By Thomas Schmid
Industry specialist Dr. Ian Benson answers Tobacco Asia’s questions about quality assurance in processes and how companies can improve their practices.
Benson is currently managing director of his own consultancy business, Ian Benson Consultative Solutions Ltd (IBCS Ltd), specializing in providing support to instrumentation suppliers looking for focus in their business strategy, market planning, product evaluation, and how to best gain market acceptance. But he likewise assists instrumentation end-users offering hands-on advice on process instrumentation strategy, vendor selection, and assessment.
Until mid-2019, Benson was the global marketing director and vice president for all four business groups within NDC Technologies, an operating company within Spectris and a global leader in the design, development, and manufacture of online measurement and control instrumentation solutions for a broad variety of industries, including tobacco. He left NDC in the summer of 2019 after deciding that he henceforth wanted to spend more time supporting businesses on his own accord.
After earning an honors degree in chemistry from the University of Bristol in 1975, Benson subsequently went for a PhD in organometallic chemistry at the same institution, extensively researching spectroscopic methods and also publishing several papers during that time. From 1978-1981 he worked as a product development chemist at Ilford Films before joining NDC, where he put in place the company’s applications engineering resources. Benson together with his team also developed a range of rapid non-contact measurement methods for moisture, nicotine, total sugars, glycerol, and volatiles in tobacco. It was this work that laid the ground for his decades-long close ties with the tobacco industry. Eventually winding up with NDC after a few more twists and turns in his long career, he is today regarded as a world expert on the implementation of online NIR measurements.
Tobacco Asia (TA): What are the most important processes that require sustained and stringent quality assurance procedures?
Dr. Ian Benson (IB): The primary process and in particular the blending of components are critical areas for quality assurance (QA). To achieve a consistent and accurate blend, the mixing of components must be precise, requiring knowledge of the tobacco moisture content to calculate dry weight. Blending in primary is important as this is what gives the in-pack cigarette tobacco constancy. Final blend moisture is most critical, because it effects processability.
But I also refer to the additions making up the final blend. There are processes which again benefit from accurate control and principally moisture measurement, such as dry-ice and HTD expansion. And while recon sheet is added earlier in the process, online QA instrumentation also is important there.
TA: And perhaps a few words on secondary processing?
IB: Well, most of the impact that online measurement can have on tobacco quality has already taken place in primary. But in secondary the rod will be measured for density and weight, and also dense end formation with appropriate instrumentation that has been integrated into the maker. QC likewise is a critical requirement in secondary to sample production and understand if the rods meet the many specified parameters, including draw, weight, delivery, diameter, moisture and possibly nicotine and sugars and even humectants. Some of that testing will necessarily be offline using lab instruments. However, makers running at speeds in excess of 20,000 sticks-per-minute do have sophisticated [in-built] control capability for weight, dense end detection, and density, which are key online measurements.
TA: What equipment is an indispensable “must-have” to adequately handle QA in each of these processes?
IB: The instrumentation required for the various processes does have many similarities. One key expectation today is that online instruments are non-contact (with the product). Infrared measurements are prevalent throughout the industry irrespective of the processes outlined, with the exception of secondary processing.
But the number one measurement in all processes remains moisture or total volatiles. Online moisture measurement is dominated by near-infrared instrumentation, of which there are many suppliers. Microwave technology is also used, but sample presentation and the need, in many instances, for some form of contact has diminished its deployment in tobacco processing.
Apart from the basic requirement to accurately measure moisture, a user should satisfy himself that the instrument is not affected by any process conditions that have nothing to do with moisture content but may be interpreted as so by the instrument, for example, tobacco distance from the instrument and variations such as tobacco temperature, color and, finally, atmospheric humidity, and temperature. The other key attribute to be confident about is how stable the instrument is as time passes. If the instrument changed with time due to electronic or optical drift, it could affect repeatability when measuring the same sample multiple times. By contrast, the measurements of nicotine, sugars, and humectants, while more mathematically complex and requiring more spectral information, are easier to validate because the likelihood of sample constituent levels to change with time is small… perhaps with the exception of humectants.
TA: What could the potential consequences be if QA in any of these processes is lacking or handled inadequately?
IB: Poor QA would be a major issue in the modern world of tobacco processing. If there is a lack of accurate and dependable moisture measurement in primary processing, it will impact output in the secondary process. This directly impacts productivity and bears the risk of making an out-of-spec product. The reason why so many companies nowadays make measurements further back in the primary process is to ensure that tobacco is ending up at the next processing stage within moisture specification. That makes it easier to control subsequent processes. There are many other negatives about poor QA. But poor blend control, inefficient tobacco expansion, poor effectiveness of mechanical processes, such as stem rolling and cutting, shortening of tobacco due to it being too dry and therefore suffering mechanical damage are just a few of these key issues.
TA: How important is it that quality assurance equipment is always up-to-date so to get the most out of it?
