Sie sind auf Seite 1von 2

mixing

Fluid mixing re-engineered


Improving a
mixing
system
design can
have
substantial
economic
benefits,
say Mike
Butcher and
Warren
Eagles
Below: An example
of site work to
check for potentially
damaging flow
induced vibrations
caused by
interaction between
flow from the
impeller and
vibrations of shaft
and internals
Bottom: The
reaction tube
bundle of the
Flexreactor which
has been developed
for pilot scale PI
production of
chemicals

MIXING technology has a major


influence on business efficiency in the
process industries with 50%, by value,
of all chemical production being
performed in batch stirred vessels. This
represents a sales value turnover of
1370b ($1290b) worldwide and 402b
($397b) within the EU (figures 1999).
Poor initial design results in mixing
related problems estimated at costing
industry 0.53% of total turnover, or
some $211b/year. Problems recorded
include a high level of commissioning
failures, low production rates, increased
downstream separation costs and
increased downtime for maintenance.
Solid suspension plant is especially
susceptible to these problems with only
60% of plant achieving the design
production figures (Shamlou, 1993).
Typically a large chemical plant
experiences around 12 operating problems
a year which are mixing related, and the
cost of resolving these varies between
about $0.4 and $4m (figures 1996).
These costs do not include the
consequences of a runaway reaction and
there are on average around eight a year
in the UK with poor mixing making a
significant contribution to these
incidents. A classic example of this is
where a company increases the batch size
in an endeavour to increase production.
Even a small increase in liquid height can
have a significant effect on surface
turbulence and drawdown of reactants,

especially with agitators such as a


single retreat curve impeller. With many
reactors using the traditional surface
dosing position for reactants it is
possible to build an inventory of nonreacted chemical in the upper layers,
which can be drawn down in an
uncontrolled manner. This type of
problem is especially noticeable where
there is a viscosity/ density difference
between the liquids, as often occurs in
a precipitation or dissolution stage.
The design of mixing processes is
not a straightforward procedure largely
due to severe limitations in the
published literature, where correlations
often disagree. Published correlations
for a solid-liquid suspension process
suggest an increase in power per unit
volume by a factor of anywhere
between 0.1 (Bourne) and 3 (Kneule)
for a ten-times increase in vessel
volume. Many studies have
concentrated on small-scale equipment
and to address this, BHR Group in 1983
formed FMP (Fluid Mixing Processes),
an industrial mixing consortium that
became focused on the mixing
processes employed in stirred vessels. In
1989, HILINE (high intensity in-line) was
formed, which focused on inline mixing
and mass transfer equipment.
In many cases the mechanisms of
mixing based on a fundamental
understanding of the physics and fluid
dynamics will have commonalities for
different mixing equipment. So in 2002
the two consortia were merged under the
name of FMP, which continues to build on
over 350 man-years of work, to provide
design guidelines and scale-up rules for
both batch and continuous stirred tank as
well as for high intensity inline mixers.
Members which benefit from the
confidential research include
manufacturers of bulk and fine chemicals,
pharmaceuticals and household products,
CFD code vendors and equipment
companies. The consortium has the
facilities and capabilities to investigate,
either individually or in combination, a
wide range of mixing equipment
including ejectors, heat exchanger
reactors, jet mixers, static mixers, tee
mixers, pumps, loop and venturi reactors
as well as the plethora of different
agitator types.
The aim is to generate rules, based on
a sound mechanistic understanding of the
various micro, meso and macro-mixing
processes for reliable design, operation

A CFD plot of solid suspension with


comparison of measured and predicted
concentrations in different positions
and scale-up. Members use FMP
knowledge and work with FMP engineers
on projects as diverse as minimising byproduct formation with multiple chemical
reactions, controlling the droplet size
distribution of creams and pastes, and on
modifications to the mechanical design of
agitator systems to avoid catastrophic
failure due to flow induced vibrations.
Recorded case studies show savings of
millions of dollars from the application of
FMP knowledge.
Systematic research may be carried out
in a given area, and particular emphasis
has been placed on establishing close
links between experimental and CFD work.
Industrial relevance is reinforced by
developing techniques such as using
video probes (to measure drop sizes) and
laser induced fluorescence for use at pilot
scale and carrying out experiments up to
production scale (2.7 m diameter for
stirred vessels and 25150 mm for inline
mixers). Techniques currently under
investigation include tomography and
particle image velocimetry. These coupled
with advanced rheological measurement
facilities will aid understanding of the
fundamental mechanisms involved in
mixing and permit the generation of

mixing

reliable scale up rules.


