Best Practice for Water/Particle Separation

A look at particle characterization technologies that can improve performance

Separation of water from fine tailings in the minerals industry has been a long-standing challenge. To speed water recycle, the density of dispersed particles is increased to form sediments. Current treatments involve highly engineered types of polymers and chemicals to promote flocculation and modify rheology for the specific slurry being processed. By inducing flocculation, scientists and engineers adjust the particle size distribution to maximize the solid-liquid separation. Since flocculation is a dynamic process with changing particle size and concentration, representative sampling and offline analysis are difficult. Offline samples are typically manipulated through dilution or dispersion preparing them for measurement. This can alter or destroy fragile particle structures and offline measurements often cannot be applied to make real-time process optimization and control decisions. With established in situ particle characterization technology, scientists and engineers can quickly measure the particle phase behavior in process without pulling samples.

In-process particle measurement tools can help optimize thickener and flotation performance and improve dewatering of oil sands mature fine tailings.

In-process particle size characterization enables the lab researcher and field operator to track the entire dynamic particle system in hydrometallurgical processing gravity thickeners or oil sands mature fine tailings (MFT) streams at standard operating concentrations, temperatures and pressures-without sampling or sample dilution. As the source of the incoming stream varies in concentration and mineral components, real-time particle size, shape, and count measurements provide predictive or feedback measurements to enable a controlled response. Scientists and engineers can understand how the particle system responds to changing process parameters and optimize the type of polymer additive, the dosage, and the shear to control solid-liquid separation and downstream throughput.

In-process In Situ Particle Characterization
In-process particle size measurements improve process understanding by avoiding significant measurement errors resulting from routine offline sampling and sample preparation methods. ParticleTrack and PVM measure real-time changes to particle size, shape and count without the need to take samples. This enables real-time optimization of process performance, whether measuring in laboratory beakers, pipelines or process environments.


PVM	ParticleTrack

Focusing on Flocculation

In-process particle measurement tools such as ParticleTrack with FBRM (Focused Beam Reflectance Measurement) technology and PVM (Particle Vision and Measurement) can help users optimize water recycle and reduce MFT. Scientists and engineers apply these tools to understand, optimize, and control the flocculation process and improve dewatering efficiency in real time. PVM technology is a probe-based video microscope that provides high-resolution images of particles and droplets as they naturally exist in process. FBRM technology tracks the rate and degree of change to particles and particle structures as they exist in process, and provides a distribution tracking particle count and dimension in real time.

Engineers and scientists identify inline particle size distributions, which correspond to the ideal flocculated state, then use real-time measurements relative to their target distribution to ensure consistency in solid-liquid separations. Statistics such as mean or number of particles measured in specific size classes are trended over time, enabling users to troubleshoot unexpected process changes and variability in the incoming slurry.

Beaming In on Process Optimization Measurement for optimization in real time–FBRM (Focused Beam Reflectance Measurement) is a highly precise and sensitive technology, which tracks changes to particle dimension, particle shape and particle count. More than a wide detection range from 0.5 to 2,000 ?m, measurements are acquired in real time while particles are forming and can still be modified enabling process optimization and control. No sampling or sample preparation is required—even in highly concentrated (70% and higher) and opaque suspensions.

The FBRM probe is immersed into a dilute or concentrated flowing slurry, droplet emulsion or fluidized particle system. A scanning laser is focused to a fine spot at the sapphire window interface (Figure a). A magnified view shows individual particle structures will backscatter the laser light back to the probe (Figure b). These pulses of backscattered light are detected by the probe and translated into chord lengths based on the simple calculation of the scan speed (velocity) multiplied by the pulse width (time). A chord length (a fundamental measurement of particle dimension) is simply defined as the straight line distance from one edge of a particle or particle structure to another edge. Thousands of individual chord lengths are typically measured each second to produce the Chord Length Distribution (CLD) (Figure c). The CLD is a “fingerprint” of the particle system, and provides statistics to detect and monitor changes in particle dimension and particle count in real time (Figure d).

Unlike other particle analysis techniques, FBRM measurement makes no assumption of particle shape. This allows the fundamental measurement to be used to directly track changes in the particle dimension, shape and count.

How FBRM works.

In-process Measurements

As mentioned previously, separations of fine solids from a water stream is a longstanding and well-documented challenge in the mining (thickening) and oil sands (mature fine tailings dewatering and drying) industries. Flocculation is a common separation improvement technique to improve water recycle and control fines accumulations. In all cases, the dewatering efficiency is directly proportional to the incoming particle size distribution, particle population, shear, polymer type, polymer dosage, and the flocculated particle strength, porosity, and aggregate size. Researchers from Suncor Energy have shown that:

Ideal polymer dosage levels correspond with the incoming solid content (particle size and count), and variability in the particle-to-polymer ratio results in under dosing or overdosing the tailings system. This results in high retention in fine solids or low porosity flocculated solids and reduced dewatering.

