Using Real-Time Analysis to Optimize Process Development

Although the focus is on optimizing the operations of a commercial manufacturing plant, the real efficiency lies in getting accurate results at the first trial itself. The development, streamlining and acceleration of efficient product design creation relies on the real-time analysis of process parameters.

This article evaluates the benefits of devoting time and money for real-time process measurement in pilot studies.

There has been an increase in the use of real-time process analysis systems in manufacturing units. The six sigma model proposes the implementation of automated analysis and control in manufacturing environments in order to minimize the variability of operations and the inefficiencies caused by it.

By automating the manufacturing process, the quality of the product improves significantly and the cost of production is drastically reduced. The economics work in favor of automation since the savings achieved through automation exceed the amount invested in it.

At the pilot scale, the decision to automate may not be easy to make from a financial point of view as the potential advantages are considerable but they are not instantly visible or easily quantifiable. For more rapid and efficient commercialization, it is important to maintain consistent operations and efficiently correlate causes and effects so that both process and product development processes are simultaneously improved. This increases the importance of time spent on R&D for many sectors.

This article explores on-line particle analysis from the perspective of making an investment in real-time measurements on a pilot scale. Various case studies demonstrate different ways in which these systems contribute towards improving process development, which results in more profit.

Pilot Plant Trials

The focus is on developing a robust process that ensures long-term profitability and consistency of production processes as the product moves from early R&D stage to commercialization. Trials at the pilot scale play a significant role in this transition. It is comparatively less expensive to operate a pilot plant, which represents a specific part or the entire existing or proposed full-scale plant, to testing the viability of a proposed processing solution.

The use of the term ‘relatively’ denotes that in most scenarios, pilot scale trials run into huge expenses, and this investment should be considered as cash spent now for returns in the future rather than expecting immediate returns. The important factor is to maximize the efficiency. Keeping in mind the contributions of real-time analysis, the following goals of pilot scale study may be considered:

  • Identification of the process design that is highly cost-effective
  • Establish the practicality of the new technology
  • Gathering representative samples for testing of the product
  • Defining the operating envelope for the process, which includes the set of conditions that ensures that the product conforms to the specifications
  • Establishing an effective control strategy

Achieving the above goals at the right time requires a rapid and complete understanding of the process. The pilot phase may ensue directly after the basic lab investigations, just to prove the feasibility of the product design. At this stage there is minimum knowledge of the various manufacturing options, which need to be discovered quickly.

Role of Real Time Measurement

The experimental efficiency and control of the pilot plant can be significantly increased by real-time measurement of key processes or the product parameters. For instance, in a trial designed to evaluate the optimal milling conditions, it is useful to compare the use of periodic particle size measurement with real-time analysis since the particle size defines the product characteristic.

A snapshot of the process is provided when a sample is taken each time for periodic, off-line analysis. This involves extraction of the material from the plant and conducting the analysis at a different place and finally returning the results. There is an unavoidable time lag between sample extraction and the acquisition of results.

In the milling process, there is an instant impact of the exiting particle when the mill parameter is modified. The exited particle is sampled, worked upon and results are provided almost an hour later. The effect of the initial change can be accurately determined through careful correlation, but precise quantification of results is possible only after multiple samples are evaluated.

In case of interrupted operation of the mill, it would be difficult to distinguish between the baseline steady state variability and the impact of the change. In comparison, real-time or continuous monitoring of the exiting material particle size proves beneficial for several reasons.

One of these benefits is the convenient observation of plant operations by operators which allows hassle free operation and the establishment of a more accurate baseline for experimentation. Gathering representative data that can be used for product testing is also much easier when a steady state of operations is maintained.

The most significant benefit is the immediate tracking of results whenever a change is made. Statistical quantification of the cause and effect can be done rapidly through continuous process analysis. Here, the rate of experimentation depends only on the process dynamics, by the time taken by the changes to trickle through the plant.

Real-time process analysis enables rapid and efficient validation of multiple experimental conditions and rapid and complete result acquisition. With real-time measurement, achieving a steadier pilot plant operation and accurate capture of even the most minute process changes is possible. This capability speeds up and improves process development work and also enhances the conviction in proposed solutions.

The potential benefits of real-time process analysis are described below.

Case study: Producing a Suitable Emulsion for Separator Efficiency Testing

Developing practical solutions for improving the efficiency of crude oil processing is the principal aim of a team from Universidade Federal de Itajuba (UNIFEI), Brazil.

One of their research areas is the evaluation and enhancement of oil-in-water emulsion separators. Feed emulsion generation is an important aspect in this project. These emulsions are similar to those found in oil processing. A maximum flow rate of emulsions up to 18 m3 per hour is maintained at the pilot scale unit for this project. The operating pressure and the concentration of oil are varied according to the requirements of a specific application.

The project is based on the consistency of emulsion production. The assessment of the separator efficiency and improving the process solutions is only possible by maintaining a stable feed. Here, real-time analysis of the particle size of the exiting material in the emulsification unit ensures steady operation of the test bed. The data obtained from a test trial that evaluates the ability of a laser diffraction on-line analyzer in monitoring the droplet size of the emulsion by altering the test bed conditions is shown in Figure 1.

