Engineering R&D: Sensors for a digital world
Advances in sensor technology are bringing industrially hardened detectors with on-board electronics to food and beverage processors.
Given the deterioration oxygen causes food and beverages during and after processing, manufacturers are increasing their efforts to minimize oxidation’s impact (see “Get the micro-oxygen out,” Food Engineering, September 2010). Trace amounts of dissolved oxygen can have serious consequences for product quality, but detecting and measuring molecules in the parts per billion range is a challenge, particularly when monitoring is performed in the processing area instead of a laboratory.
In discreet amounts, dissolved oxygen is even desirable, though monitoring and controlling the specific amount requires a high level of process control. Beer fermentation is a prime example. Though fermentation itself occurs without oxygen, a controlled amount of air is blown into the wort for optimal propagation of the yeast. Sensors monitor oxygen levels during air dosing and while the wort is held in a fermentation tank, but conventional sensors can be inexact, are susceptible to interference and require frequent maintenance and replacement of sensor caps and electrolyte components. CIP chemicals add to the maintenance challenge.
Fortunately, the state of the art in sensor technology is rapidly improving, thanks in large part to the demands of biotechnology. The integrity of sensor data is critical for the process control systems of pharmaceutical manufacturers and other biotech companies. Sensor suppliers have responded with a number of advances that simplify the technology, reduce costs and make devices more robust. Breweries are among the beneficiaries of those improvements. Among the leaders in advanced sensor technologies is the Hamilton Company, Reno, NV. Best known for its precision syringes and lab devices, Hamilton also operates a sister company in Switzerland where research and development in sensors, robotics and other disciplines are conducted.
Improvements in sensor technology are under the direction of Philipp Arquint in the Hamilton Company’s Bonaduz, Switzerland offices. After earning a master’s degree in software engineering from the University of Applied Sciences HTW Chur in Switzerland and a master’s in electrical engineering from the Swiss Federal Institute of Technology Zurich (Albert Einstein’s alma mater), Arquint was awarded a doctorate in sciences from University of Neuchâtel in Neuchâtel, Switzerland. While completing his doctorate, he worked as a computer engineer at Hamilton Company, where he developed an ELISA test for a pharmaceutical company. Since 2000, Arquint has served as head of sensor innovation at Hamilton Company, focusing on development of chemical sensors, calibration solutions and fittings.
FE: What improvements are occurring in oxygen sensors?
Arquint: In the old days, around 2008, the only oxygen detectors commercially available were electrochemical sensors that were passive and needed to be connected to a transmitter. Our research efforts in recent years focused on eliminating the traditional transmitter by designing a sensor with a microprocessor built into it, allowing the sensor to communicate directly with the control system.
We refer to these intelligent sensors as the ARC family. The name is supposed to suggest a bridge into the future. The newest additions are the Oxygold B for measuring trace quantities of oxygen in beer, soda and other beverages with high CO2 levels and the Conducell PWSE, which provides low conductivity measurements in water and ultra-pure water and CIP validation that caustics have been evacuated from a line before production resumes.
FE: What were the key issues you faced?
Arquint: Miniaturization of the components was the biggest issue. Three years ago, our customers had a transmitter the size of a shoebox on the wall to amplify and relay the sensor’s signal. The challenge was to miniaturize the transmitter down to the point where it would fit into a shaft with a 12mm (0.47 inch) diameter. A partner with expertise in semiconductor design produces customized electronic chips for our application. They make the electronics, and we assemble the components in Bonaduz.
Another challenge was validating that the sensors would not pose a risk as an ignition source. Regulations surrounding explosion protection are very strict. The requirements are very demanding in terms of the design of electronics when they are placed in a potentially explosive atmosphere.
FE: Why was explosion protection a major hurdle?
Arquint: The requirements for intrinsically safe electronics, which is the basis for explosion protection, collide with the fundamental requirements for miniaturization. Explosion protection requires larger components and isolation distances between interconnections. The components’ maximum operating values in terms of temperature, current, voltage and power should never exceed 66 percent of the specified values. As a result, there had to be an intelligent trade-off between miniaturization and intrinsic safety.
To reconcile these conflicting demands, we collaborated with a company specializing in development of intrinsically safe electronics and with the Fraunhofer Institut für Integrierte Schaltungen in Erlangen, Germany, which has a huge expertise in the design of analog and digital integrated circuits. The challenge is bringing both of these worlds together.
FE: What is the underlying technology used for detection?
Arquint: Some use electrochemical technology, while others are optical. Until about five years ago, optical sensors for dissolved oxygen did not exist; the only option was electrochemical. Both technologies are in use in breweries, depending on customer preference.
