WO2001067014A1 - Method for detecting faults in a cooling system - Google Patents

Abstract

The invention relates to a method for detecting faults in a cooling system (1) comprising several cooling points (2, 3, 4). According to the invention, the cooling point (2, 3, 4) subjected to the highest load is continuously determined according to a predetermined schema. In doing this, it is desirable to be able to detect possible faults as early as possible. To this end, a load pattern of the cooling points (2, 3, 4) is established and a warning is generated when at least one predetermined parameter deviates from said load pattern.

Description

Process for detecting faults in a cooling system

The invention relates to a method for detecting faults in a cooling system with several cooling points, in which the most heavily loaded cooling point is continuously determined according to a predetermined scheme.

The invention is described below using the example of a cooling system in a supermarket, in which the cooling points are formed by individual sales counters. In these sales counters, chilled or frozen goods are kept ready so that a customer can look at them and remove them from the cupboards. However, the invention is not limited to supermarkets or the like, but can basically be used wherever a cooling system has several cooling points that can be operated with different loads. The term "cooling system" is intended to include all systems that serve to generate refrigeration, including those that would otherwise be called freezing systems.

A method of the type mentioned at the outset is known from US Pat. No. 4,084,388. The most heavily loaded cooling point is determined here in order to control the compressor arrangement. On the one hand, the compressor arrangement is intended to provide so much cooling capacity that the most heavily loaded cooling point can maintain the desired temperature. On the other hand, it should also no longer provide performance. In the known case, this is done by regulating the suction pressure of the compressor arrangement. This enables considerable energy savings. With cooling systems of this type, errors always occur. These errors are critical because the goods to be cooled can spoil if they do not remain at the set low temperature. A common way to monitor errors is to monitor the temperature. If the temperature at a cooling point exceeds a predetermined value for an extended period of time, an error is displayed. This is referred to below as a "temperature warning". Such a warning is often only given long after the actual error has already occurred. The error must then be remedied relatively quickly so that the chilled goods do not spoil or suffer a loss in quality. The urgency of troubleshooting is one of the reasons why such repairs are costly. This is especially true if they have to be carried out at night or at the weekend.

The invention has for its object to indicate a possible error as early as possible.

This object is achieved in a method of the type mentioned at the outset by creating a load pattern for the cooling points and generating a warning if the load pattern deviates from at least one predetermined parameter.

Here, the "continuous" determination of the most heavily loaded cooling point does not necessarily mean that this determination is carried out continuously over time. Rather, it is sufficient to carry out the determination in smaller time intervals, the time intervals should be adapted to the thermal time constant of the cooling system.

Many of the errors that occur in cooling systems only become temperature warnings after a relatively long time, i.e. than elevated temperatures. For example, it can take several days before the tires that begin to show up in the channels of a refrigerated display case, in which air flows from and to the evaporator of the display case, as a warning. If the temperature warning is triggered at night or at the weekend, a cooling technician is called in in many cases. An earlier warning would make it possible for the operator to carry out the necessary maintenance, for example defrosting, himself within normal working hours.

If you now create a load pattern of the cooling points, then it can be assumed in the undisturbed case that most or each of the cooling points is determined once as the most heavily loaded cooling point. If you determine the load pattern for the undisturbed, i.e. error-free case for a longer period of time, certain properties will be recognizable. For example, the load ratios of the individual cooling points are in a certain, almost constant relationship to one another or at certain times certain cooling points are particularly heavily loaded or the loads on the cooling points always change evenly over time. If a deviation from such a load pattern now occurs, the present invention assumes that this is the announcement of an error, ie an error at an earlier stage, which is otherwise not noticeable. For this reason, a warning is then generated. Because the warning is relative is generated early, the operator of the cooling system can go to troubleshooting during normal working hours or simply carry out the necessary maintenance work.

