Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
GAS MONITORING SYSTEM
Document Type and Number:
WIPO Patent Application WO/2021/043613
Kind Code:
A1
Abstract:
The present invention refers to an optimized method of monitoring a circuit breaker. Furthermore, the present invention refers to a circuit breaker being adapted to perform such method. Additionally, the present invention refers to a monitoring system. Furthermore, the present invention refers to the use of such circuit breaker or monitoring system.

Inventors:
RIVERA CLARO OSCAR JAVIER (DE)
Application Number:
PCT/EP2020/073636
Publication Date:
March 11, 2021
Filing Date:
August 24, 2020
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
SIEMENS ENERGY GLOBAL GMBH & CO KG (DE)
International Classes:
H01H33/56; G01M3/26; H02B13/055
Foreign References:
JP2001186613A2001-07-06
JP2007263584A2007-10-11
US20140055274A12014-02-27
EP3425385A12019-01-09
Attorney, Agent or Firm:
MAIER, Daniel (DE)
Download PDF:
Claims:
Patent claims

1. Method of monitoring a circuit breaker containing a gas with at least one gas characteristic providing a numerical value containing the steps of a) collecting a dataset referring to the at least one gas characteristic inside the circuit breaker, wherein the da taset contains the numerical value of the at least one gas characteristic during a specific condition or specific time of the day, b) calculation of a standard deviation of the at least one gas characteristic of the datasets of at least 3 days of the last 10 days, c) comparing the standard deviation of the gas pressure with a predefined threshold value, d) triggering a first action in case the standard deviation exceeds the threshold value.

2. Method according to claim 1, wherein the first action includes at least one action selected from the group consist ing of a notification of an operator, sending a first action dataset to a database, triggering a security mechanism and triggering a further measurement of the at least one charac teristic.

3. Method according to any of claims 1 to 2, wherein the first action is stopped in case an interaction with the gas has taken place within the period of collecting the datasets.

4. Method according to any of claims 1 to 3, wherein the collected dataset is stored in a database.

5. Method according to any of claims 1 to 4, wherein the datasets refer to the at least one gas characteristic inside the circuit breaker during a specific time or interval of the day and/or wherein the at least one gas characteristic con tains the gas density and/or the gas pressure. 6. Method according to any of claims 1 to 5, wherein the gas is SF6.

7. Method according to any of claims 1 to 6, wherein the predefined threshold value is at most 1 %.

8. Method according to any of claims 1 to 7, wherein the standard deviation of the at least one gas characteristic of the datasets of at least 4 days of the last 8 days.

9. Method according to any of claims 1 to 8, wherein the method includes e) in case the standard deviation is lower than the threshold value the datasets collected during a predefined prior time period are utilized to calculate a linear fitting line, f) wherein the dataset collected in step a) is compared to an upper limit of the linear fitting line, wherein the upper limit of the linear fitting line is the highest value of the linear fitting line within the predefined prior time period, wherein a second action is triggered in case the dataset col lected in step a) is less than 100% of the upper limit of the linear fitting line.

10. Method according to any of claims 1 to 9, wherein the method includes g) in case the standard deviation is lower than the threshold value the datasets collected during a predefined prior time period are utilized to calculate a linear fitting line providing a gradient, h) wherein the gradient is compared to a predefined gradient limit and/or adaptive gradient limit, and wherein in case the gradient is lower than the gradient limit a third action is triggered.

11. Circuit breaker containing a monitoring device being adapted to realize a method according to any of claims 1 to 10. 12. Monitoring system for a circuit breaker, wherein the monitoring system is adapted to perform a method according to any of claims contains 1 to 10.

13. Use of a circuit breaker according to claim 11 or a mon itoring system according to claim 12 to monitor the gas con tent of a circuit breaker.

14. Computer program product with program commands to per form the method according to any of claims 1 to 10.

15. Device for providing a computer program product accord ing to claim 14, wherein the device stores the computer pro gram product and/or provides the computer program product for further use.

Description:
Description

Gas monitoring system

The present invention refers to an optimized method of moni toring a circuit breaker. Furthermore, the present invention refers to a circuit breaker being adapted to perform such method. Additionally, the present invention refers to a moni toring system. Furthermore, the present invention refers to the use of such circuit breaker or monitoring system.

