Today’s availability of reliable real-time particle sizing technology presents an opportunity to transform the efficiency of a wide variety of manufacturing processes. There are those who remain wary, however, perhaps unsure of the best way to engineer a solution, or unconvinced of the potential benefits. In this article we take a look at what real-time laser diffraction particle size analysis has to offer and address some of the issues most commonly raised by those considering a switch from laboratory to process analysis, focusing on the practicalities of making a change.

Understanding the Potential Benefits of Real-Time Measurement
The technique of laser diffraction has become the method of choice for particle size analysis across the majority of solids handling sectors. Fast, non-destructive and amenable to automation, it is efficiently harnessed in real-time systems for process use, in-, at- or on-line, as well as in powerful highly productive laboratory instrumentation. This presents users with the choice of how best to apply the technology to meet their analytical needs.
    Laser diffraction particle size analysis is often first applied in the laboratory, to support research and development, and for QC. Laboratory measurement is highly reproducible and generates secure particle size distribution data that can be used to, for example: improve product performance; develop better manufacturing processes; and sensitively detect out-of–specification material. Such tasks are eased by the capabilities of modern instrumentation which has over several decades been refined to closely match industrial requirements. The best laboratory analyzers now provide embedded support that guides the user through every step of the measurement process, to lighten the analytical workload and ensure the highest quality data.
    The speed and ease of use of laboratory systems means that they are also used to support routine plant operation, for basic plant monitoring and for troubleshooting. This can be a successful approach for processes with slow dynamics that change infrequently and are easily maintained at steady state. However, because sampling, and the time taken to transfer and measure material, introduce both a potential source of error and a time lag, laboratory analysis can be less than ideal for those aiming for exemplary, and automated, process control.
    At-line systems integrate analysis with process operation and are a first step towards tailoring analysis more closely to operational requirements. With an at-line system operators can ramp up analytical frequency to meet processing requirements, during start-up or grade change for example, and lag times, between sampling and the receipt of results can be significantly reduced. However, it is on- and in-line particle sizing systems that accelerate measurement to the point of real-time monitoring, thereby providing the most secure foundation for fully automated process control.
    Laser diffraction particle sizing systems specifically designed for process analysis measure up to four complete particle size distributions per second, so measurement frequency can be exactly matched to monitoring and control requirements, even for rapidly changing processes. Once real-time monitoring is in place, plant performance can be continuously assessed and it becomes possible to instantly observe the impact of changing a process variable, a benefit that adds considerable value during development and into commercial manufacture. At the pilot stage continuous monitoring accelerates learning and the extension of process understanding, while in production it supports the ongoing optimization of plant performance, especially during periods of transient operation – at product changeover, for example, or when starting up. For manufacturers, the tighter, often automated, control achieved with real-time analysis therefore typically delivers substantial economic gains.
    In summary, real-time monitoring brings value by supplying the timely information flow required for truly optimal process control. Automated process measurement simultaneously offers an opportunity to substantially reduce the manual effort required for analysis, and also for process control, further enhancing economic return.

Assessing Particle Sizing Technology
One of the primary concerns for those drawn by the economic promise of installing an on-plant system is reliability. Across the manufacturing industries operating plant is being judged against increasingly stringent reliability criteria with maintenance simultaneously being reduced as far as is practicable for safe and secure operation. Highly reliable process instrumentation that requires minimal manual input is essential and this should be reflected in the equipment design.
    The best process analyzers are designed from the ground up specifically to meet the demands of reliable 24/7 operation, in what can be a harsh environment. Modified laboratory systems, in contrast, tend to be less successful but reliability is best judged by results. Some laser diffraction on-line particle sizing systems have now been operating continuously, on abrasive duties, with minimal maintenance for over a decade. Evidence to support reliability claims and dispel concerns about this issue is therefore readily available for those looking for reassurance.
    The issue of how best to interface an analyzer with the process is also a focus for attention in the early stage of a project, and justifiably so, since the design of the process interface is crucial to overall success. Sampling is now the major source of error in any laser diffraction analysis so with an on-line system it is vital to ensure that the stream is consistently and representatively sampled and that analysis has no downstream impact on the process.
    There are many proven process interface solutions to choose from, ranging from simple eductors to two-stage auger samplers for flow rates in excess of 100 teph, but this is an area that can benefit from expert support.