In many cases, a fault is not caused by elements of the cooling system itself, but by human error. For example, the door is left open on a refrigerated display case or in a cold room. Another example that cannot be detected directly with the aid of a temperature warning system is the incorrect stacking of goods, in which goods are stacked too high in a refrigerated display case, so that the cold layer in the display case cannot be maintained. The showcase works inefficiently because surrounding warm air flows into the refrigerated showcase and has to be cooled. The refrigerated display case must therefore provide more cooling capacity in order to maintain the desired temperature. This means not only uneconomical operation of the individual refrigerated display case, but also inefficient operation of the associated compressor system, because here the suction pressure is regulated, for example, by the most heavily loaded refrigeration point. In the example case, however, the load would be caused by the error, which could still be removed relatively easily at this stage. If one recognizes such an incorrect load at an early stage, there is no danger that an error caused by overloading occurs outside of normal working hours, which entails an expensive repair.

The load is preferably continuously determined locally at each cooling point. In this case, you can determine the most heavily used cooling point at any time, because the load data is available locally at every cooling point. With a microprocessor, for example, when the most heavily used cooling point is to be determined, all the cooling points can be queried in order and the most heavily used cooling point can be found by comparing the load data. Here, too, the term "continuous" does not necessarily mean continuous, but the exposure data can also be determined discretely in time.

It is preferable to filter the load over time. This avoids that load impacts, as can be caused, for example, by opening or closing a door of a refrigerated display case, result in an incorrect load pattern. Rather, the cooling capacity to be applied over a certain period of time is included in the load.

A temperature-dependent variable at the cooling point is preferably used as the load criterion. A temperature-dependent variable can be determined with relatively simple measures, namely with a temperature sensor that is almost always present. There is therefore no need to intervene in the refrigerant circuit. A display system by the process

Use, can therefore be easily installed in existing systems. The effort involved in determining the load therefore remains small.

Preferably used as a temperature dependent

size

A m - ^{Tγ T} CutO t ^al REL ~ m

'cutin ^ CutOut where T _{v is} the temperature at the cooling point, T _Cut0ut, for example, in the refrigerated showcase, a temperature at which the cooling body is switched off and ^τ c _u t i _n th ^e temperature at the cooling position is switched on. In other words, Δτ _REL expresses the deviation from the temperature T _Cut o _ut ^{to the} current cooling point, weighted in relation to the distance between the temperatures T _CutIn and T _Cut o _u t _{- in} particular in connection with a "two-point control" in which If the temperature T _Cut ι _{n is used} for switching on and the temperature T _Cut o _ut ^for switching off the cooling point, this load variable is relatively easy to obtain. Filtering can take place, for example, by _newly determining a load L from

_ ^L _{M t} * Const. + A _REL ^eu Konst + 1

where L _{ALT is} the previously determined and, if necessary, filtered load.

This filter eliminates sudden changes in the load expression.

After determining the most heavily loaded cooling point, this is preferably regarded as the most heavily loaded cooling point until its state changes or a predetermined time is exceeded, and only then is the most heavily loaded cooling point again determined. This procedure has the advantage that the effort for determining the most heavily loaded cooling point can be kept to a minimum. If a cold store is identified as the most heavily loaded, this first assumes that this cooling point cools. When the cooling point has reached the necessary or desired low temperature, it switches off. This change in state can be used as a signal to search again for the most heavily used cooling point. Another criterion would be, for example, that a defrosting process begins at this cooling point, which is either started from time to time or depending on other criteria, such as a frost load. Finally, it is also possible that the cooling point does not change its state over a predetermined period of time. In this case, a security barrier is used, so to speak, and after this predetermined period of time has elapsed, it is checked whether the cold store is still the most heavily used.

For each cooling point, the times are preferably summed up in which the cooling point in question is regarded as the most heavily used cooling point. This simplifies the evaluation and creation of a load pattern. The only thing that is checked is how long a cold store is to be regarded as the most heavily used cold store. Only these times are then used for the evaluation.

The evaluation can now be done in different ways. An immediate warning is preferably generated if a cooling point is the most stressed cooling point for more than a predetermined continuous period of time. This can indicate an error that should be checked as soon as possible. The predetermined coherent period of time may, for example, be of the order of an hour. Such an error occurs, for example, when the door of a cold room or a refrigerated display case is open has been left or goods have been stacked too high in the refrigerated display case. Such an error is not immediately noticeable. However, it has the consequence that the corresponding cooling point must be regarded as the most heavily used cooling point over a relatively long period of time. In this case, a warning is issued immediately so that the operator of the cooling system can intervene.