Optical devices, which rely on a chemical layer that is excited by light to produce a signal, remove some of the issues associated with electrochemical, such as the maintenance required to prevent clogging and the fact that they incorporate ceramics that can crack and present a danger. But optical sensors take a long time to polarize, can’t measure oxygen under certain conditions and are not as sensitive. The nominal sensitivity is eight parts per billion (8 ppb), though detection can drift over time to 15 ppb, and we don’t recommend optical for breweries. With electrochemical units, the nominal sensitivity is 1 ppb.
FE: In what types of applications are the sensors being applied?
Arquint: Nuclear cooling plants use this technology, as do many process industries operating in an industrial environment. Our primary focus, though, is biotech, and we have many pharmaceutical clients who must strictly control the fermentation process for some of their ingredients. The food industry is quite close to pharmaceutical, in terms of the requirements for sterility, temperature ranges and the growing demand for digital designs. Miniaturization of the electronics into the sensor complicates the design, since the devices must be able to tolerate the hot, aggressive media in an autoclave. They also must tolerate caustics in the cleaning cycle, and CIP is particularly harsh on optical membranes.
FE: What calibration issues are associated with the sensors?
Arquint: We calibrate them in the factory, or the customer can do the same thing in his laboratory, put it on the shelf and then simply put it into service when an older sensor fails. No recalibration is needed because, unlike passive sensors that have no memory, all the settings are stored in the sensor itself.
FE: Where are you now focusing your development work?
Arquint: The future will be connecting the sensor directly to a PLC in a purely digital connection, without any analog signals running through the plant. With all the signal noise in an industrial setting, analog signals experience interference problems. On the other hand, analog lets you take someone without any advanced training, give him a small voltmeter and send him out into the plant, where he can understand what the signal is telling him.
Digital communication using Profibus or Foundation Fieldbus is widespread in factory automation, but in process control and especially process analytics, we estimate 60 to 80 percent of applications still are using 12-20 milliampere (mA) analog signals, so we also use the Hart protocol, which is old but has a large installed base of devices and takes advantage of both the analog and digital worlds.
Philipp Arquint, head of sensor innovation, Hamilton Company, Bonaduz, Switzerland. Source: Hamilton Company.
Given the deterioration oxygen causes food and beverages during and after processing, manufacturers are increasing their efforts to minimize oxidation’s impact (see “Get the micro-oxygen out,” Food Engineering, September 2010). Trace amounts of dissolved oxygen can have serious consequences for product quality, but detecting and measuring molecules in the parts per billion range is a challenge, particularly when monitoring is performed in the processing area instead of a laboratory.
In discreet amounts, dissolved oxygen is even desirable, though monitoring and controlling the specific amount requires a high level of process control. Beer fermentation is a prime example. Though fermentation itself occurs without oxygen, a controlled amount of air is blown into the wort for optimal propagation of the yeast. Sensors monitor oxygen levels during air dosing and while the wort is held in a fermentation tank, but conventional sensors can be inexact, are susceptible to interference and require frequent maintenance and replacement of sensor caps and electrolyte components. CIP chemicals add to the maintenance challenge.
Fortunately, the state of the art in sensor technology is rapidly improving, thanks in large part to the demands of biotechnology. The integrity of sensor data is critical for the process control systems of pharmaceutical manufacturers and other biotech companies. Sensor suppliers have responded with a number of advances that simplify the technology, reduce costs and make devices more robust. Breweries are among the beneficiaries of those improvements. Among the leaders in advanced sensor technologies is the Hamilton Company, Reno, NV. Best known for its precision syringes and lab devices, Hamilton also operates a sister company in Switzerland where research and development in sensors, robotics and other disciplines are conducted.
Improvements in sensor technology are under the direction of Philipp Arquint in the Hamilton Company’s Bonaduz, Switzerland offices. After earning a master’s degree in software engineering from the University of Applied Sciences HTW Chur in Switzerland and a master’s in electrical engineering from the Swiss Federal Institute of Technology Zurich (Albert Einstein’s alma mater), Arquint was awarded a doctorate in sciences from University of Neuchâtel in Neuchâtel, Switzerland. While completing his doctorate, he worked as a computer engineer at Hamilton Company, where he developed an ELISA test for a pharmaceutical company. Since 2000, Arquint has served as head of sensor innovation at Hamilton Company, focusing on development of chemical sensors, calibration solutions and fittings.
The Bitburger brewery in Bitburg, Germany is an early adopter of Hamilton’s dissolved oxygen sensors with built-in electronics that eliminate the need for signal transmitters. The brewery field tested an optical oxygen sensor for five months to monitor finite amounts of air in wort pipes (foreground) during transfer to the fermenter (upper right). At left, a close-up of the sensor. Source: Hamilton Company.
