Continuous Monitoring System for Wine Fermentation

Transcription

1 Continuous Monitoring System for Wine Fermentation Margarida Seixas a, Raul Morais b, A. Valente c, C. Couto d a Universidade de Trás-os-Montes e Alto Douro, Vila Real, Portugal, mseixas@utad.pt b CETAV/UTAD, Vila Real, Portugal, rmorais@utad.pt c CETAV/UTAD, Vila Real, Portugal, avalente@utad.pt d Universidade do Minho, Guimarães, Portugal ccouto@um.pt Abstract An automatic system to continuously monitor and control wine fermentations will enable winemakers to produce better quality wine and maintain competitiveness of Portugal in international wine markets. This paper presents an alternative approach of continuous monitoring system, in a fermenting tank, to create a supporting system to the most efficient and reliable decision. So far, the collecting data of the fermentation variables and equipment tests are done. Now this work still in a phase of the data treatment and as soon as sufficient information is obtained to characterize the process the mathematical models will be carried out in order to equate optimization and control strategies. Key words: Wine, Fermentation process, Fully automatic monitoring. 1 Introduction Wine follows a long and complex path from grape to wine glass. In addition, increased competition, larger production volumes and a raise in discerning wine consumers create greater challenges for the wine maker. Biotechnological industry, like the wine production, has evolved fast in the most recent years, greatly because of the efforts that science and engineering has developed in order to provide monitoring off the processes that allows avoidance and correction of errors and fails. This development provides energy and coasts saving. Wine-cellar techniques have advanced significantly in the last decades. Whereas experience and intuition were sufficient for the wine maker in the past, today these assets are supported by advanced technology and a scientific approach to the various processes involved in wine production. A good monitoring, followed by a good control strategy, plays an important role in fermentation industry. The majority of the established control techniques are based on mathematical models of the controlled process. The state of the process variables must be sampled on-line so that the control strategy works. In this context, the main goal is to develop intelligent identification techniques to measure, on-line, the fermentation-related variables during red wine making and overcome all the deficiencies of the typical fermentation control systems, based in the operator decisions and measurements taken off-line. This integrated monitoring system will provide the winemaker with predictive indices that will serve much like warning signs on dangerous roads. Thus, they will be alerted to potentially high-risk situations so that early, effective and selective must treatments may be made. We are working to pinpoint key stages in the fermentation of grape must that could decide if a good wine becomes a great wine. The fully computer controlled system will be designed to give up-to-the-minute analytical information about wine as it is being made. When fully operational, with criteria set before starting, the system will be able to automatically take samples from fermentation tanks at hourly intervals 1392

