The present invention relates to a method for adjusting and controlling an ozoniser installed in commercial open-type refrigerator benches (for supermarket and shop) for conserving, displaying and selling fresh food not packaged.

As is known, a commercial refrigerator bench is an open cabinet which delimits a refrigerated volume adapted to conserve, display and sell fresh products to consumers. The products are typically stored at a temperature of between -2 °C and +5 °C. A refrigerator bench may be of the so-called "service" type, where the goods displayed are provided to the customer by an operator, of the so-called "self" type, where the customer picks up the goods directly and there is no operator, or mixed, where the customer can take the goods, as well as being served by an operator.

A refrigerator bench may also be channelled, i.e. obtained by combining multiple individual refrigerator benches to form a channel of refrigerator benches that may or may not have the same features as sale of goods (e.g. meat, deli, meats and dairy products in combination) . A refrigerator bench normally has a length in the range of between 0.75 m and 3.750 m, with the possibility of having refrigerated corners. A channel of refrigerator benches may potentially be even 20 m long, being typically made up of several different types and lengths of refrigerator benches. A refrigerator bench can sometimes also be provided with partial transparent closures, or night curtains, which ensure a predetermined night energy saving. Examples of commercial refrigerator benches for conserving, displaying and selling fresh food are shown in figures 1 and 2.

The difference between a commercial open-type refrigerator bench and a home or professional refrigerator or a refrigerating room, both of the closed type, consists mainly in the fact that the open condition, which is necessary for the display and sale of products, generates continuous contact between the air contained in the refrigerated volume and the air in the area where the refrigerator bench operates. The continuous input of warmer and humid air in the refrigerated volume involves problems of particular adjustment and control other than those that occur in a closed refrigerated system, especially as a result of the thermo-hygrometric transformations of the air inside the refrigerator bench, which occur as a result of thermostatic pauses and defrost cycles.

A thermostatic pause is substantially defined as a stop of the refrigerating machine of the refrigerator bench. This stop is necessary to keep the temperature of the refrigerated volume within certain threshold values. When the temperature of the refrigerated volume exceeds a certain upper threshold value, the compressor of the refrigerating machine is powered and the refrigeration cycle starts, which returns the temperature below the limit defined by such an upper threshold value. In contrast, when the temperature of the refrigerated volume falls below a certain lower threshold value, the compressor stops again and the temperature rises. The frequency of the number of thermostatic pauses depends on the difference between the refrigerating capacity delivered by the compressor and the power required by the refrigerator bench. The greater the difference, the more frequent are the thermostatic pauses.

During a thermostatic pause, the relative humidity inside the refrigerator bench increases significantly, since the evaporator tube of the refrigerating machine heats up in absence of the passage of coolant and by exceeding the dew value of the air, it loses the capacity of condensing and dehumidifying the air itself. The trend of humidity in the air in an open refrigerator bench is shown in the graph in figure 3. From the graph we see that during the operation of the refrigerator bench, the air humidity reaches minimum values around 70%, before rising during the thermostatic pause up to values around 83%. It is further noted that, during the stop of the refrigerator bench for defrosting, the relative humidity rises almost reaching saturation values (98%) . The thermostatic pause further has a positive effect on the reduction of frost on the evaporator since, when the compressor stops, the frost accumulated on the evaporator melts, thereby increasing the level of relative humidity.

In a refrigerator bench, defrosting is the operation by which any frost accumulated in refrigerator bench itself is thawed. In a commercial open-type refrigerator bench, defrosts always take place automatically and are programmed as a function of the operating time. Frost always forms on an evaporator when its (evaporation) temperature is lower than 0 °C. When humid air comes into contact with the cold surface of the evaporator, the water content in the air freezes and is deposited as frost. With time, the frost builds up on the evaporator, thereby reducing the heat exchange surface thereof, so at some point, it becomes necessary to stop the refrigeration for a sufficiently long period (30-40 minutes) so that all the accumulated frost thaws and the water accumulated on the bottom of the refrigerator bench is drained by gravity through a drainpipe .

As mentioned above, the values of relative humidity in an open refrigerator bench are very variable and fluctuate over time, so the presence of a humidity sensor would not find proper application in context of adjustment, especially for the condensation that would form on the sensor and for its poor dynamic response. The possibility of simulating the pattern of humidity through an algorithm, during pauses and defrost, then assumes a special importance.

