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

• Bulk sample size - sufficient sample needs to be taken such that, if

possible, all products to be evaluated can be compared on the same
substrate over a range of dose levels. If this is not possible then
careful thought needs to be given to experimental design. If several
different batches of substrate have to be used then it is recom-
mended that products are directly compared at similar doses on the
same batch of substrate, rather than using a different batch for each
product. The number of "control" tests required also needs some
careful consideration.
• Sample ageing - the rate of sample ageing needs to be identified. If
it is rapid then it may be necessary to use different batches of
sample since, in this case, the ageing effects could be more
significant than any batch-to-batch variation.
• Sampling point - wherever possible the sample obtained should be
taken from a point immediately prior to the addition of the pre-
treatment chemical under investigation. One particular problem
area is that of liquor re-circulation. In closed circuits the liquor
being returned can contain some of the pre-treatment chemicals
that are under evaluation, so it can prove to be very difficult to
obtain samples that are truly untreated. It is also desirable to take
the sample from a fast flowing vertically downward or upward flow
pipe as opposed to a horizontal pipe or open channel. This is due to
the increased possibility of sedimentation or stratifying in a
horizontal pipe, particularly in a slow flowing region.
• Sub sampling - all actions possible should be taken to ensure that
sub samples are representative of the whole and that there is no
trend developed as the bulk sample is depleted. A typical example
would be that if there is insufficient mixing during sub sampling
then there is likely to be a gradual increase in suspended solids
content due to settlement, particularly of coarse material.
• Temperature - it is important that the tests are conducted at a
temperature that is as close as possible to that encountered in the
practical situation. Temperature can affect the speed of the separation
process itself, the product of choice and the final results obtained in
terms of cake solids, etc. Where the substrate is treated at elevated
temperatures it is recommended that, as well as conducting the test
work at the appropriate temperature, the substrate is not allowed to
cool between sampling and testing. The reason for this being that if
the substrate is allowed to cool, soluble constituents may precipitate
out onto the surface of the suspended solids. On reheating they may
not always re-dissolve, thus changing the surface of the solids to
be treated. Changing the solid's surface in this way can have a
 2 • Chemical pre-treatment 51

significant effect on the efficacy of pre-treatment chemicals, since

surface adsorption is a major mechanism, and could also result in
incorrect choice of product.
• Other reagents or conditions - it is necessary to ensure that
reagents, other than the one under investigation, that are being used
are also present in the substrate at the correct level and at the
correct time e.g. it needs to be ensured that pH is correct.
Dependent upon where it has been possible to take samples, other
reagents may or may not have to be added, for example reagents
will have to be added if the only sampling point is before the
reagent is present in the system, or if a further reagent is added
after the pre-treatment chemical being evaluated.

2.3.1.2 Substrate characterisation

It is important to characterise the test substrate for a number of reasons
that includes identifying batch to batch sampling variations that could
affect the efficacy of different products, confirming or otherwise that
the sample obtained is typical of the substrate to be treated, providing
information that may help in choosing the most appropriate pre-
treatment chemical. Typical measurements could include:

• Total solids content

• Suspended solids content
• Dissolved solids content
• pH
• Conductivity
• Appearance (colour, odour, etc)
• Ash content or volatiles content.

2.3.2 Reagent preparation

The reagents described above are available in a wide variety of
physical forms. In most cases it is necessary to prepare an aqueous
solution, if the products are solids or emulsions as supplied, or dilute
from a concentrate. In determining the most appropriate solution
strength to use there are two main considerations. Firstly wherever
possible the reagent solution used should be of a similar strength to
that likely to be used in practice on full scale. The second consideration
is that of accuracy. The strength and volume of solution used should
take into account the ease of weighing/measuring the solid or
52 Solid/Liquid Separation: Scale-up of Industrial Equipment