2 or on demand, giving winemakers access to information on factors affecting fermentation such as: ph, redox potential, density, and temperature readings. This is a huge advantage over current methods, where one operator might take a day to manually process up to a few samples in the laboratory. This document is a template for Microsoft Word versions 6.0 or later. If you would prefer to use LaTeX, download LaTeX style and sample files from the same Web page. 2 Biochemical Process Description The first stage of winemaking involves a complex biochemical phenomenon that transform the sugars, glucose and fructose into ethyl alcohol, or ethanol, carbon dioxide and numerous products which are known as secondary products because they are present only in small quantities. Most red wines are produced by two different kinds of fermentation. In the primary or alcoholic fermentation, sugar such as glucose and fructose, contained in grapes is transformed in ethylic alcohol, or ethanol with carbon dioxide release. This complex process, assured by the action of living yeasts cells, may be translated by the following equation (Cooke, 1988): C 2 + CO 6H12O6 CH 3CH 2OH 2 ( 1) The secondary or malolactic fermentation is a completely different phenomenon from the alcoholic fermentation, being always followed by the activity of a bacterium, such as Leuconostoc, Pediococcus and Lactobacillus genera. This bacterium is responsible for the transformation of malic acid into lactic acid, with carbon dioxide release (Eisenman, 1998). 2 H + 2C4H 4O6 HC3H 5O3 2CO2 ( 2) 3 Physical-chemical Factors That Affect Fermentation The fermentation process involves many chemical reactions and its kinetic depends of several physical factors. Two of the most important parameters, that affects this kinetic, are ph and temperature. The yeasts grow and are reproduced quickly due to rich nutrients that there are in the must, obtained from the pressed grape, and they produce the consumption of fructose and glucose. The speed of the fermentation will be a function of the temperature, humidity, concentration of nutrients, among others. The maintenance of these optimal conditions assures a good evolution of this process and assures the survival of microorganisms. 3.1 ph The ph is a measure of the number of hydrogen ions present in a solution. A variety of chemical reactions takes place in wine, and many of these reactions are affected by the total number of hydrogen ions present. If the wine has a low ph the sugar fermentation progresses more evenly, and malolactic fermentation is easier to control. The situation is much different when wine ph values are high. Bacteria multiply rapidly in high ph wines, and unwanted bacterial fermentations become more troublesome. High ph wines are less biologically stable, and they have poorer chemical stability. Red and white wines have poorer color when the ph is high. Wines with high ph values always require more attention and greater care than wines with low ph values. Therefore, the ph control must be done carefully and the necessary corrections must be made on time by the addition of an acid or alkaline solution. 1393

3 3.2 Temperature Temperature is another essential parameter in the success of fermentation. Microorganisms growing at a temperature lower than the optimal reveal a delayed growth and a reduced cellular production. On the other hand, if temperature produced during alcoholic fermentation reaches C the must becomes vulnerable to bacteria which change the sugars into mannitol, producing an undrinkable liquid. In order to obtain a good efficiency, fermentations need a superior temperature in the beginning, to stimulate the actions of yeast, and after that, the process could continue with a lower temperature. 4 System Elements, Design and Available Functions The success of fermentation depends upon the existence of defined environmental conditions for biomass and product formation. Thus temperature, ph, degree of agitation, oxygen concentration in the medium and other factors may have to kept constant during the process. The provision of such setting requires careful monitoring of the fermentation so that any variation from the specified optimum might be corrected by a control system. As well as aiding the preservation of constant conditions, the monitoring of a process may offer information of the progress of the fermentation. Such information may indicate the optimum time to harvest, or, that the fermentation is progressing abnormally. Thus, monitoring tools should be capable of being associated to a suitable control system as well as producing information indicating fermentation progress. The prototype in study intends to guarantee all the conditions for a most advantageous fermentation and offers all the capabilities of monitoring and actuation in every one of the factors that affect, most, the fermentation kinetic. Criteria, which are monitored in this system, are listed in the Table 1. Table 1 Process Sensors and their possible control functions Category Sensor Control Function Physical Temperature Heat/cool Pressure Density Chemical ph Acid or alkali addition Redox Additives to change redox 5 Sensors Technical Overview 5.1 ph Sensor Measuring ph involves comparing the potential of solutions with unknown [H+] to a known reference potential. The ph meter detects the change in potential and determines [H+] of the unknown by the Nernst equation: E = E 2,3RT unknow H + log nf int ernal [ + ] [ H + ] 0 ( 3) Where E, is the total potential difference (measured in mv), E 0, the reference potential, R, the gas constant, T, the temperature in Kelvin, n, the number of electrons, F, the Faraday's constant and [H+], the hydrogen ion concentration. This prototype uses Sentrons rugged NON-GLASS ph probes. The Sentrons high-tech Ion Sensitive Field Effect Transistor (ISFET) ph probes replace fragile glass electrodes with sturdy silicon sensors. The output signal has an accuracy and resolution, respectively, of ±0,02pH and 0,01pH, in all ph range (0 to 14 ph). 1394