In detail, the operation of a commercial open-type refrigerator bench 10 (figure 4) is as follows. The cold produced by evaporator 12 of refrigerating machine (which can either be built into the refrigerator bench or installed remotely in order to operate with a plurality of separate refrigerator benches) is transmitted through an air current 14 generated by one or more fans 16. The air is introduced into the tank, which constitutes the refrigerated volume containing the goods to keep to a predetermined temperature, through a first grid slot 18 placed on a side of the tank, called delivery side, and is returned into evaporator 12 on the other side of the tank, called return side, through a second grid slot 20. Evaporator 12 is placed underneath the tank and is traversed by the air to be cooled.

The tank is a refrigerated volume not physically delimited on at least one side, so as to keep the refrigerator bench open. Due to the formation of a curtain of air that is produced by the refrigerated air circulation, there is a virtual delimitation of the tank. The curtain is formed both due to the presence of external still air, normally warmer and lighter, and because of the air trapped inside the tank, generally colder and therefore more dense. The air speed, going towards the outside of the tank, decreases progressively up to stopping. In the tank area where the cold and dry air of the refrigerated compartment mixes with the humid and hot air from the environment, there also occurs a high humidity input in the refrigerator bench, resulting in the formation of frost on the evaporator (cold component of the system) . The refrigerator bench, for mere energy saving reasons, may therefore be provided with closing sliding or non- sliding partitions, in order to limit the mixing of the air during the products handling, both by the customers and by the sales operators.

The diagram in figure 5 shows, on the enthalpy diagram of humid air, the transformation states of air itself. Ambient air, in conditions of temperature t _a and relative humidity cp _a , mixes with the cold air inside the refrigerator bench, reaching temperature t and relative humidity φ2 (condition of inlet in the air return slot 20) . From here, crossing the conduit and going through the fan, the air warms up to the temperature t2 and relative humidity φ2 conditions (condition of inlet into the evaporator) . The air then passes through the evaporator, cooling down and reaching the saturation state (0) inside such an evaporator. At the outlet from the evaporator, the air is at temperature ti and relative humidity ci. The air finally passes through the delivery conduit, warming up to exit from the delivery mouth at the conditions of temperature ti and relative humidity ci .

From the humid air graph in figure 5 we can get the total heat Qo removed from the refrigerator bench, which corresponds to the result obtained by multiplying the flow rate of cooled air Mi by the enthalpy difference in the air at points 2' and 1', respectively. Ozone is supplied by a distribution system, which could also consist of a laterally perforated tube placed on the suction side. The sensor that measures the ozone concentration is placed on the suction side above the ozone distribution system for safety reasons. The ozone generator is placed in the vicinity to the refrigerator bench and such an ozone generator is connected to both the sensor that measures the ozone concentration and the ozone distribution system placed in the tank.

The air that exits from the delivery side is refrigerated and, depending on the operation of the refrigerator bench, more or less dry and rich in ozone. Coming in contact with the goods and the surfaces of the refrigerator bench, ozone is gradually consumed. The air mixes with the ambient air and is then sucked from the return side. Here, the ozone concentration is measured and, accordingly, the production of ozone in the ozoniser is regulated, unless it is the step when the ozone concentration is estimated. The air then passes into the ozone distribution system, where the desired concentration is restored. Subsequently, the air sucked by the fan is passed through the evaporator, which cools it down. The air then enters into the tank, is mixed with the outside air and the cycle resumes.

Ozone is an allotrope of oxygen characterised by a high oxidizing power which gives it a marked microbicidal activity. By its nature, ozone is not stable over the long term and therefore it cannot be produced and sold in cylinders, as is the case for other industrial gases. This drawback leads to the fact that ozone must necessarily be prepared at the time of use by means of special devices known as ozonisers.

The ozone generation principle provides, according to different modes, the dissociation of the molecular oxygen and the intermediate formation of oxygen radicals, which react with oxygen. The main ozone formation modes are:

- UV radiation (photochemical process);

- electrolytically;

- by an electric discharge (crown effect) ;

- using the "cold plasma" technology.

