Process Analysis Technology Why PAT Must Come Out of the Silo
The 11th colloquium on Process Analytics was held in the historical dome hall of TU Vienna under this motto. The nearly 200 participants were unanimous: process analytics must come out of the ivory tower and finally shed the perception of complexity. How is this to happen? How will the much admired trialogue between device manufacturers, science and industrial users be wrapped up finally?
Irrespective of whether petro-, polymer chemistry, biotechnology or classical synthesis methods — process analysis technology can optimize methods, increase yields and reduce costs. There is consensus among all the participants that the only thing left to be decided is the uses, methods like Raman-, MIR-, NIR- or UV/VIS spectroscopy have already been introduced extensively.
But PAT remains stuck. A lively, intensely discussion-loving community which pursues many approaches and options notwithstanding, one experiences a practical shock as soon as one looks at the chemical industry.
Spectroscopy is the Predominantly Used PAT-Technology
The lion’s share in the field of installed measurement points is classical process measurement technology, i.e., pressure, temperature, through-flow and filling level. The number of measuring points in the companies is increasing very slowly.
There are more than 80 PAT-compatible analysis methods, starting from gas measurement and gas warning technology, refractometry, density-, viscosity- and humidity measurement technology, para-magnetic oxygen measurement and photometry right up to chromatography and spectroscopy. Of course, among the complex methods only spectroscopy is used predominantly. And a critical view of the scenario shows a conspicuous, universal overweight, as far as method development is concerned. A circumstance that even PAT experts find alarming.
Making PAT understandable – but how?
The value addition of PAT is frequently not seen clearly, especially in the context of complex analysis methods and sensor systems, because it is still trapped in academics, emphasises Dr. Thomas Steckenreiter, Director Support und Services at Bayer Technology Services, repeatedly.
It would not be any easier in the future either, because the link between PAT and Industry 4.0 anticipates even more complex argumentation chains. Moreover, the options that smart sensor technology unfolds cannot be explained in three simple sentences.
How to make clear the benefits of PAT?
Good advice is therefore expensive. How can PAT come out of the academic silo? And how does one convey the benefits of the system, whose R&I scheme is at least as complicated as that of a sub-system? These questions are being handled by the organisation team since the inception of the Process Analytics Work Group managed by Dechema and GdCH, and they formed the core theme of the event in Vienna.
Best Practices are the Key
Bombardment of scientific knowledge through Power Point presentations and endless slides that mutually connect spectra, the content of which can be understood only by analysis experts, do not constitute the crux of the matter, instead they are only the reason why no one understands PAT essentially.
The organisers recognize this and have avoided it this time. An important step is practice-oriented contributions like the one by Dr. Ulrich Schünemann, who is entrusted by the Engineering Department of BASF with the implementation of process analysis technology and for which “infighting at the production site” has become a daily routine.
PAT is Really Important for Basf
“Process analysis technology plays a crucial role in the technology strategy of BASF”, he emphasises. This approach notwithstanding, in Ludwigshafen too, everything that glitters is not gold. For Dr. Michael Kloska, who is at the forefront as Head of the Specialized Centre of Process Analysis Technology BASF, the introduction of PAT in already existing companies is primarily to convince the company managers.
Continuous processes are normally well orchestrated, he emphasises. For batch processes, the installation costs of the fair stands are frequently cited as the reason for deciding against process analysis technology.
When will mini-formats arrive?
Perhaps miniaturization is an approach to support the PAT systems rightly. Because in the field, PAT is frequently installed in the analysis agency, and this requires not only infrastructure but also manpower for maintenance capacities, if e.g., the sample extractors pack up.
Especially in mass spectrometry, sample feeding rapidly assumes proportions of an Achilles Heel. Even measurement delays render the implementation of an adaptive process regulation difficult. All these are reasons for which the Krohne measurement technology specialist used mass spectrometry for his latest research project to subject it to a shrinkage.
“The miniaturization of sensors enables the implementation of complex analysis field devices for online or at-line application”, says Winfred Kuipers of Krohne Messtechnik in Vienna. Sample extraction is still a problem, but with Micro-Electro-Mechanical Systems (MEMS) Krohne has a rich pool of experience which one is now using to marry MEMS chips with sensor technology.
The proof-of-concept has already worked, and the team around Kuipers has succeeded in reducing the size of the total system to a 19" casing, wherein MS itself is half the size of a one-cent coin. The long-term objective is to achieve the size of a shoe box, says Kuipers.
A New Platform for Embedded Spectrometer
Away from the analysis agency, the technology enterprise Tec 5, which specializes in things “embedded”, has developed a new platform for embedded spectrometer systems called “tecsAAS“, which performs acquisition of spectral data, data processing and communication with the process interface independently.
Designed for individual measuring points in which a second measurement channel can be used for online reference, this platform is not connected to sensor types or spectral ranges.
A modular software (firmware), comparable to the process software Multispec Pro II, it permits parameterization of the individual steps of the measurement data recording and processing right up to evaluation on the basis of the chemometry models created on the PC.
A classical programming is normally nor required, stresses Hanns Simon Eckhardt of “tec5“. The only drawback of the clever idea: the presented use sample comes from agrarian technology, also the technique is available for the chemical industry. That shows clearly that chemical industry has to speed up PAT-applications.
Spectroscopy for Measuring Quality and Quantity
Even the exotic areas of spectroscopy are booming. This is confirmed by the use example of the Vienna based Institute for Chemical Technology and Analytics. Jointly with the OMV Refinery in Schwechat, researchers are developing a gas sensor on the basis of a quantum cascade laser, which is already being used in a pilot hydrogenation plant for research purposes.
The technology is interesting because one can measure concentrations not only qualitatively but also quantitatively, which is not so easy at present. This individual case involves measurement of the hydrogen sulphide concentration in the process gas flow of a hydrogenation plant – and that too in the sub-ppm range, as Harald Moser of TU Vienna emphasises. Polymer
Production and Process Management in Yeast Fermentation
The proof that the sensitivity of the sensor correlates sufficiently with the concentration has already been provided. Only a risk analysis remains to be performed.
The next step then is the test in the refinery under Atex conditions. How useful will it be to common users to store analysis jobs performed in the lab earlier directly on the assembly line, is illustrated by two different examples:
The continuous polymer production and the process management in yeast fermentation at a brewery. The PAT instrumentation of continuous processes has not matured in the least, says Peter Mayo, who is engaged at Borealis Olefine in incorporating Time Domain Low Resolution NMR in the process flow.
In the matter of determining the quality parameters of polyolefin powders and pellet flows, the technology in quality assurance and development laboratory is standard. But time is money in polyolefin production.
PAT to be on the Way
A typical polypropylene plant produces 30 to 45 tonnes per hour, estimates Mayo. It is easy to imagine how much money the plastic producers stand to lose if the laboratory result yields an off-spec analysis. Wait times of six hours for one lab test on Xylene dissolution or even 24 hours, till the lab scientist determines the mechanical properties of the blend, should belong to the realm of the past with the adapted TD NMR.
A similar high potential also exists in the optimization of brewing processes, which indicate high capital commitment on account of the long process times. Another example of the colloquium which shows: values that one can raise with PAT are more than sufficiently. The responsible persons must now accelerate the popularization of the advantages.