4 5.2 Redox Potential Sensor Redox Potential, also known as ORP, stands for Oxidation-Reduction Potential. The relative tendency of different substances to lose electrons (relative reduction potentials) varies depending on the number of electrons in the outer shell and on the size of the atom or ion. ORP values are affected by all oxidizing and reducing agents, not just acids and bases. The life expectancy of bacteria in wine is related to ORP. The potential, E, that the noble metal ORP Electrode have relatively to the standard hydrogen half reaction in any REDOX reaction is: [ Oxid] [ Re d] RT E = E 0 log nf ( 4) Where E, is the total potential difference (measured in mv), E 0, the reference potential, R, the gas constant, T, the temperature in Kelvin, n, the number of electrons, F, the Faraday's constant, [Red], is the concentration of the reduced form of the substance and [Oxid], is the concentration of the oxidized form of the substance This prototype uses a Hanna ORP controller and its voltage output signal has an accuracy and resolution, respectively, of ±2mV and 1mV, in a ORP range of to +2000mV. 5.3 Pressure Sensor. Pressure, by definition is a derived parameter. It is the combination of a mass measurement imposed upon an area. Electronic pressure transducers incorporate a strain gauge diaphragm or piezoelectric cell and signal conditioning electronics to convert the pressure into an electrical signal, such as 4-20mA transmitter r mv output. The prototype uses two Vibrocontrol piezoresistives sensors with a current output signal (4-20mA). These sensors have a measurement range from 0 to 1bar pressure with an accuracy of ±0,1%. 5.4 Temperature Sensor. The instrumentation model build for this work use a temperature sensor based on platinum resistance thermometers - PRT - currently named PT100. This kind off sensors offers excellent accuracy over a wide temperature range - from -200ºC to 850ºC. The principle of operation is to measure the resistance of a platinum element. The PT100 type has a resistance of 100 ohms at 0ºC and 138,4 ohms at 100ºC. The relationship between temperature and resistance is approximately linear over a small temperature range: for example, if you assume that it is linear over the 0 to 100ºC range, the error at 50ºC is 0,4ºC. 6 Fermentation Tank Constitution The fermentation tank prototype intends to assure optimal fermentation conditions in an aseptic environment (Pumphrey, 1996). Therefore, this was equipped with airing, agitation and monitoring systems of the main process variables, namely ph, temperature and redox. This design - smooth internal surfaces using, whenever possible, welds - allows the cleaning and maintenance operations. In order to decrease the operation and maintenance costs of the fermentative process, all the electro-mechanical system was conceived in a way to reduce, to the minimum, the energy consumption as well as the losses by evaporation. The prototype tank has also an airing and agitation system in their upper side. The agitation is achieved by hand vertical movements that provide must homogenization and contact between liquid and grapes in order to give color and flavor to the wine. Agitation is an operation that creates or speeds the contact between two or more phases. The proper agitation is of main importance to fermentation, since it insures 1395

