As

discussed above, it is vital that the tests are undertaken at the system
temperature and reflect any temperature excursions that naturally exist.

Pre-filtration: The pre-filtration of the process fluid should be set at a

maximum rating of one tenth of the diameter of the microfiltration
fibre or channel tube size. The other factor when considering pre-
filtering for the system is the contamination sensitivity of pumps and
valves used in the system which may be damaged by erosion caused by
solid contaminants carried in the process fluid.

Test rig set up: The test rig should be connected up to the process line
and run for a minimum of 500 hours to prove the long-term flux
stability. The required volume concentration factor of the contaminants
should be achieved on each batch of fluid processed.

When working with MF, it is often preferable to work at constant

permeate flux, allowing TMP to increase due to membrane fouling
(TMP < 2 bar). Under these conditions, back pulse cleaning can be
very effective in sustaining the permeate flux.

4.3.2.6 Case study

An example of how a scale up trial was used to size a tangential flow
filtration system is detailed below.
The application was an automotive industrial wash used to clean engine
blocks from an internal combustion engine before assembly. Poor wash
performance could result in parts being inadequately cleaned affecting
production. Results from a pilot site trial are shown in Figure 4.34.
The minimum permeate flux of 42 LMH was achieved with a membrane
area of 0.5 m^ for the pilot site trial with a fluid volume concentration
factor of 35%. The required batch concentration was typically achieved

1
125
100
X
^ 75 *
50 ' ^ -. • • .. . . • -^\-. H.\ • - - . . . ^ • - ^ - . -
CD
I 25
0) 0 J , 1 - I --•
0 100 200 300 400 500

Time, hours
F i g u r e 434 Flux maintenance. Microfiltration permeate flux from site trial of an
industrial wash
 4 • Membrane filters - microfiltration and ultrafiltration 183

in 100 hours of processing. Chemical cleaning after the trial for 8

hours restored the clean water permeate flux to 94% value of the new
membrane, which was considered an acceptable recovery.

Conclusions of site trials: The trial results demonstrate that a steady

permeate flux was achieved over 500 hours of operation. The peaks in
permeate flows are due to the batch concentration with the lowest
permeate flux being achieved at the highest concentration, which
would be expected.
This trial demonstrates that a steady permeate flux has been achieved
at a minimum of 42 LMH. This would allow confident selection of the
most appropriate membrane, accurately scaled up to achieve the
required permeate flux to maintain the wash at the low contaminant
level required for acceptable component cleanliness.

Calculation example: The example wash had 10 m^ capacity tank and

the fluid required changing every two weeks, as oil level in the wash
fluid reached 1%, which was known to affect component cleanliness.
A tangential flow filtration system that produced a clean permeate flow
of 400 litres/hour or 9,600 litres/day would maintain the wash at
maximum of 0.2% oil contamination while at full production and
under normal ingression rates. Therefore the required permeate flow
for the wash is 400 litres/hour of contaminant free wash fluid.

Scaling up the permeate flux from pilot site trials:

Pilot minimum permeate flux = 42 LMH
As before,

Jt

Therefore

42

The tangential flow membrane area == 9.5 m^.

This represents a scale-up factor of 20 times from the pilot testing to a
full size system, which is considered acceptable after operating for 500
hours.
184 Solid/Liquid Separation: Scale-up of Industrial Equipment

The above system was installed on an industrial wash machine with a

membrane area of 10 m^ and gave an average permeate flux of 45 to 60
LMH over one year of operation.
Chemical cleaning of the membrane was carried out monthly to help
maintain the flux. The unit successfully maintained the wash fluid at
low levels of oil contamination and the fluid was kept in service for
one year from the original 2 weeks fluid life that had been experienced
before installing the system.