Alternatively or in addition, a long-term warning can be generated if a cooling point is the most stressed cooling point for more than a predetermined portion of a predetermined period. Usually, i.e. in the undisturbed case, it will turn out during a predetermined period, for example during the opening hours of a supermarket of 12 hours, that each cold store has a certain share of the times when it is the most heavily used cold store. Due to different locations or different loads, the proportions of the individual cooling points can of course shift somewhat. In most cases, this statistical division can be determined beforehand by calculation or by trial and error. If this division changes somewhat without external influences, such as regrouping of the showcases or the like, being apparent, then this indicates a developing error which must be investigated and, if necessary, eliminated.

Finally, a warning can be generated if the load pattern of a period differs by more than a tolerance range from a load pattern of a predetermined earlier period. On a statistical average, it can be assumed that the load pattern remains unchanged over time. smaller Fluctuations are of course possible at any time without having to indicate an error. However, if there are major deviations, these indicate an error that must be warned of. For this purpose, one can compare immediately successive periods, for example successive sales days of a supermarket. But you can also compare other periods with each other, for example all Mondays, Tuesdays, etc. However, with larger intervals, differences in the load patterns become more noticeable. You have to choose the time intervals so that a warning is generated early enough, but differences are clearly recognizable when they occur.

The invention is described below with reference to a preferred embodiment in connection with the drawing. Show here:

1 shows the schematic structure of a cooling system,

2 shows a schematic illustration to explain the temperature control of a cooling point,

Fig. 3 is a flow chart for explaining the control of an evaporator and

Fig. 4 is a flow chart for explaining the determination of a load pattern.

1 shows a cooling system 1 with a cooling circuit, which in the present case has three cooling points 2, 3, 4 connected in parallel. Each cooling point has an expansion valve 5, an evaporator 6 and an evaporator control unit 7. To indicate that if elements of the cooling points 2, 3, 4 of the same type need not be the same elements, these elements are additionally identified with a, b, c for the cooling points 2, 3, 4.

The evaporators are connected to a compressor system 8, which in turn is connected to a condenser 9 which has a plurality of fans 10 in order to dissipate heat. The capacitor 9 is connected to a collector 11. The compressor system 8 and the condenser are controlled by a control unit 13 which regulates the condenser pressure, for example with the aid of a pressure sensor 12.

The entire cooling system is controlled and / or monitored by a central control unit 14.

The basic mode of operation of such a cooling system is known. A refrigerant in the gas phase is compressed in the compressors 8. In the process, its temperature rises. The heated, compressed gas is passed through the condenser 9. There, the fans 10 dissipate heat so that the gas liquefies. The liquid coolant is collected in the collector 11, which acts as a coolant buffer. From the

Collector 11 passes the coolant through the expansion valves 5a, 5b, 5c into the evaporators 6a, 6b, 6c, where it evaporates while absorbing heat from the environment, in order to reach the compressors as gaseous coolant.

Each cooling point 2, 3, 4 is now controlled by the evaporator control unit 7a, 7b, 7c assigned to it. These local evaporator control devices 7a, 7b, 7c are connected to the central control unit 14. each Control unit 7 controls the expansion valve as a function of the temperature, which is determined with a temperature sensor 15, in the following manner, which is to be explained with reference to FIG. 2.

The temperature sensor 15 determines a temperature T _v , for example the temperature in a refrigerated display case. Two temperatures T _Cu i _n ^unc * T _Cu _ tout are specified as long as T _{v is} less than T _Cut i _{n /} no cooling takes place (t <t ₂ in FIG. 2), ie no coolant into the evaporator 6 is directed. If T _v exceeds the temperature T _Cut ι _n , the cooling of the cooling point begins (t ₂ <t <t ₃ ) and coolant is passed into the evaporator 6. Cooling lasts until the moment when T _v T again _Cut0ut ^un ~ terschreitet. At this moment (t ₃ ) the expansion valve 5 will close and only open when T _v again exceeds the temperature T _Cut ι _n .