concentrate to the accuracy required, and subsequently the volume of

the reagent to be added to the test substrate. Both of these factors can
result in solutions of lower or higher strength being used in laboratory
evaluations compared to the full scale situation, the effect that this may
have on ease of mixing and subsequent results is discussed in Section
2.3.3.
For the majority of types of chemicals discussed, the reagents are
either readily soluble solids or easy to handle concentrates. The
exception to this is the high molecular weight flocculants group where
care in solution preparation for both laboratory test work and full scale
plant operations is required. These products are hydrophilic, and as
soon as they come into contact with water they begin to swell. They
continue to absorb water until fully hydrated and then form an
homogeneous solution, this can take up to two hours for solid grade
products. In the initial swelling stage, if the individual particles are not
kept separate they can clump together and form large lumps that can
then be very slow to dissolve. To overcome this effect there are
specific laboratory techniques, one of which is described in detail
below. For full-scale use, there are a number of specialised preparation/
dissolving units available on the market.
The normal solution strength prepared in the laboratory usually ranges
from 0.25% to 1.0% dependent upon the specific product being
prepared. This concentrated solution is usually diluted around ten fold
to form the working strength solution.

2.3.2.1 Example
500 cm^ of make-up water is placed in a 600 cm^ beaker and stirred
using a mechanical paddle type stirrer at a speed sufficient to create a
vortex. 2.5 g of dry polymer is sprinkled into the vortex taking care
that the polymer doesn't form clumps. (If clumps do form the solution
should be discarded). Once all the polymer has been added the stirring
speed is reduced to a level that ensures there is full movement of all the
solution, but low enough to prevent shearing of the polymer. Stirring is
continued for two hours or until a fully homogeneous solution is
formed. This solution can then be diluted to the appropriate working
strength solution.
If a 40% active emulsion polymer is used in place of the solid grade
polymer, then 6.25 cm^ of product should be expelled into the vortex
via a syringe. To ensure that an accurate solution strength can be
calculated, the syringe should be weighed before and after the
emulsion is expelled so that the exact quantity of emulsion added can
be determined.
 2 • Chemical pre-treatment 53

2.3.3 Mixing and application techniques

It is well known that choosing the correct mixing regime and
application technique is as important as ensuring optimum product
choice (Gregory, 1981). It is also known that the flocculation
efficiency of a fixed dose of flocculant can be minimal if poor
dispersal occurs due to insufficient mixing or be similarly minimal due
to floe destruction due to excessive mixing shear (Weir and Moody,
2003). The maximum flocculation efficiency occurs with intermediate
mixing where efficient flocculant dispersal is achieved and subsequent
mixing leads to floe forming collisions. Therefore it is important that
when conducting any application test work that both mixing regime
and application technique are given careful consideration. As with
solution strength, conditions that are likely to be encountered in the
full-scale situation should be replicated as closely as possible.
Factors to consider include:

• Solution strength
• Solution viscosity
Solids concentration of substrate
Mixing regime - length of time, shear applied, type of mixing
Single point or multi point addition.

The types of product where these considerations take on most

importance are the flocculants whether they are natural products, such
as starch, or high molecular weight synthetic flocculants. Flocculation
is not a permanent effect. Floe formation, subsequent breakage and
reformation occurs over a short, but deflnitive period of time (Farrow
and Swift, 1996). Sufficient mixing needs to occur to provide intimate
contact of flocculant and substrate to ensure flocculation occurs, but
too much mixing will cause the floes to break and so reduce the
effectiveness of solid/liquid separation. In full-scale plant it is not
always possible to ensure that conditions are conducive to optimum
flocculation, for example the best flocculation may be achieved at very
low solution strengths but this may result in too much water being
introduced into the process or insufficient fresh water may be
available; when flocculated substrates are introduced into centrifuges
they encounter very high levels of shear which cannot be avoided etc.
Therefore when looking at any of the application techniques which
follow, the first task is to decide the most appropriate mixing regime to
use.
54 Solid/Liquid Separation: Scale-up of Industrial Equipment

2.3.4 Dose level calculations and definitions

Before commencing any test work there are a few basic calculations
and definitions that need to be known and understood.