Applying the correct shear during the flocculation process is necessary to optimize the particle size distribution for maximum dewatering. Insufficient or excessive shear produces a particle size distribution with reduced permeability and non-ideal dewatering.

Choosing the correct polymer for the incoming tailings can maximize dewatering effectiveness.

By implementing in-process particle measurement technology, real-time flocculation performance is measured atfull process concentrations allowingoptimization of MFT dewatering at the bench or in the field. Figure 1 depictshow in-process measurements allow users to measure changes to the particle size and number with increasing polymer dosage. The distribution progressively shifts to lower counts and larger size as the dosage increases.


Figure 1—Inline ParticleTrack distributions measuring flocculation over time.		Figure 2—Distributions tracking flocculation and dispersion over time: top—initial distribution; middle—after flocculant addition; bottom—after flocculant addition and extended shear.

By implementing in-process particle measurement technology, real-time flocculation performance is measured atfull process concentrations allowingoptimization of MFT dewatering at the bench or in the field. Figure 1 depicts how in-process measurements allow users to measure changes to the particle size and number with increasing polymer dosage. The distribution progressively shifts to lower counts and larger size as the dosage increases.

When shear is introduced to improve the polymer dispersion, the initial aggregate size distribution often increases. However, depending on the strength of the flocculants and intensity of shear, it may result in dispersion and solids breaking up over time resulting in a distribution that shifts to smaller particle counts. (Figure 2). These fine dispersed particles have a negative impact on dewatering and optimizing the changes in the process and maximizing the degree of flocculation improves downstream dewatering.

When combined to track a batch flocculation process, ParticleTrack statistics and PVM images identify four stages of the shear progression curve of flocculated MFT (Figure 3):

1. Incoming steady state MFT dispersion and poor water release.

2. Polymer dosage followed by flocculation.

3. Floc breakdown and maximum dewatering.

4. Floc shear and dispersion reverting to the MFT and less water release.

Optimization and adjustment of the dosage and shear is critical to improve the dewatering performance for each incoming MFT particle distribution.

Figure 3—PVM images and ParticleTrack trends measuring four stages of flocculation and dispersion.

The real-time mean chord length measurement over time describes the aggregate size distribution based on the aggregation efficiency and breakage rates. Dewatering performance is predicted based on the corresponding particle size, flow rate and hydrodynamics. Researchers at CSIRO Australia have used FBRMtechnology to follow flocculation in turbulent pipe flow to quantify the kinetics of aggregate growth and breakage. They have shown that varying the slurry flow rate for a fixed polymer dosage (Figure 4) greatly affects the extent of aggregation and the degree of breakage at longer times. By measuring the real-time particle dimension over a range of conditions, a population balance model for flocculation canbe applied using the outputs in hydro-dynamic modeling to optimize operating conditions in pipes, channels or thickener feed wells.


Figure 4—The effect of slurry flow rate on the flocculation mean size (chord length, μm) over time.		Figure 5—ParticleTrack is applied to predict conditions for the optimum settling rates and reaction time after flocculant addition.

While aggregate dimensions are taken as indicative of likely settling properties, the corresponding settling rates are actually found to diminish at longer reaction times (Figure 5). This is attributed to flocculant deactivation, which prevents additional growth at longer times, reducing dewatering performance.

CSIRO has also shown a complex mechanism between solids concentration and flocculation efficiency. For a fixed dosage, a maximum relationship exists between floc particle size and solids concentrations (Figure 6). By tracking the rate and degree of change to particles and particle structures in real time, engineers can optimize the flocculant dosage/solids concentration ratio to maximize fines capture and dewatering efficiency.

Finally, the particle size, shape, surface area, and surface chemistry can accelerate or slow growth kinetics and reduce the strength of the aggregated structure. CSIRO reported that the flocculation kinetics can vary significantly for different feed types (Figure 7). CSIRO has applied this approach to the study of MFT flocculation through a 2010 study partially fundedby CONRAD-the Canadian Oil Sands Net-work for Research and Development.


Figure 6—Mean dimension tracking the efficiency of flocculant dosage with increasing solids concentration.		Figure 7—Effect of incoming minerals on the efficiency of a given flocculant.

Improved Control

By using in-process particle characterization tools, flocculation kinetics are measured in real time. This provides scientists and engineers with information to optimize the polymer type, dosage, and shear based on the variation in the incoming particle size, concentration and flow rates. ParticleTrack and PVM enable scientists and engineers to:

Understand how the particle system responds to changing process parameters;

Optimize particle flocculation to improve downstream separations performance; and

Control particle distribution to achieve consistency, process throughput and process stability.

Anjan P. Pandey is a senior technology and applications consultant for Mettler Toledo, with extensive experience in particle characterization projects. Benjamin Smith, currently head of marketing for Mettler Toledo, is a chemical engineer with a background in particulate processing and separations. Terry P. Redman is a principal technology and engineering consultant for Mettler Toledo. Readers who wish to obtain a unedited copy of the original white paper, with references included, may do so by visiting www.mt.com/wp-mining-separation.