Arquint: In the old days, around 2008, the only oxygen detectors commercially available were electrochemical sensors that were passive and needed to be connected to a transmitter. Our research efforts in recent years focused on eliminating the traditional transmitter by designing a sensor with a microprocessor built into it, allowing the sensor to communicate directly with the control system.
We refer to these intelligent sensors as the ARC family. The name is supposed to suggest a bridge into the future. The newest additions are the Oxygold B for measuring trace quantities of oxygen in beer, soda and other beverages with high CO2 levels and the Conducell PWSE, which provides low conductivity measurements in water and ultra-pure water and CIP validation that caustics have been evacuated from a line before production resumes.
FE: What were the key issues you faced?
Arquint: Miniaturization of the components was the biggest issue. Three years ago, our customers had a transmitter the size of a shoebox on the wall to amplify and relay the sensor’s signal. The challenge was to miniaturize the transmitter down to the point where it would fit into a shaft with a 12mm (0.47 inch) diameter. A partner with expertise in semiconductor design produces customized electronic chips for our application. They make the electronics, and we assemble the components in Bonaduz.
Another challenge was validating that the sensors would not pose a risk as an ignition source. Regulations surrounding explosion protection are very strict. The requirements are very demanding in terms of the design of electronics when they are placed in a potentially explosive atmosphere.
FE: Why was explosion protection a major hurdle?
Arquint: The requirements for intrinsically safe electronics, which is the basis for explosion protection, collide with the fundamental requirements for miniaturization. Explosion protection requires larger components and isolation distances between interconnections. The components’ maximum operating values in terms of temperature, current, voltage and power should never exceed 66 percent of the specified values. As a result, there had to be an intelligent trade-off between miniaturization and intrinsic safety.
To reconcile these conflicting demands, we collaborated with a company specializing in development of intrinsically safe electronics and with the Fraunhofer Institut für Integrierte Schaltungen in Erlangen, Germany, which has a huge expertise in the design of analog and digital integrated circuits. The challenge is bringing both of these worlds together.
FE: What is the underlying technology used for detection?
Arquint: Some use electrochemical technology, while others are optical. Until about five years ago, optical sensors for dissolved oxygen did not exist; the only option was electrochemical. Both technologies are in use in breweries, depending on customer preference.
Optical devices, which rely on a chemical layer that is excited by light to produce a signal, remove some of the issues associated with electrochemical, such as the maintenance required to prevent clogging and the fact that they incorporate ceramics that can crack and present a danger. But optical sensors take a long time to polarize, can’t measure oxygen under certain conditions and are not as sensitive. The nominal sensitivity is eight parts per billion (8 ppb), though detection can drift over time to 15 ppb, and we don’t recommend optical for breweries. With electrochemical units, the nominal sensitivity is 1 ppb.
FE: In what types of applications are the sensors being applied?
Arquint: Nuclear cooling plants use this technology, as do many process industries operating in an industrial environment. Our primary focus, though, is biotech, and we have many pharmaceutical clients who must strictly control the fermentation process for some of their ingredients. The food industry is quite close to pharmaceutical, in terms of the requirements for sterility, temperature ranges and the growing demand for digital designs. Miniaturization of the electronics into the sensor complicates the design, since the devices must be able to tolerate the hot, aggressive media in an autoclave. They also must tolerate caustics in the cleaning cycle, and CIP is particularly harsh on optical membranes.
FE: What calibration issues are associated with the sensors?
Arquint: We calibrate them in the factory, or the customer can do the same thing in his laboratory, put it on the shelf and then simply put it into service when an older sensor fails. No recalibration is needed because, unlike passive sensors that have no memory, all the settings are stored in the sensor itself.
FE: Where are you now focusing your development work?
Arquint: The future will be connecting the sensor directly to a PLC in a purely digital connection, without any analog signals running through the plant. With all the signal noise in an industrial setting, analog signals experience interference problems. On the other hand, analog lets you take someone without any advanced training, give him a small voltmeter and send him out into the plant, where he can understand what the signal is telling him.
Digital communication using Profibus or Foundation Fieldbus is widespread in factory automation, but in process control and especially process analytics, we estimate 60 to 80 percent of applications still are using 12-20 milliampere (mA) analog signals, so we also use the Hart protocol, which is old but has a large installed base of devices and takes advantage of both the analog and digital worlds.
Looking for a reprint of this article?
From high-res PDFs to custom plaques, order your copy today!