5 the air dispersion in the solution, the temperature homogenization, ph and nutrients concentration in the must, color and aroma of wine by the grape s skin contact with the juice. To allow off-line measurements and solution samples a collect system was also introduced in the bottom of the tank. 7 System Operation Overview The stainless steel fermentation tank, pictured on Figure1, is surrounded by several sensors that facilitate fully automatic monitoring and measurement of ph, redox, pressure and temperature inside and outside of the fermentation tank. All the sensors of the system require periodic maintenance to clean and calibrate them. This action must be made every time at the beginning of the fermentation. To facilitate this operation all the sensors are attached to the tank with a specific connection that provides isolation and easy installation. The fermentation-related variables like ph, temperature and ORP are directly measure by the particular sensors but to estimate the amount of alcohol in wine, you make an automatic hydrometer using two pressure sensors, one in the top and the other in the base of the fermentation tank. The hydrometer is a weighted glass tube that floats in a liquid and measures the specific gravity, or density of the solution. As the sugar is converted to alcohol, the specific gravity of the solution reduces. Fortified wines like Porto require a special fermentation process or the addition of Brandy to help reach higher alcohol content. The pressure measured at the bottom of a liquid filled tank is proportional to both the height (head) and density of the liquid. For liquids with densities different from water, liquid density can be determined from pressure, gravitational force and liquid column height with the formula: 68947,6P = gh ρ ( 5) Where P, is the pressure in pounds per square inch (PSI), ρ, the liquid density in grams per cubic centimeters (gcm -3 ), g, the gravitational force and h, the liquid column height in centimeters (cm). The acquisition is performed with an acquisition board - containing a 12 bits ADC that communicate to a central computer to allow the user to know about the continuous progress of fermentation in the tank. The samples are collected each minute, continuously. The PC function is to coordinate the development of the entire process, as well as allow its monitoring and modification. According to the acquired parameters and based on the control algorithm, decisions could be made in order to improve the fermentation process. A reliable, automated fermentation monitoring system allows the winemaker to detect fermentation problems while they may still be remedied. Just like other biotechnological processes, the wine fermentation needs an attentive analysis. In this circumstance, the system has an additional benefit to winemaker: data are all save in specific files for posterior studies, if is necessary. With this prototype, the winemaker need only attaching the sensors to the tank then the computer may be adjusted to sample at a time interval of his choosing. One personal computer performs the measurements that will be displayed and saved in the computer. With such a system in place, winemakers would not only be advising of problem fermentations, they would anticipate when the individual fermentations would be complete. 1396

6 Fig.1 - Fermentation tank elements. 8 Results So far, the collecting data of the fermentation variables and equipment tests are done. In the graphics of Figure2 we can observe the evolution of four fermentation related variables computed directly (ph, temperature and ORP) or indirectly (alcohol content) from measures made with the sensors in the tank prototype. The tests and results reveal a suitable performance of the system and a proper overview of the variables monitored in the fermentation process. Redox Potencial (mv) 400,00 300,00 ORP 200,00 100,00 0, , ,00-300,00 September

7 Temperature (ºC) 30,00 25,00 20,00 15,00 InsideTemperature OutsideTemperature 10, September 2004 Alchool Content Alchool Content September ,20 ph 4,00 3,80 3,60 3,40 3, September 2004 ph Fig.2 Fermentation related-variables evolution during this biotechnological process. 9 Conclusions and Future Work So far, the collecting data of the fermentation variables and equipment tests are done. Now this work still in a phase of the data treatment and as soon as sufficient information is obtained to characterize the handling of the data will be carried out and mathematical models will be tested in way to equate strategies that will improve the performance of the entire system. In conclusion, we can say that computerized approach makes it possible to monitor the critical elements of the fermentation process, and use them in an automatic control system. By giving the winemaker a more accurate assessment of the fermentation, earlier in the winemaking process than is currently possible, the build system will save the winery money and improve the quality of the finished wine. In a posterior phase, will be developed a turbidity sensor to be integrated in the fermentation tank and a new monitoring technology based in the sensorial fusion and micro sensors. 1398

8 We purpose, also, to make a consistent mathematical description of fermentation-related variables, such as temperature, ph, redox and density, during all this biotechnological process and identify the principal stages of this process. We intend, also, build a simple user interface and allow the monitoring of the system can be carried out from a central or remote location and a system responsible for fault detections that helps, once again, the operator mission. This automatic system of monitoring intends, in the near future, to maximize product yield and production capacity; and decrease product degradation, production costs, downstream losses and batch-tobatch variations. 10 References Boulton, Roger (1977) The prediction of fermentation behavior by a kinetic model, Am. J. Enol. Vitic., Vol 31, Nº1, Cooke, George M. and Lapsey, James T. (1988) Making Table Wine at Home, Division of Agriculture and Natural Resources, University of California, Davis. Eisenman, Lum (1998) The Home Winemakers Manual. Pumphrey, Brian and Julien, Christian (1996) An Introduction to Fermentation Fermentation Basics, New Brunswick Scientific. 1399