Circuit breakers are essential components in modern energy generation and distribution systems. Herein, providing the required security of the circuit breakers is especially im portant and has to be ensured even under extreme conditions.

A well established method to ensure the security of the cir cuit breakers includes the use of gases like SF 6 in such cir cuit breaker. Herein, such gas like SF 6 allows to cool the circuit breaker as well as to quench the arc. The use of SF 6 , for example, is a reliable and secure method to solve corre sponding problems. However, based on the high greenhouse ef fect the loss of SF 6 resulting from, for example, leakages are to be prevented.

To ensure that no detrimental environmental effect is result ing it is necessary to secure a safe handling and storage of the SF 6 gas. While the circuit breaker including its valves can be designed accordingly it is still a challenge to pre vent such loss SF 6 on the long term. While the system might be secure under the planned conditions minor failures during refill actions or contaminations in real life long term usage rendering valves slightly leaky may result in significant losses. Thus, it is still a task to provide a long term solu tion of enable a safe monitoring. Only providing a secure monitoring method allows to detect a leakage in time and to counteract. Yet it was noted that existing systems, for exam ple, still provide a ratio of accuracy and prevention of false blind results being to be optimized. These problems are solved by the products and methods as dis closed hereafter and in the claims. Further beneficial embod iments are disclosed in the dependent claims and the further description. These benefits can be used to adapt the corre sponding solution to specific needs or to solve further prob lems.

According to one aspect the present invention refers to a method of monitoring a circuit breaker containing a gas with at least one gas characteristic providing a numerical value containing the steps of a) collecting a dataset referring to the at least one gas characteristic inside the circuit breaker, wherein the da taset contains the numerical value of the at least one gas characteristic during a specific condition or specific time of the day, preferably a specific time, b) calculation of a standard deviation of the at least one gas characteristic of the datasets of at least 3 days of the last 10 days, c) comparing the standard deviation of the gas pressure with a predefined threshold value, d) triggering a first action in case the standard deviation exceeds the threshold value. Surprisingly, it was noted that utilizing such simple method allows to securely detect a be ginning loss of gas very early before major problems arise. Simultaneously, the rate of false positive gas loss detec tions is significantly decreased.

In case multiple gas characteristics providing a numerical value are determined according to the inventive method the action is triggered in case at least one numerical value ful fills the requirement. However, for typical embodiments it is preferred that the numerical values of at least 50%, more preferred at least 75%, of the gas characteristics fulfill the requirement to trigger the action, wherein the required number of the gas characteristics is rounded to an integer. For example, it can be preferred that at least 50% of 3 moni tored gas characteristics fulfill the specified requirement meaning 2 of the 3 gas characteristics. To provide a high se curity against false positive results it can even be pre ferred that the numerical values of all gas characteristics monitored with the inventive method fulfill this requirement to trigger the action. The aforementioned also applies to the embodiments described hereafter, especially those referring to the second and third action unless explicitly specified otherwise.

According to one aspect the present invention refers to a circuit breaker containing a monitoring device being adapted to realize an inventive method.

According to one aspect the present invention refers to a monitoring system for a circuit breaker, wherein the monitor ing system is adapted to perform an inventive method.

According to one aspect the present invention refers to a use of an inventive circuit breaker or an inventive monitoring system to monitor the gas content of a circuit breaker.

According to one aspect the present invention refers to a computer program product with program commands to perform the inventive method.

According to one aspect the present invention refers to a de vice for providing an inventive computer program product, wherein the device stores the computer program product and/or provides the computer program product for further use.

To simplify understanding of the present invention it is re ferred to the detailed description hereafter and the figures attached as well as their description. Herein, the figures are to be understood being not limiting the scope of the pre sent invention, but disclosing preferred embodiments explain ing the invention further.

Fig. 1 shows a scheme of an inventive method. Fig. 2 shows a scheme of a different embodiment of an in ventive method.

Fig. 3 shows a scheme of a further different embodiment of an inventive method.

Preferably, the embodiments hereafter contain, unless speci fied otherwise, at least one processor and/or data storage unit to implement the inventive method.

According to one aspect the present invention refers to a method as specified above.