2.3.4.1 Substrate solids concentration

In quoting substrate solids concentration, it is necessary to distinguish
between suspended solids content and total solids content (suspended
plus dissolved solids). In high suspended solids systems this is less
important as the dissolved solids content is only a small proportion of
the total solids, however in low solids systems the dissolved solids
content can be higher than the suspended solids content. Lack of
clarity could result in misleading conclusions regarding throughput,
dose level calculations etc.
It is also important to quote the solids content as weight solids per
substrate volume (w/v) or weight solids per total substrate weight (w/w).
In contrast to the comments above, regarding dissolved solids, in the low
solids system the difference between these two measures is often quite
small, whereas in high soHds systems, particularly those containing high
density suspended solids the difference can be significant.

2.3.4.2 Dose level

Often dose levels are seen quoted as parts per million (ppm), but this can
be very misleading as it is dependent upon whether this refers to parts
per million of total substrate or parts per million of dry solids. In order to
clearly distinguish between these two alternatives it is more appropriate
to quote dose level as either grams polymer per tonne dry solids (g/tds)
or in milligrams polymer per litre of substrate (mg/1). At the start of a
series of test work it is possible that the solids content of the substrate is
not yet known and it is therefore easier to calculate dose levels based
upon mg/1 that can subsequently be converted to g/tds as follows:

Dose level (g/tds) = ^ ^ ^ " ^ ^ (2)

where M is the dose level (mg/1), and S is the suspended solids content
(% w/v).
Calculation of the volume of reagent to add is carried out as follows

Volume of reagent required (cm^) = x (3)

C x 1000 1000
where D is the dose level required (mg/1), C is the solution strength
(%), and F i s the volume of test substrate (cm^).
 2 • Chemical pre-treatment 55

2.3.5 Qualitative screening tests

A very good starting point when encountering a substrate for the first
time is to carry out qualitative tests to determine whether or not certain
reagents have any effect. Once some effect is observed then it is
worthwhile progressing to specific quantitative tests. Since coagulation
and flocculation is a very visual effect, simple screening tests can be
devised using visual observations. Typically, the required amount of
sample (at least 200 cm^ is recommend, although if sample is in limited
supply, useful information can be obtained from as little as 50 cm^) is
taken in a beaker, using glass or clear plastic to facilitate observation.
The beaker containing the feed sample is taken in one hand and a
beaker containing the measured out polymer solution is taken in the
other. The feed is poured from the one beaker into the other and then
back and forth from one beaker to the other whilst visual observation is
made of the floe build-up. The number of beaker pours is noted (ten
times should be sufficient to see an effect, and the effect may be seen
with much less, however to simulate the mixing encountered in a
centrifuge then 50 pours or more may be required). An alternative is to
induce mixing by mechanical stirring using a paddle blade.
Observations that can be made include assessments of floe size,
settlement rate, liquor clarity and the speed of effect.
Often, the dose level required is unknown, and since it is possible to
overdose with flocculants and coagulants there can be some danger in
testing at a single dose. Where dose requirement is unknown a simple
approach is to take a single substrate sample and begin by adding a
relatively low dose level e.g. 0.5 mg/1, mixing the substrate and
flocculant, and then making further additions, doubling the total dose
until an effect is seen. If a total dose of >500 mg/1 is reached without
any observable effect then it is unlikely that further additions will
improve the situation.
These tests alone should not be used to recommend a product for full
scale testing, but are appropriate in helping to identify the range of
products that are worth evaluating in greater detail.
There are a number of sophisticated tools on the market (Barrett and
Glennon, 1999) that can more accurately measure particle size
(although these are often limited to 1 mm or possible 2 mm maximum
size) and in some cases provide images of floe shape etc., but these are
for the moment generally limited to use in more fundamental research
studies. However, it is anticipated that as the technology develops, they
will become increasingly used in the industrial situation.