4,4 TFF/DFF applications

4.4.1 Water processing

As we have seen, membrane systems can be operated in DFF and TFF
modes. There are no specific rules to follow when deciding which flow
method should be used for which configuration of membrane. This is
better defined by the characteristics of the fluid in question. Indeed,
many hollow fibre membranes are used in DFF and vice versa.
The use of membrane systems has experienced exponential growth
over the last 5 years due to the ability to deliver drinking water and
wastewater within compliance limits. With a modular design and
sophisticated automation, membrane plants are now being built with
safety, flexibility and minimal operator intervention. Another driving
force for the popularity of membranes is that their costs have never
been lower, spurred by increased competition, more efficient process
operation and manufacturing capacity. The single factor that dictates
the membrane plant size or the number of membranes is the mem-
brane's ability to handle fouling. A well-designed membrane, one that
has high porosity and permeability, will inherently provide better flux
and economics. However, to make use of those physical properties,
membrane fouling must be controlled in a practical and economical
manner. Pre-treatment of the feed water usually has a quantifiable
benefit on the sizing of membrane systems but this obviously has to be
offset against the associated additional operating cost.
Another performance improvement is the effective use of "enhanced
flux maintenance" (EFM) strategies that can effectively increase the
membrane flux, thereby reducing the cost of membranes installed quite
significantly. Keeping the membranes clean as often and as long as
possible allows the most efficient use of the membrane area and
porosity.
 4 • Membrane filters - microfiltration and ultrafiltration 185

4.4.1.1 Water quality analysis

An assessment of the required equipment and treatment options of a
source water normally begins with the review of a water quality analysis.
Historical data, if available, provide the best indication of seasonal
changes and variations. Unfortunately, this information is not always
available, so it may be necessary to procure a sample and obtain an
analysis from a certified laboratory. Water quality results may also be
used to determine the need for on site or laboratory pilot testing. It
should be noted, however, that a single sample, while valuable, provides
only a limited snapshot of the actual water constituents. Parameters that
are often included in an analysis of drinking water include the following:

• Total Organic Carbon (TOC)

• Turbidity
• Total Suspended Solids (TSS)
• Iron
• Manganese
• Hardness
• UV254.

Other constituents may need to be measured depending upon the water

source, finished water requirements, and national and local regulatory
standards.

4.4.1.2 Pilot testing

Due to variations in water quality and type, it is often advisable to
conduct small-scale trials, or pilot tests, on the source. This evaluation
typically lasts 2-6 months and is intended to obtain continuous data
that is used to effectively size and cost the full-scale membrane
system. The trial usually consists of an optimisation phase (2-4 weeks)
followed by one or more design runs (4 weeks). Pilot tests are used to
accomplish the following:

• Demonstrate successful operation with the specific water source

• Confirm finished water quality
• Optimise system performance and costs
• Determine full-scale design criteria -
Flux
Cycle duration
1 8 6 Solid/Liquid Separation: Scale-up of Industrial Equipment

Operational set points

Flux maintenance protocols
• Assess power and chemical consumption
• Provide customer training and orientation.

Pilot Equipment: An automated single module test rig is typically used

for pilot evaluation. The system is equipped with a single hollow fibre
MF module. The physical characteristics of a PVDF hollow fibre
membrane are described in Table 4.5. A schematic of the MF system is
shown in Figure 4.35 with a photograph included below in Figure 4.36.
There are four basic modes of operation for the membrane unit:

Forward filtration'. The feed pump draws water from the feed tank and
pumps through the membrane filter directly to the top of the module.

Table 4.5 Typical MF module specifications.

Outside membrane area 50 m2

Module length 2 metres
Module diameter 6 inches (152 mm)
Nominal pore size 0.1 i^m
Maximum pore size N/A
Membrane material PVDF
Flow direction Outside-in
Temperature operating range Freezing point - 40°C
Feed water turbidity operating limit 0-500 NTU

Legend:
LOWER FT: flow transmitter
RF DISCHARGE PT: pressure transduce
TI: Temperature transmitter

Figure 4.35 Pilot test rig schematic

 4 • Membrane filters - microfiltration and ultrafiltration 187

Figure 4.36 Pilot test rig

Part of this flow can be recirculated to the feed tank, a process known
as excess recirculation.

Simultaneous air scrubbing and reverse filtration (SASRF): SASRF is

a frequent, short-duration hydraulic cleaning of the membrane to main-
tain optimal performance. During SASRF, air is injected into the filter
module while simultaneously pumping filtrate from a storage tank into
the top of the module. The combined water-air flow creates strong
turbulent and shearing forces to dislodge deposits on the membrane
surface. The waste is discharged through the upper discharge port of
the module. A reverse flush (RF) is used after the SASRF to flush out
the solids dislodged during air scrubbing.

Forward flush (FL): FL is another form of hydraulic cleaning for the

membrane that follows a SASRF. The feed pump draws water stored in
the feed tank and pumps the water through the membrane filter in the