It was noted that the inventive method can be beneficially utilized to secure an immediate action to, for example, pre vent a significant loss of SF 6 . According to further embodi ments it is preferred that the first action includes at least one action selected from the group consisting of a notifica tion of an operator, sending a first action dataset to a da tabase, triggering a security mechanism and triggering a fur ther measurement of the at least one characteristic. Such da tabase can be a distributed database providing a security against manipulation. This also applies to the embodiments as disclosed hereafter. For example, a blockchain system can be utilized to protect against manipulation of the entries of the database. For example, the database may further automati cally inform service personal to immediately plan some maintenance schedule to review and/or repair the circuit breaker. Herein, it has to be understood that sending such dataset to a database does not necessarily only include that said dataset is added to the database or replaces an existing dataset. It can, for example, also trigger an amendment of the data contained in the database like increasing the number of days of gas loss, increasing the amount of gas lost and/or resetting a value of gas contained in the circuit breaker to reflect some gas refill or the like.

The term "distributed database" as used herein refers to a decentralized database like a blockchain, a distributed ledg- er, a distributed data storage system, a distributed ledger technology based system, a manipulation proof database, a cloud, a cloud service, a blockchain in a cloud or a peer-to- peer database. Herein, such distributed database can be a public database like a public blockchain or a non public da tabase like a private blockchain. Typically, it is preferred that such blockchain is non public and can only be accessed by authorized persons. Herein, such access right might also be granted for external persons to allow a review of specific data like the gas loss statistics or overall gas loss to de termine whether corresponding regulations are fulfilled or not.

The term "data block" as used herein refers to a block of a distributed database like a blockchain or a peer-to-peer da tabase. It may contain data regarding the size of the data block, a block header, a counter of monitored data, data re garding the specific at least one gas characteristic, sensor data and/or monitored data. Said block header can, for exam ple, contain a version number, a checksum, or a time stamp.

Such method as described above can be beneficially utilized to provide an action triggering, for example, a notification of the operator. Surprisingly, it was noted that the corre sponding long term information allows to start a replacement process in time allowing to include the replacement process in normal maintenance processes reducing the overall costs while simultaneously raising the security and reliability. This also applies to the second action and third action as described hereafter.

For typical embodiments it was noted that the present method is especially useful to trigger a first action sending a first action dataset to a database. Such first action dataset send to the database can be a dataset recording a determined gas loss in the database and/or a dataset noting inside the database a gas refill. Automatically identifying such gas re fill is surprisingly beneficial. Although, the operator is typically enabled to enter such data manually it has been noted that it rarely happens. Utilizing the inventive method as described herein allows to securely detect such gas refill and automatically enter such data. This allows to greatly in crease the reliability and monitoring of such circuit break er. Also, such information can be utilized to trigger a se ries of collecting datasets to, for example, automatically provide a reliable new value of the current gas content of the circuit breaker. For example, five measurements with at least half an hour distance can be taken to detect the cur rently contained amount of gas after a refill. Allowing to greatly increase the reliability and safety of the gas moni toring. This especially also applies to the second action as described hereafter.

For further typical embodiments it was noted that the present method is especially useful trigger a first action triggering a further measurement of the at least one characteristic.

This allows to automatically provide an improved detection of abnormal data like some deviation resulting from an activity of the circuit breaker to trigger a further measurement of the at least one characteristic. This enables to ensure reli able datasets to be entered into the database to provide a consistent monitoring of the circuit breaker. Also, it allows to predict a gas loss in the early stages removing the risk that some circuit breaker activity might hide a starting gas loss. This also applies to the second action as described hereafter.

Furthermore, it was noted that it is typically beneficial to include a review process to further reduce the number of in correctly triggered first actions. Based on certain interac tions automatically influencing the gas characteristics it is possible to provide corrective data to, for example, identify an allowed deviation from the expected behavior and prevent such false positive detection. According to further embodi ments it is preferred that the first action is stopped in case an interaction with the gas has taken place within the period of collecting the datasets. Such interaction includes, for example, a refill of the gas content or replacement of the gas content of the circuit breaker or a complete exchange of the circuit breaker. For example, a corresponding action can be entered in the database, wherein the method receives the corresponding information and avoids an unnecessary noti fication of the operator. Also such information can be uti lized to trigger a series of collecting datasets to, for ex ample, automatically provide a reliable new value of the cur rent gas content of the circuit breaker. This allows to re duce the work to the operator to enter a gas refill, while the reliable measurement of the change of the gas content is automatically performed not requiring additional interaction.