The first ozone industrial applications concerned the chemical disinfection of water and wastewater treatment processes. Ozone is also used, in gaseous or liquid phase, in the post-harvest treatment of fruit or vegetables or as a preservative agent in preservation and storage environments. Several studies have proven that ozone is effective in significantly prolonging the "shelf life" and reducing the undesired microbial load of fruit and vegetables, such as berries, grapes, peppers, tomatoes and other products, including fish products, meat and poultry. The ozone microbicidal action is carried out by means of the ability of such a compound to permanently compromise vital cellular components of the microorganisms (DNA, enzymes and proteins) by oxidative damage, without releasing any type of chemical residue potentially dangerous to the consumer's health.

The simultaneous use of ozone and refrigeration is one of the most interesting joint applications for the purposes of proper storage and control of the hygiene and health quality of food. The ozone microbicidal action is, in fact, supported by the bacteriostatic and preservative power of cold.

On the basis of the above advantages of using ozone in the food sector, the Applicant has conducted a series of tests on prototypes of refrigerator benches for conserving and displaying non-packaged foods and has verified that there are many benefits resulting from such an application.

The advantages are of particular importance when non-packaged fresh meat is conserved in the refrigerator bench. The tests conducted have in fact shown not only the ozone ability to sanitize all the surfaces with which the meat is in contact, but also the fact that ozone has an antibacterial effect which ensures a 40% longer preservation of food compared to standard storage conditions.

One of the major problems of refrigerator benches for the preservation of fresh meat is to be able to keep the meat well preserved for relatively long periods (more than 24 hours), therefore refrigerator benches are generally always emptied at the end of the day because otherwise the meat would oxidize and would be unsellable. With the ozoniser, on the other hand, it would be possible to greatly increase the storage period, thus avoiding the emptying of the refrigerator bench with all the burdens in terms of times and costs that this would entail.

However, the Applicant has also verified that the standard apparatuses known today for ozone generation, when installed in a refrigerator bench, lose their control capability (that is to say, the capability to adjust the ozone concentration in the bench) , resulting in poor reliability of the results and possible overrun of the TLV (Threshold Limit Value) , which for ozone is 0.1 ppm. In fact, operators and customers are in continuous contact with the refrigerator bench, which is not closed or if it is provided with closures, they are opened to allow picking the products. It is also established that an ozone concentration higher than 0.07 ppm oxidizes the product, so it would undermine the sale of the product itself rapidly.

The problem that is presented in the most dangerous way, in fact, is that the measurement and the relative dispensing of ozone inside the refrigerated volume are not reliable in terms of amount of supplied ozone, which could reach average values well above the limits imposed by law. All of this despite the fact that the apparatus is programmed to stay well below said values. By investigating further, the Applicant has found that the problem lies within the constraints imposed by the construction technology of the sensor for measuring the ozone concentration which controls the ozone generator in feedback. This is in particular because at temperatures below 5 °C, the course of the signal output from the sensor is no longer linear, as well as for the fact that humidity in certain operating states of the bench affects the sensor's ability to make reliable measurements.

The sensor for measuring the ozone concentration, in addition to being as said above very sensitive to variations in temperature and humidity that occur continuously inside a refrigerator bench, whereby the measuring accuracy is very variable, exhibits other drawbacks. In fact, the sensor for measuring the ozone concentration is provided with a protective membrane of Teflon or similar material, particularly sensitive to both humidity and to cleaning agents, such as in particular alcohol and other petroleum products normally used for cleaning the refrigerator bench. These agents, even in moderate concentrations, cancel the sensor reading capacity and often damage it irreversibly. The traditional control system used by the standard apparatus, of the PID type, is therefore not able to respond appropriately to the operating variables of the bench which, by its characteristic (it is an open bench or at most with some closures that however do not solve the problem) never operates under steady state conditions.

Document JP 2001 061460 A describes a cooled and humidified warehouse with ozoniser to sterilize the environment. The warehouse is provided with an insulated double wall, internal and external, which therefore delimits a structure completely closed on all its six sides. A hollow space is provided between the two walls which is refrigerated by a traditional external refrigerator group. An inner wall of the warehouse is conductive and allows the input of "cold" . However, means for moving the air inside the warehouse are not provided. In JP 2001 061460 A, three ramps of pre-programmed value (and not calculated dynamically) are used to prevent overshooting upon starting the ozone generator. Since the ozonation times are pre-programmed, they are then fixed while the value of the ozone concentration is derived based on the relative humidity measured value. Finally, JP 2001 061460 A makes no reference to corrective predictive logics applicable during the ozonation process.