IB: I am not certain that QA equipment being up-to-date and the latest model is as important as making sure the equipment is doing what is was purchased for. In fact, buying the latest instrumentation is more about avoiding product obsolescence and also taking advantage of improved measurement capability, either with existing measurements or the addition of new ones. Take moisture as an example: As instrumentation has evolved, some manufacturers have focused on improving the measurement so that it relates more directly to the oven test, reducing calibration sensitivity to different tobacco blends, which may correspond to color; though not always. This would be an incentive to upgrade, provided the benefits are there for the end-user.
Important, then, is to ensure that the instrument used is calibrated properly against the primary reference method and that it is periodically checked through a statistically sound routine to make certain it is “still telling the truth.” Even today, instruments can still show calibration drift with time, and this can be due to instrument design. If this drift is not detected and corrected, the instrument is not giving a dependable control signal, massively reducing the benefit of measurement.
TA: Are you saying that some tobacco companies are neglecting to keep their quality assurance equipment calibrated correctly?
IB: I am certain there are many companies that are not getting the best out of their instruments. This can be due to poor calibration initially, but also changes in the instrument that are not picked up by the in-house QC checking routine if it is not diligently applied, assuming one is even in place. But, it also can be down to poor routine maintenance. I do recall visiting one factory where the online instrument was reported not to be working as well as it used to. Closer inspection showed that the light source, so critical to any infrared instrument, was nine years old and emitting virtually no light!
TA: Why does this sort of thing happen?
IB: It is not usually blatant negligence but rather commonly due to a multitude of different factors. For example changes in personnel are ever more frequent and the new person may just not know how an instrument works and what routine maintenance is required. Calibration and testing strategies can be designed by a central authority, but the local plants may not have the time or interest to following procedures. Some plants may have no ability to do formal testing and instrument checking, or perhaps no primary reference methods on-site. I would also say that instrument suppliers may undersell the need for interaction with a QA instrument, stating that it is “install and forget.” But in reality, that is not true -- or sensible in any critical process control loop. The unit may not require any routine maintenance, but it is vital to check whether it is performing properly and that its calibration remains valid.
TA: Looking at “Big Tobacco”, how have the multinationals’ partnerships with key suppliers in formal product evaluations changed over the years?
IB: In the past -- and I am really going back to the 1980s here -- there was tremendous opportunity for cooperation between instrument suppliers and [the large multinational] manufacturers to help each other. Instrument companies’ products needed validation to be sure that the end user would get the benefits intended. The big companies dedicated huge resources to this practice. And both parties gained from it. It was a win-win situation.
Instruments improved greatly as they were put through the mill by dedicated test specialists and scientists in the tobacco companies. When it came to validation, an instrument would literally be pulled to pieces, checking it in detail and making sure it really would do the job. Very important, this also would yield feedback on the issues and improvements needed.
Sadly, over time the tobacco companies have reduced these resources, and in some cases more or less eliminated. That makes instrument evolution and improvement more difficult. The tobacco industry has been under cost pressures and these “softer” activities are inevitable casualties. And while a level of cooperation commitment may still be there, the in-house accessible technical resource is less so.
TA: Have you observed similar partnership changes between smaller or national tobacco companies and their key suppliers?
IB: Instrument evaluation in smaller or national companies has been a mixed bag of activity. Some of them are taking it very seriously, while with others the attention to detail can vary from superficial to taking comparative test samples for QC checking with little thought to correspondence to what that gauge was reading when the sample was taken, making it hard to draw any solid conclusions. You could say this lack of resourcing and awareness in [smaller and national] companies means that supplier decisions may be taken too lightly and revert solely to cost.
TA: Where is QA in processes headed in the medium to long term in the light of increasingly restrictive tobacco product legislation?
IB: I would think that in an increasingly regulated tobacco world one needs to be getting things right from the outset and quantifying this as accurately as possible. Whereas in the past the online measurement of nicotine was of little interest mainly because there was nothing one could do to modify the levels, the need now perhaps is about traceability and to ensure that nicotine does not exceed a given level. At least then you have a chance to modify the end product composition, perhaps through blend adjustment.
As products take on a more pharmaceutical stance in terms of production methodology, measurement and control to ensure variability is minimized thus becomes more important. This is especially true of new-generation products such as heat-not-burn. And with traditional combustibles becoming more regulated in terms of additives and composition, I think it will remain just as important as ever to measure product, ensure its consistency, quality, and conformance.
TA: What are your recommendations for a medium-sized tobacco company looking to enhance its QA procedures? And is it worth the investment?
IB: I think it would be worthwhile for any company interested in improving QA to talk to someone who can advise them on best practice and what really needs to be done to make a significant improvement. They must avoid falling into the trap of excessive instrumentation with a huge price tag and little chance of success. Existing process equipment should be audited by an expert, who would then address the company’s objectives and recommend what would help it to get there. I always think the Pareto rule applies here. That is, 80% of issues will be solved with only 20% of the possible cost!