Additionally, it was noted that the inventive method can ben eficially include a unique identifier of the circuit breaker to correlate the datasets to a specific circuit breaker. Sur prisingly, it was noted that during long term usage short no tice replacements of the circuit breakers can take place re sulting in deviations from the normal behavior. Also, bene fits are obtained for handling a corresponding circuit break er pool, wherein circuit breakers requiring a more intense maintenance or even replacement can be easily identified. Ac cording to further embodiments it is preferred that the da taset contains a unique identifier of the circuit breaker.

It was further noted that such collected datasets are benefi cially stored in a database like a distributed database. Some central storage possibility allows to easily monitor and com pare the available data. Using typically available communica tion methods allows to also include review processes by ex perts to more closely evaluate the correspondingly received data while corresponding actions performed by local field personal and operators are swiftly forwarded and executed. According to further embodiments it is preferred that the collected datasets are stored in a database, more preferred a distributed database.

It was further noted that acquiring the datasets in correla tion to a specific time or time interval provides very bene ficial results for typical embodiments. While corresponding effects like influences from the surrounding temperature should be expected to be less emphasized based on such trivi al aspect it was surprisingly noted that the influence was big enough to give a major benefit for typical applications. According to further embodiments it is preferred that the da tasets refer to the at least one gas characteristic inside the circuit breaker during a specific time or interval of the day. Typically, it is preferred that the datasets are col lected during nighttime. Herein, it was noted that it further smoothened transition over longer time periods is observed and more reliable data is acquired. According to further em bodiments it is preferred that the datasets are collected within the time period from 11 p to 6 am, more preferred from 1 am to 5 am, even more preferred from 2 am to 5 am. Surprisingly, it was further noted that no significant devia tions have been observed for different times of the year. Ap parently, the inventive scheme is utilizing the correct time frame to remain highly sensitive while simultaneously ignor ing slow overall changes of the temperature resulting from the seasons.

While the typical gas characteristics available to the person skilled in the art may be utilized for the inventive method it was noted that certain gas characteristics are especially useful. Surprisingly, they are not only easy to measure, but also provide an especially high reliability. According to further embodiments it is preferred that the at least one gas characteristic contains the gas density and/or the gas pres sure.

While the inventive method can be utilized with different gases available it was noted that the inventive benefits are especially high for SF 6 taking into account typical applica tion cases. According to further embodiments it is preferred that the gas is SF 6 - Surprisingly, the inventive method pro vides an especially high sensitivity and low amount of false positive detection in this case. Taking into account that the loss of SF 6 should especially be restricted based on environ ment reasons applying the inventive method in this context renders retrofits or replacement of existing systems utiliz ing this gas especially interesting.

Additionally, it was noted that despite differences to be ex pected a generic threshold value surprisingly can be utilized for typical applications, especially utilizing SF 6 . This al lows to easily apply the inventive methods to different cases without requiring adaptions still providing an improved reli ability. According to further embodiments it is preferred that the predefined threshold value is at most 1 %, more pre ferred at most 0.5 %, even more preferred at most 0.1 %.

Furthermore, it was noted that a further improvement of the sensitivity can be achieved utilizing an optimized number of days utilized to calculate the standard deviation. According to further embodiments it is preferred that the standard de viation of the at least one gas characteristic of the da tasets of at least 4 days of the last 8 days, more preferred the least 5 days of the last 7 days, is calculated. Typical ly, it is even preferred that the standard deviation is cal culated based on the datasets of each of the last 5 days.

Unless specified otherwise terms like "calculate", "process", "determine", "generate", "configure", "reconstruct" and com parable terms refer to actions and/or processes and/or steps modifying data and/or creating data and/or converting data, wherein the data are presented as physical variable or are available as such.