Starting from this known technique, the object of the present invention is to provide a method for controlling an ozoniser, installed in refrigerator benches for conserving and displaying fresh food, which is able to solve the drawbacks described above and, in particular, the risk of blinding the sensor in the presence of high humidity, allowing the operator to carry out the cleaning step of the refrigerator bench without damaging the ozoniser sensor.

The features and the advantages of a method according to the present invention will become apparent from the following exemplary and non-limiting description, made with reference to the accompanying schematic drawings, in which:

figures 1 and 2 show embodiment examples of commercial refrigerator benches for conserving, displaying and selling fresh food;

figure 3 is a graph showing the air humidity trend, as a function of thermostatic pauses and defrosting cycles, in an open refrigerator bench;

figure 4 shows the operation of an open refrigerator bench and the relative air circulation scheme;

figure 5 is a graph of the humid air transformation points ;

figure 6 is a diagram of a refrigerator bench which shows in particular the components of the ozoniser; and figure 7 shows the division of the operating states of a refrigerator bench where high and low humidity phases are present within the single states, where the ozoniser is controlled by the sensor or by means of algorithms that bypass the sensor itself.

In general, these objects have been achieved by the

Applicant through an innovative ozoniser control method where the normal PID controller (traditional analogue model-based adjustment) is replaced by a suitable mathematical algorithm based on modelling a discrete system divided between observable states, in which the information obtained from the sensor is reliable, and states where the sensor reading is no longer reliable. Therefore, particular attention had to be put in evaluating the feedback error signal changes, taking into account the gradient and its sign. It was necessary to reconsider the output values, correlating them to the previous input and feedback values, implementing a corrective type modelling.

Furthermore, in order to ensure the best possible adjustment level, it was necessary to add a predictive model which, on the basis of the results obtained, adjusted both the error variable and the estimate of the ozone concentration output from the respective generator device to obviate the reading faults mentioned above.

In particular, in case of high relative humidity, the control unit bypasses the sensor for a fixed or variable time frame and commands the ozone generator according to other logics. In particular, the exclusion of the sensor from the control function of the ozone generator occurs when an algorithm for estimating the humidity present in the refrigerator bench, an algorithm that acts as a humidity sensor, deems that a humidity threshold value is exceeded in such a refrigerator bench beyond which the command via the ozone concentration sensor could lead to errors. Note that an actual humidity sensor cannot be installed in the refrigerator bench, both due to the problems of poor measurement speed and high inertia that would arise, and because it would be subject to operate at the lower limit, if not beyond, of its permissible temperature range.

In addition, the risk of damaging the ozone concentration sensor by contact with aggressive substances has been solved by removing power to the sensor during the cleaning steps (the operator, at the beginning of his/her shift, at the end of his/her shift or on request, cleans the bench with cleaning products) and then, before restarting it, over a period of time, for example for about 20 minutes, by immersing it in an ozone stream. In doing so, the high oxidizing power of ozone cancels the aggressiveness of cleaning agents, ensuring the functionality of the sensor.

As mentioned above, the inventive aspect underlying the present patent application provides for the division of the operating steps of a refrigerator bench into finite states, identifiable depending on temperature, algorithm estimate of the humidity in the bench and purposes, to then define within control parameters of the ozoniser these states, which selectively activates the control on basis of the relative ozone concentration sensor or excludes it for a fixed or variable time frame. The estimate of relative humidity is obtained as a function of thermostatic pauses and defrosting cycles of the refrigerator bench.

The refrigerator bench system has been assimilated to a finite state machine, in which such states have been defined as:

a) refrigeration,

b) thermostatic pause,

c) defrost,

d) cleaning.

Two subsystems can be defined within each of these states, respectively, a phase which can be observed and reached through a correct use of the sensor (measurement conditions ensured by the temperature and relative humidity range) , and a subsystem which cannot be reached within which the sensor measurements could control the ozone generation incorrectly (when out of the measurement range due to high relative humidity) . The non-reachability of unreachable subsystems is mainly due to the high relative humidity that is sometimes present in refrigerator benches.