For the following explanation, the evaporator is in the Tcu state when it cools and in the T _Cutout state when there is no cooling.

From time to time, ie at fixed time intervals, the respective evaporator 6 can be defrosted. Defrosting can be carried out in different ways, for example by electrical heating of the air surrounding the evaporator 6, or by hot refrigerant, which is not shown in detail, but is known per se. The frost that has built up on and around the evaporator will melt. This state of the evaporator during defrosting is called the defrosting state. In addition to the three types of state mentioned above, the cooling point can be in a closed state, for example when it is switched off, or it can be loaded with sensor errors, which the evaporator control unit 7 can determine in each case.

The compressor control unit 13 regulates the capacity of the compressor system based on the suction pressure, which is determined by the sensor 12. The compressor capacity of the system is increased or decreased, depending on the capacity required to maintain the desired suction pressure. The compressor control unit 13 also controls the fans 10, ie the ventilation system, so that the desired pressure in the condenser is maintained. This pressure can be caused by a

Increase or decrease in the air flow through the condenser can be achieved to achieve sufficient heat dissipation to the environment.

The central control unit 14 identifies the most heavily loaded cooling point based on the load determined by the respective evaporator control unit 7, which is explained further below, and warns an operator according to principles also specified below.

The specific way in which the load of the individual evaporators is determined only plays a minor role. Basically, the warning can be generated in very different systems. In addition to the above-mentioned states, a cooling system can also work with more or fewer states, for example with suction pressure control or modulating thermostat control, in which the degree of filling of the evaporator 6 is regulated to a predetermined temperature. on systems in which the evaporator, the compressor systems and the condenser ventilation are controlled by a central control unit, or on other types of cooling systems, for example brine systems or forced circulation systems.

The basis of the warning generation is the load on each individual cooling point. In the present example, the determination of the load takes place in the control unit 7 of each individual cooling point 2, 3, 4 and the difference between the desired temperature and the current temperature at the cooling point, ie T _v , is used as the starting point. Of course, other methods can also be used. The design of the cooling system often decides how the load must be determined. For the present invention, however, it is not critical how the current load is actually determined.

Fig. 3 shows schematically how the load is determined at each individual cooling point 2-4. First, it is examined whether the cooling point is currently defrosting, i.e. whether it is in the defrosting state. If the cooling point is currently in the defrosting state, go back to the beginning of the diagram. If the cooling point is not in the defrosting process, check whether it is in the c _u i _n state, i.e. is currently cooling. If this is not the case, the load is set to zero. This also applies if, for another reason, it cannot be positively determined that the cooling point is in the T _Cu i _n state.

If it is in the _Cu ι _n state, the load is determined according to the following scheme. One is calculates the current load using the following formula

L _{Al t} • Const. + ATszz

% eu Konst + 1

where ÄTREL means:

Ar ^T V ~ ^T CutOut

REL ^T C tIn ^{~ T} CutOut

Δτ _REL expresses the deviation from the temperature T _Cu - _t o _ut ^to ° - ^he current cooling passage from, said deviation is weighted taken to the bandwidth between ^τ C _u tm ^un d ^τ cutθut- ^Basically could already this expression as a value use for the load. However, in order to avoid sudden changes in the load expression, the calculation of the load expression is provided with a simple filter that avoids such sudden changes.

^L _A i _t i ^{st d} > - ^e load on the cooling point 2-4, as it resulted from the previous run. Then the current load L _{New is} saved.

The scheme shown in Fig. 3 is run through at predetermined small intervals. It is not necessary for one run to follow the other seamlessly. This scheme is handled locally in each evaporator control unit 7. At the start of this flow diagram, the temperature and the control status are registered. This step can also be done elsewhere in this flow chart. This means that information about the load on each cooling point is continuously available.

4 how a warning is generated will now be explained.

For this purpose, the sequence shown in FIG. 4 is also repeated.