Experiments investigating the impact of oil concentration and production pressure on the droplet size of an oil-in-water emulsion.

Figure 1. Experiments investigating the impact of oil concentration and production pressure on the droplet size of an oil-in-water emulsion.

Table 1. Experiments investigating the impact of oil concentration and production pressure on the droplet size of an oil-in-water emulsion.

Concentration (ppm) Emulsion valve pressure (bar) Obscuration (%) Dv50 (µm) Dv90 (µm)
(1) 560 3 60 25.52 74.69
(2) 1150 3 73.3 27.34 75.57
(3) 1670 3 82.1 27.84 77.06
(5) 1670 5 87.2 21.25 61.66
(6) 1670 8 93.1 13.25 43.13
(7) 1670 10 95.9 9.28 33.02
(8) 1830 10 97 8.72 32.27
(9) 1830 12 98.2 5.98 26.63

When the concentration is increased in tests 1 to 3, there is a negligible effect on the volume moment mean of the particles D[4,3] and on the Dv90, below which the majority of the particle population lies, but there is an increase in obscuration (red line). The light reaching the detector from the analyzer source laser reduces as the obscuration increases. Higher obscuration implies that higher oil concentration increases the opacity of the process stream.

The applied pressure during emulsion production is increased steadily in tests 5 to 7, which drives down the Dv50, below which the minority of the particle population resides, and also the Dv90. The oil concentration is increased further in test 8, which in turn increases obscuration to 97.5%. The pressure is further increased to 12 bar in the final test (9) to produce an emulsion with very fine droplets, with a Dv90 of 26.63 µm.

It can be seen from the time axis of the screen shot that all the tests were completed in one hour, which demonstrates the contribution of the on-line analyzer in allowing the operator to rapidly find, set and control the experimental conditions to develop a suitable emulsion.

The analysis of unstable emulsions away from the test rig could be tedious, resulting in a delay in the identification of optimal production conditions. Moreover, an ongoing process for controlling the production test bed enables better research into the efficiency of the downstream separation processes.

Case study: Innovative Equipment Design

The engineering teams at Ferrari Granulati, Italy, who is a market leader in the manufacture of marble powders and granulates, have narrowed down on a vertical roller mill (VRM) after conducting extensively assessing various mill types for producing a new marble powder. Choosing VRM based on the energy efficiency and the footprint of the equipment was challenging. Usage of these mills is not done regularly for production of powders of a particular fineness: a Dv50 varying from 3 to 8µm and a Dv98 between 15 and 50 µm.

Engineers at Ferrari Granulati collaborated with STM, a mill manufacturer, to start a project for developing a new milling circuit using a VRM for delivering in-spec material. The end component of the circuit was a classifier for the final product separation. The initial step was the installation of the Insitec on-line particle size analyzer to gather information on particle size.

Gathering information on the factors influencing the particle size and intuitively modifying the hardware was essential to tailor the mill towards an optimum performance specific for each application. The need for continuous measurements to streamline the process was proved in this case study and identified by the team. Aided by real-time particle size analysis, the team was able to quickly assess the benefit brought about by the modifications and work towards developing an optimal processing solution.

VRMs are generally not used to produce powders of a Dv98 below 50 µm because materials of such sizes tend to slide under the compacting rollers unaffected by them. Powdering in such mills depends on the thorough crushing of the particles between the table and the roller. The team therefore, focussed on developing a solution to overcome this intrinsic limitation.

Figure 2 shows a simple mechanical transducer placed on the arm of a compaction roller. This arrangement allowed the team to control the depth of the powder on the table for attaining maximum impact of the rollers.

Schematic of a milling circuit, with vertical roller mill, for the production of ÿ ne marble powders.

Figure 2. Schematic of a milling circuit, with vertical roller mill, for the production of ÿ ne marble powders.

The powder depth is an on-going measurement takenby the transducer, when the measured value dips below the optimum 15 mm, the rotational speed of the table increases automatically so that more powder is retained on the table to re-build the depth. The table typically rotates at low speeds of approximately 25 to 30 rpm. Another limitation of using very fine powders in VRMs is that they tend to agglomerate. This separator in this circuit is capable of classifying agglomerates as over-sized sent to the mill for further grinding, thereby, setting an internal recycle that consumes energy.

By installing an in-line pin mill just upstream the classifier can help overcome this limitation. The unit rotates at 300-400 rpm roughly, which is slow for a pin mill, to generate the energy required for breaking-up a majority of the agglomerates before they travel to the classifier. This set up optimizes the mill throughput and improves the energy efficiency of the unit and also avoids excessive wear.

The team had to first fix the air flow through the circuit. With proper air flow, the fallen powder is lifted back onto the table to be milled further. Another role of the air flow is the transportation of material from the mill to the classifier. Moderate air flow would suffice for coarser particles, but in case of finer particles lower air flow is required for pneumatic transport or classifier operation compared to that needed to retain material on the table.

