In the mix: how to keep a lid on process costs

Mixing has often been considered something of a ‘black box’ process. Provided everything ends up mixed in the end, who cares what actually goes on inside the pipe or vessel, right?

Yet this attitude could be costing time and energy. It may even be degrading product quality in some applications – if over-mixing causes too much damage when aiming to get an even distribution of fruit pieces into a batch of yoghurt, for example.

Perversely, the fact that mixing is a fundamental operation in virtually all process industries is one reason why so many mixing operations are sub-optimal, according to Chemineer sales manager Irfan Rehman: “A big issue is the very large installed base. Some of these applications have changed over time so the agitator may not be used for what it was originally designed for.

“Sometimes users have applied ‘brute force and ignorance’ to achieve their result when they could reduce power consumption with a more efficient impeller. Sometimes it is just historic, over-designed mixing equipment that needs updating, especially impellers, where new efficient ones exist.”

You can look outside the vessel but it’s what’s inside the vessel that controls the power consumption

Irfan Rehman, Chemineer sales manager 

Energy consumption is a key consideration in mixing, where the arrival of developments such as the UK Government’s mandatory Energy Savings Opportunity Scheme (ESOS) only add to the pressure.

Like many operations, improved controls may be able to improve some aspects of the mixing process, but Rehman says that when it comes to optimising energy use, there’s only one place that users should be looking: “You can look outside the vessel but it’s what’s inside the vessel that controls the power consumption.”

While he describes most of the progress made on energy savings in mixing as incremental, rather than radical, new impeller technology can make a big difference in particular applications. For example, Chemineer’s recent introduction of the JT-2 impeller for transitional flow can reduce power draw by up to 50% for the same blending performance as competing impellers, according to the company. The JT-2 is designed for nonlaminar flow in the sort of higher-viscosity, non-Newtonian fluids that can crop up across a range of industries (think slurries, ketchup or paint, for example).

Maintenance burden

Rehman adds that a poorly designed system can result in other ongoing costs. “Maintenance can be an issue. The mechanical seal and gearbox may have to be maintained more regularly in a poorly designed system.”

Optimising power consumption can be a challenge both for agitator-based mixers and static mixers alike. Static mixers may not need a motor, but they may increase pumping costs if they’re over specified.

It’s easy to see if your mixing system is inadequate, since it won’t get the job done. But how can you tell if the process is optimised, overspecified or otherwise inefficient? That’s where flow mapping and modelling comes in.

Computional fluid dynamics (CFD) is an obvious way to go, relying on building a virtual model that simulates what’s going on in the vessel. As well as modelling the fluid flows of different agitator or impeller designs, CFD can also be used to analyse parameters such as heat transfer and the mixing and reaction rates of chemicals.

Maintenance can be an issue. The mechanical seal and gearbox may have to be maintained more regularly in a poorly designed system

However, CFD should be used with caution, warns Rehman: “CFD has to be used very carefully. It’s just a complementary tool that should be used in conjunction with lab work.”

It’s also not an easy option. The flow patterns in stirred tanks are complex, which makes traditional CFD a time-consuming process. A CFD technician may need days to define all of the equations and run the program. However, the technology is speeding up as new techniques are developed.

The physical modelling techniques used in lab work include digital particle image velocimetry (DPIV). This involves building a transparent model of the process and shining an Argon-ion light sheet through it to illuminate particles as they flow around.

A CCD camera captures the images, a computer digitises the images, and software turns the motion of the particles into a velocity field. According to Chemineer, combining CFD and DPIV provides the most accurate evaluation possible and DPIV can also be used to validate the models used in CFD.

Laser Doppler Anemometry (LDA) is a technique that measures mean velocity and turbulence data with pinpoint accuracy. It relies on the interference pattern formed when two laser beams of the same wavelength cross. As a particle passes through the intersection, it reflects light at specific frequencies that depend only on the velocity of the particle and the interference pattern.

The resulting data can be used to build a highly accurate picture of the velocity within a small volume of fluid. According to Chemineer, one of the most challenging problems in modelling mixing is to get a direct measure of mixedness. Laser induced fluorescence (LIF) is a technique designed to get round this by using materials such as rhodamine or uranine, which fluoresce under light at certain wavelengths. That can be used to track how these materials diffuse through agitated vessels and static mixers.

Tomography techniques

While they are extremely useful, all these techniques have one thing in common, which is that none of them are looking at the actual installation. But there is one technique that can be used in situ to check performance – tomography.

“The big difference is we’re actually taking measurements on the process, rather than looking at the behaviour on a model. So we’re visualising reality, rather than modelling a ‘pure’ form of the process,” says chief executive of Industrial Tomography Systems (ITS), Ken Primrose. “We get quite a lot of work validating computer models for fine tuning the parameters they’re using.”

Tomography systems in industry work in the same way as medical CAT scans using the same software algorithms. They rely on the fact that the conductivity of the material at any given point will vary according to its composition and how it’s moving.

The big difference is we’re actually taking measurements on the process, rather than looking at the behaviour on a model. So we’re visualising reality, rather than modelling a ‘pure’ form of the process

Ken Primrose, chief executive of Industrial Tomography Systems

The Mix-itometer from ITS uses sensors to measure the concentration and mixing index at more than 200 locations throughout the process to generate a cross-sectional view of what’s happening. The resulting slices can then be stacked to generate 3D images.

The technique is widely used in everything from nuclear waste and catalyst production to food and personal care, and is also used by some of the biggest mixing suppliers to validate their ongoing design improvements.

The system can be used in a number of ways. For instance, ITS can come in and carry out an assessment. Alternatively, users might want to keep a system on-hand for periodic checkups, or may even want it built permanently into the production equipment.

“We can tell people exactly when something is mixed, which, unless you’re sampling something everywhere, is hard to pin down otherwise,” says Primrose. In slurry storage applications, for example, it means users can switch off the mixer when it’s mixed and turn it on when it start to settle. “That’s how you minimise energy consumption,” he adds.