It is first assumed that the most heavily used refrigeration point is known. Their temperatures (T _v , _Cu tou ^τ cutin) ^and their control status are registered. Then it is checked whether this cooling point is still in the T _Cut ι _n condition. It must have been before. Otherwise, it would not have been selected as the most heavily used refrigeration point. If it is in this state, a check is carried out to determine whether this state has already existed over a period t that is greater than a predetermined period t _{2 cycle} . If this is not the case, the loop returns to the

Beginning back.

If the condition of the cooling body has changed, or the cycle time t _Zy) ςi _us has been exceeded, the central control unit 14 again searches for the loaded most cooling point. Here, it simply compares the load information provided by the evaporator control units 7.

If it turns out that another cold store is the most heavily loaded, the cold store identification data and the time are saved. This data is then used to calculate load times when it is run again. If it turns out that the most heavily loaded cooling point is the same as before, a loading time t _{MLC is} calculated from the previously stored time and the now determined time. This loading time t _MLC is then checked to determine whether it exceeds a predetermined maximum value t _L i _M iτ. Such a maximum value t _LIMIT is typically one hour. If this value has been exceeded, a warning is issued. Such a warning can indicate, for example, that the door to a cold room or a refrigerated display case has been left open or that goods in a refrigerated display case are stacked too high.

The stored data is processed statistically, i.e. the duration of their exposure time is saved for the individual cooling point. However, this always only applies to the most heavily used refrigeration point.

This results in load times for all cooling points in the cooling system, i.e. Times in which one cold store is the most heavily used cold store. One can now statistically evaluate whether this portion of a refrigeration point in the total time, for example the opening period of a supermarket, changes over time or whether there are changes in such a way that this time increases. If such changes occur, a warning can be issued for the respective cooling point.

This warning informs the operator of the cooling system that there may be a problem. In most cases, he can solve this problem by maintenance or by changing the cooling point itself, for example by closing a door or the Modify a stack of goods, help. Other typical faults, such as a damaged fan, a defrosting heating element, a loss of coolant filling, can also be detected. Such errors are characterized by the fact that one cooling point or a group of cooling points can only maintain the desired temperature with problems and is therefore identified more frequently than the other as the most heavily used cooling point. If a cooling point has problems maintaining the desired temperature, this will not initially trigger a temperature warning, because such an error is not yet so serious that the cooling point cannot keep the temperature below the warning limit. With the procedure shown, however, it is possible to detect an error long before such an error triggers a temperature warning and urgently requires maintenance work.

No additional measures are required to operate the cooling system. You only have to evaluate existing sensors accordingly.

Claims

claims 1. A method for discovering faults in a cooling system with several cooling points, in which the most heavily loaded cooling point is continuously determined according to a predetermined scheme, characterized in that a load pattern of the cooling points is created and a if the load pattern deviates from at least one predetermined parameter Warning generated. 2. The method according to claim 1, characterized in that locally continuously at each cooling point Load determined. 3. The method according to claim 2, characterized in that temporally filtering the load. Method according to one of claims 1 to 3, characterized in that a temperature-dependent variable at the cooling point is used as the load criterion. A method according to claim 4, characterized in that as a temperature-dependent variable _Aτ _ Ty ~ Tcutout REL m _ m ¹ CutIn ¹ C toout used. 6. The method according to any one of claims 1 to 5, characterized in that after determining the most heavily loaded cooling point, it is regarded as the most heavily loaded cooling point until its state changes or a predetermined time has passed and only then again the most heavily used refrigeration point is determined. 7. The method according to any one of claims 1 to 6, characterized in that for each cooling point, the times are summed up in which the cooling part in question is regarded as the most heavily loaded cooling point. 8. The method according to claim 7, characterized in that an immediate warning is generated when a cooling point is the most heavily loaded cooling point for more than a predetermined continuous period. 9. The method according to claim 7 or 8, characterized in that a long-term warning is generated when a cooling point is the most heavily loaded cooling point for more than a predetermined portion of a predetermined period. 10. The method according to any one of claims 1 to 9, characterized in that a warning is generated if the load pattern of a period differs by more than a tolerance range from a load pattern of a predetermined earlier period.