The energy consumption is adversely affected by excess air flow, hence, the selected solution was to operate with an air flow that is enough for the classifier operation.

To transport the material from the table to the pin mill inlet a screw auger was installed. This allowed the material to be classified along with other milled material. Once the air flow is fixed for effective transport through the circuit and optimum classifier operation, the volumetric air flow rate is maintained constant during the production of any particular product for stabilizing operations.

Ferrari Granulati considered real-time availability of data through the entire development process as the primary reason for the success of the milling circuit. The company believes that in the absence of on-line analysis they would have had to compromise on the ability to develop a solution to tackle the challenge of using VRM for this application.

The milling circuit showed high stability and responsiveness in the automated production of high quality powders in a cost-effective manner. Further, the consumption of energy was minimal and the waste was eliminated.

Case study: Developing a Generic Processing Solution

The manufacturing practices followed in the pharmaceutical industry are under stringent evaluation. With a history of comparatively inefficient batch production, the pharmaceutical industry is now looking at implementing continuous processing methods in its operations in order to achieve real-time product release.

Fuelled by the Process Analytical Technology (PAT) initiative laid down by the FDA, many pharmaceutical manufacturers are working on the development of highly efficient processing strategies based on appropriate analytical techniques. Inhalable pharmaceutical formulations and tablets need to conform to a specific size during production.

Milling is the commonly used technique for processing the active and inactive particles according to the specifications. There are a number of reliable and commercial real-time particle size measurement systems that can be used in automation, which are can suitably substitute processes where real-time analysis is not yet practical.

Based on the automation potential of milling, a global pharmaceutical manufacturer has commercialized an automated milling solution with a wide scope of applications. The newly developed solution is simple, yet effective, and uses a comminutor mill. The material is broken up by the rotating blades that apply cutting and impacting actions at the same time whilst it enters the mill through the throat. The screen filters the oversized particles and sends them for further grinding, whilst it allows particles that have a defined size to pass through.

The size of the exiting particles is mainly influenced by the rotor speed, while it also depends on the specification of the screen and blade profile. The use of the Insitec on-line particle analyzer in the pilot scale trials proved the system was able to monitor the material exiting the mill continuously and in real-time.

With the solution in place, the next step was the integration of the mill and the analyzer so that the mill speed could be automatically controlled based on the particle size data. Automation of the mill operation was done through a closed control loop. The human mill machine interface (HMI) is the used by the operator to interact with the automated mill. This interface is ran on a dedicated PC.

This interface can be used to input the particle size set points for running the control loop, start and stop the mill remotely, receive results on particle size and also perform background testing.

To maintain the particle size specification, mill PLC will automatically adjust the speed within a specified control range based on the value of the particle size set point. Proportional (P) control was used to tune the loop and an average Dv50 with a 30 second rolling average was the chosen feedback parameter.

During testing the set point was lowered from its initial 58 µm value to 50 µm and finally set to the original value. At the first set point stabilization of the mill occurred after one minute roughly, then the second set point was reached at 30 seconds after the change with the final transition completed in less than 2 minutes.

The effectiveness of the control system is evident from the performance data. The capabilities of the automated solution were proved at the pilot scale trials, which led to its commercialization and applicability to any number of mills operated manually. One of the extensively used tasks in pharmaceutical production is milling of a batch of materials obtained from a crystallizer with the aim of bringing in uniformity in particle size.

Working with manual control is a tedious proposition involving presetting processing conditions according to their specifications followed by maintenance of strict control for eliminating variances throughout the batch.

Unlike manual control, automation enables the instant identification of appropriate operating conditions and making suitable adjustments according to the feed variability. In this case study, the efficiency of on-line technology and the practicality of automation were demonstrated. Based on the data, commercialization of this efficient processing solution could be pursued confidently. The scope for high returns from both the pilot scale trial and subsequent rolling of the trial across the manufacturing plant is evident.


Real-time process analysis helps improve the quality of the product being manufactured at a competitive cost in a production environment. This knowledge gain is important for pilot trials rather than mass production. For pilot trials, real-time analysis is the best way of gaining a better understanding at competitive costs.

Better control in the operation of the pilot plant and improved investigative research can be achieved via real-time analysis of the process. A successful pilot phase ensures faster and hassle free transition to profitable manufacture.

This information has been sourced, reviewed and adapted from materials provided by Malvern Panalytical.

For more information on this source, please visit Malvern Panalytical.


Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Malvern Panalytical. (2019, September 03). Using Real-Time Analysis to Optimize Process Development. AZoM. Retrieved on July 19, 2024 from

  • MLA

    Malvern Panalytical. "Using Real-Time Analysis to Optimize Process Development". AZoM. 19 July 2024. <>.

  • Chicago

    Malvern Panalytical. "Using Real-Time Analysis to Optimize Process Development". AZoM. (accessed July 19, 2024).

  • Harvard

    Malvern Panalytical. 2019. Using Real-Time Analysis to Optimize Process Development. AZoM, viewed 19 July 2024,

Ask A Question

Do you have a question you'd like to ask regarding this article?

Leave your feedback
Your comment type

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.