The research conducted so far may be
divided into seven areas.

blending of difficult liquids


Work on blending difficult liquids
encompasses the hydrodynamics and
mixing of viscous and non-Newtonian
fluids into similar and dissimilar bulk
fluids and includes the blending of
Newtonian and shear thinning liquids over
a range of viscosity and density
differences. Investigations are underway
into the effect of geometry, operating
conditions and impeller/jet interactions
on the performance of CSTRs and also
visco-elastic and yield stress fluids,
systems designed specifically for viscous
materials and blending performance of
rotor stator mixers. The aim is to compare
and unify mixing measurements such as
coefficient of variance (CoV), mixing time
() and RTD in STRs, CSTRs, static mixers
and rotor stator devices leading to better
equipment selection.

liquid-liquid dispersion
Much experimental data has been
collected on various equipment for
creating dispersions. This has led to
predictions of drop size distributions with
more recent work focusing on higher
dispersed phase concentrations. A more
diverse range of flow regimes and fluid
properties is to be investigated, including
liquids with increased viscosities and
viscosity and density differences. This
includes the addition of surfactants to
produce micron and sub-micron droplets.
Better understanding of droplet
formation, break-up and coalescence will
be gained from fundamental studies into
the local flow structures, shear rates and
length and time scales. This knowledge
will be applicable to the kinetics of
dispersion processes, aiding creation of
design and operating guidelines for batch
and inline systems. One target for this
knowledge is the manufacture of
emulsions and controlled dispersions, the
nature of which are often essential to the
functionality of a companys final product.

gas-liquid systems
Developments in the gas-liquid area are
enhancing guidance on mixing for
reaction. Correlations for gas hold-up, kLa
and gassed power draw already exist for a
range of impeller geometries, tank
geometries and static mixers. This is
being expanded to a broader configuration of equipment (for example, multiple
impellers), liquid properties (viscous
liquids, complex rheology, organic liquids)
and gases with a range of solubilities.

Dependent on the industrial


application, surface aeration can either be
desirable or not, therefore guidelines
outlining the conditions that promote/
avoid surface air entrainment are being
prepared along with an investigation into
de-foaming techniques.

solid-liquid systems
Solid-liquid mixing covers two main duties
suspension and distribution (sinking
solids), and drawdown (floating solids).
Investigations are being extended to
include a greater range of concentrations,
particle sizes, wetability, shapes and
properties as well as tank internals and
inline devices. More fundamental studies
will focus on the modification of
turbulence due to the presence of solids.

three phase systems


This new area within FMP will concentrate
on solid-liquid-gas mixing, building on
the existing knowledge of STRs to
investigate the effect of a third phase on
mass transfer, blending, turbulence, holdup, and so on. The target is to produce
qualitative guidelines for the deviations
from two-phase behaviour. This will be
extended into liquid-liquid-gas and other
devices to address the industrial needs of
the members.

reactive mixing
Work in process intensification has
confirmed that different scales of mixing
(micro, meso or macro) can
control/influence final product
composition. Developing rules for
designing and scaling up reaction
processes based on data, correlations and
design methods for a wide spectrum of
reactor types, will have the aim of making
the mixing in the reactor fit the reaction
needs.
Existing work on low added fraction,
equal viscosity and turbulent systems in
static mixers will be developed to apply to
different viscosity ratios in the turbulent,
transitional and laminar reactive mixing
regimes, and also high added fractions
and relatively low mixing intensities
where larger scales of mixing are
important. The Bourne azo-coupling

scheme and other chemical reaction


probes will be used extensively but where
reaction timescales are closer matched to
the mixing timescales new schemes will
be developed.

CFD and flow modelling


The CFD area identifies best practice within the field and validates code against
high quality experimental data, resulting
in a CFD good practice guide. Codes will
be tested, as they are developed, and
turbulence predictions evaluated, using
tools such as large eddy simulations
(LES). The complex problem of multiphase simulations is also being validated
against new data to further strengthen
guidelines provided to members.
This area has also developed a range of
embedded and linked models enabling
simulations of micromixing effects in
reactions, solid distribution and
precipitation processes.
The information is made available
electronically, giving access to an expanse
of data including CFD meshes and
simulations, large sets of LDA data, raw
and processed experimental data and
video footage of interesting mixing phenomena. This allows mixing experts within
companies to integrate the information
with their in-house design information.
Improving a mixing system design can
have substantial economic benefits. For
one company, carrying out a study into
reactor fouling in their polymerisation
plant resulted in improvements that
brought about savings of $5m a year.
Such studies can often result in reduced
energy consumption. A UK survey by the
Energy Technology Support Unit (ETSU)
indicated that given a 20% uptake of
improvements in mixing-related technologies a saving of $90m was possible.
Extrapolated to the EU this equates to
550m. The bulk of these savings result
from the consequential benefits of
improved separation, reduced waste and
by-product production, and a corresponding reduction in raw material requirement.
Additional benefits include enhanced
safety and environmental impact, the
costs of which are increasing rapidly with
the development of legislation.

A CFD simulation
of precipitation
reaction which
has integration
of population
balance into
the CFD solver
allowing
interaction of
mixing and
crystallisation
kinetics.
From left:
Supersaturation,
number
concentration
and mass
concentration

Mike Butcher
and Warren
Eagles work for
BHR Solutions

Das könnte Ihnen auch gefallen