Since such high humidity situations are however short events, the unstable, non-reachable subsystem can be returned to a level of stability and reachability by appropriate corrections. In other words in such situations, or when the humidity estimate algorithm deems that a threshold humidity value has been exceeded in the bench, the data detected by the ozone concentration sensor are bypassed by acting on the ozoniser through particular alternative algorithms. In these non-observable sub-states it is therefore necessary to provide for special ozone concentration controls.

To the ozone concentration control, there are two possible types of management, equally effective. The first one involves the modelling of the refrigerator bench at time intervals. In this case, the control has a consolidated dynamics whereby the transition takes place after a fixed time. Alternatively, the control of the ozone concentration is implemented as a function of the ozone concentration gradient when the ozone concentration curve exhibits a gradient that exceeds a certain threshold value. As shown in the diagram in figure 7, in fact, during the use of the refrigerator bench, states alternate for selectively relying on the sensor or bypassing it depending on the humidity present in the refrigerator bench.

In the first step (Al), the refrigeration is started and there is air with high relative humidity. Under these conditions, the sensor must be bypassed and the ozoniser must be activated with alternative commands. This state is maintained for about one minute. After this step, there is a central refrigeration step where air is dry and allows the use of the sensor without encountering control errors. Upon reaching the set temperature, the thermostatic pause (Bl) starts and the same ozoniser control can be used with actuated sensor. As the thermostatic step (B2) continues, the humidity increases and the system returns to exclude the ozoniser control sensor.

At the end of the cycle, the defrost step (CI) starts with air at 5 °C and relatively humid. This step lasts about 5 minutes. The next step is defrosting (C2) with very humid air. During this step (which usually takes at least 40 minutes), the sensor progressively loses its reliability and can be either switched off or powered. In the final part of the defrost step (C3), with air saturated with humidity and high risk of condensation, the sensor is completely blind and the operation therefore takes place by bypassing the sensor itself .

During the final cleaning step (P) , during which the operator cleans the surfaces of the bench with detergents, the sensor is either switched off manually by the operator, or the disabling is programmed via an appropriate scheduler, whereby the transition to this state is done by external input known to the algorithm. When the refrigerator bench is cleaned by the operator, in order to allow the use of detergents, the sensor is in fact switched off for a set period of time and ozone production is contextually interrupted. At the end of this step, with the sensor still switched off, the ozoniser is restarted for another set period of time, different from the previous one (for example for 7-10 minutes or for a "guarantee" time sufficient for the ozone to sanitize the tank of the refrigerator bench from the presence of pollutants) . Thereafter, the sensor is restarted since the ozone produced in the previous 7 minutes has degraded the harmful molecules present in the refrigerator bench and on the sensor, released by the detergents used by the operator.

The sensor is particularly sensitive to external polluting agents such as alcohol, ammonia, various hydrocarbons and, in general, to standard industrial and non-industrial detergents. The sensor exposure to these contaminants requires a step in which said sensor provides completely unreliable measures, and often there is a high risk of breakage of the sensor itself. Therefore, the refrigerator bench cleaning step is very critical for the reasons above. During the first set period of time, i.e. when the sensor is deactivated, ozone degrades the components that affect the sensor itself both as measurement reliability and as durability. At the end of this first set period of time, the sensor is restarted (powered) and the normal adjustment begins. This system ensures that the sensor becomes operational in the shortest possible time, without issues regarding "pollutants" that affect both the operating life and the measurement reliability of the sensor itself.

For further caution, the sensor housing may be: - heated with a thermostat resistor (such as NTC or PTC) in order to prevent condensate;

- housed with the sensitive part downwards to avoid and/or facilitate the disposal of any condensate;

- in the case of channelled benches, powered for additional 5 minutes after the end of the defrost step;

- temperature-monitored in order to compensate exactly the drift in temperature of the sensor itself.

It has thus been seen that the method according to the present invention achieves the objects mentioned above by providing, on the basis of appropriate humidity algorithm estimates, steps in which the ozoniser is controlled on the basis of the data measured by the ozone concentration sensor and steps in which the control is accomplished by bypassing the sensor and acting according to alternative control logics .

The method of the present invention thus conceived can be subjected to numerous modifications and variants, all falling within the same inventive concept; moreover, all details may be replaced with technically equivalent elements. In the practice, the materials used as well as the dimensions thereof may be any, according to the technical requirements.