Switching from Lab to Line
Once a process analyzer has been installed, an important activity is to compare the results obtained with established off-line data sets. In an ideal situation the two match precisely boosting confidence in the newly installed system. Sometimes however slight differences are observed potentially undermining the application of real-time data.
    When investigating possible reasons for a discrepancy there are several issues that may usefully be considered. The first is the location of sampling points and the sample dispersion conditions applied, both on- and off-line, to ensure the measurement of de-agglomerated material (which is usually the intent). If the process analyzer is installed in a different area of the plant or positioned differently from the laboratory sampling point then the samples extracted in either case may not be identical. Sample dispersion conditions should be rigorously evaluated, against analytical aims, as part of the method development process, to ensure that the methodologies being applied are appropriate and robustly equivalent.
    Other factors that can give rise to a discrepancy include the well-know phenomenon of powder ‘relaxation’. Many manufacturers recognize that their powders change slightly over time, as they cool, away from the process environment, and/or settle. Where this is the case then it is reasonable to question whether it is realistic to expect results measured in QC, some time after sampling, to agree with those measured in situ. Furthermore there might be quite specific reasons for a change in particle size. For example, material measured on-line, at the exit of a mill, might have a finer particle size than the same material measured some time after sampling when, away from the high energy environment of the mill, irreversible agglomeration has taken place.
    Satisfactorily rationalizing any observed differences is undoubtedly reassuring and is a realistic goal. On the other hand it may also be productive to make a pragmatic judgment about how close the two sets of data should or actually need to be. The main reason for installing a process analyzer is usually to monitor deviation from a set point. It may therefore be argued that it is the ability of the system to continuously and accurately detect such deviation that is crucial rather than the precise measurement of an absolute value. In QC, in sharp contrast, the goal is to detect whether a product meets a rigidly defined specification so absolute values are vital.
    This difference in the way in which the analytical data is to be used leads some manufacturers encountering a small offset between laboratory and process data to choose simply to scope correlations between the two and modify the process set point accordingly. For example, if Dv50 (the median particle size) is the control basis for the process but laboratory Dv50 results are two microns lower than the equivalent plant measure then the set point for the plant is input as two microns higher than the specification requires.
    What is crucial when tackling issues around any offset is that it does not become a reason for failing to exploit the potential of real-time monitoring. Having access to both laboratory and process analyzers can be a major benefit here since the laboratory system can support on-line system set-up and, on an ongoing basis, help to develop any understanding needed to optimally apply all the available analytical technology.
   
Capitalizing on an Investment
The final element of switching from lab to line analysis, and the one that distinguishes the most successful projects, is to imaginatively and productively use the resulting data to really advance production goals. Real-time monitoring provides the information required to gain a detailed understanding of the process and to push the plant right up to its operating limits. Many find that the best way to do this is to automate process control, a strategy which further lightens the operational workload and delivers consistently optimal operation.

    When plants are operated manually, on the basis of periodic analysis, there is often a ‘see-sawing’ effect around the set point. The operator takes action, observes the impact of that action some time later, when results are returned from the lab, and then corrects, sometimes swinging too far in the opposite direction. Where this is the norm, the tendency will be to operate in a way that allows for a margin of error, so that variations in performance do not take the product out of specification (see figure 4).
    In contrast, with real-time measurement and automated control in place it becomes possible to precisely control the plant, exactly where required. The need to compensate for poor control by operating conservatively is removed and the plant can be pushed much harder. This change typically translates into benefits such as more consistent product quality coupled with lower energy consumption, or higher throughput with the same energy consumption. In general those that derive maximum return from real-time measurement not only automate control but tailor control strategies to specifically target the long-standing plant issues and bottlenecks that impact most on profitability, thereby leveraging the highest possible payback.

Looking Ahead
The availability of laser diffraction particle size analyzers specifically tailored to the rigors of 24/7 operation offers those already familiar with laboratory analysis the option for wider exploitation of the technology. The steady increase in real-time process analysis means that questions relating to reliability, economic return, installation, the reconciliation of off- and on-line data, and optimal use of the resulting data stream can be answered with confidence, providing reassurance for new users. Into the future the trend towards continuous process monitoring, wherever the technology is available, seems likely to continue, increasing access to full automation and better process control. However, reliance on laboratory systems looks set to remain, for gold standard QC and for ongoing research. By harnessing laser diffraction technology through from lab to line, users can tailor the information supplied to optimally meet needs across the manufacturing lifecycle, thereby maximizing the value of this vital particle characterization technique.  
    
    Carl Levoguer is product marketing manager, Malvern Instruments. For more information, call +44 (0) 1684 892456 or visit www.malvern.com.