However, it can also be beneficial to increase the number of days included in the calculation of the standard deviation. For example, such embodiments can be especially usefully uti lized in application cases providing a more consistent tem perature profile like locations at the sea. Taking into ac count the specific location and condition of the application allows to even further increase the sensitivity and reliabil ity of the method. According to further embodiments it is preferred that the standard deviation of the at least one gas characteristic of the datasets of at least 5 days, more pre- ferred at least 7 days, even more preferred at least 9 days, of the last 10 days is calculated.

Furthermore, it was noted that even in case no loss of gas is noted using the steps of the inventive method as specified above a further method step can be beneficially performed to identify abnormal behavior. According to further embodiments it is preferred that the method includes e) in case the standard deviation is lower than the threshold value the datasets collected during a predefined prior time period are utilized to calculate a linear fitting line, f) wherein the dataset collected in step a) is compared to an upper limit of the linear fitting line, wherein the upper limit of the linear fitting line is the highest value of the linear fitting line within the predefined prior time period, wherein a second action is triggered in case the dataset col lected in step a) is less than 100%, more preferred less than 99%, even more preferred less than 98.2%, of the upper limit of the linear fitting line. For example, such predefined pri or time period can be at least 30 days, like 30 days, at least 60 days, like 60 days, or at least 90 days, like 90 days. To provide a very reliable linear fitting line it is preferred for typical embodiments that the linear fitting line is calculated for a period of at least 60 days, more preferred at least 90 days. The linear fitting line is calcu lated as known to the person skilled in the art. For example, the mean error square is minimized to find the most suitable linear fitting line possible.

According to typical embodiments it is preferred that the second action contains sending a notification of an operator and/or sending a second action dataset to a database and/or triggering a security mechanism, more preferred sending a no tification of an operator and/or sending a second action da taset to a database. This allows to note, record and/or react to even minor deviation at an early stage for even the first detected abnormality. Additionally or alternatively it is possible to obtain long term evaluation data utilizing such datasets. Herein, an evaluation is preferably triggered in case the standard devi ation is lower than the predefined threshold. The evaluation includes calculating a linear fitting line and evaluating said linear fitting line to allow an overall assessment of the gas loss behavior of the circuit breaker. According to further embodiments it is preferred that the method includes g) in case the standard deviation is lower than the threshold value the datasets collected during a predefined prior time period are utilized to calculate a linear fitting line providing a gradient, h) wherein the gradient is compared to a gradient limit, and wherein in case the gradient is lower than the gradient limit a third action is triggered. The gradient limit can, for ex ample, be a predefined gradient limit, an adaptive gradient limit or a combination of both. Herein, such gradient limit can be predefined and be selected based on the circuit break er to be monitored. While the person skilled in the art may select a corresponding value based on his experience and cor responding test it was surprisingly noted that said gradient values can also be selected based on corresponding standards as described hereafter. This greatly simplifies retrofitting or setting up a new circuit breaker or taking over existing circuit breakers in case the corresponding data is limited and time for tests is limited. For example, for embodiments of high quality circuit breakers fulfilling new standards it is typically preferred that the gradient limit is selected from the range from -0.01 to -0.0005, more preferred from -0.01 to -0.001. For example, for embodiments monitoring es pecially short time periods like 30 days it is typically ben eficially that the gradient limit is selected from the range from -0.02 to -0.001, more preferred from -0.015 to -0.0015. In case, for example, an older circuit breaker providing a less secure sealing is monitored over a longer time the gra dient limit is preferably selected from the range from -0.025 to -0.0015, more preferred from the range from -0.021 to -0.003. In case such older circuit breaker is monitored for a shorter period like 30 days it is typically preferred that the gradient limit is selected from the range from -0.04 to -0.005, more preferred from -0.04 to -0.001. Also, an adap tive gradient limit based on the collected data can be uti lized. Herein, the gradient for the first predefined prior time period defined is calculated and the gradient limit is set correlating to this. For example, the gradient limit can be calculated as follows: gradient limit = calculated gradient — (X x ( calculated gradient) 2 ) wherein X is an adaption factor being, for example, selected from the range from 0 to 10, more preferred from the range selected from 0 to 7, more preferred selected from the range from 0 to 5. Also, this system can be combined utilizing such predefined gradient limit first and changing to such adaptive gradient limit after acquiring the data for the first prede fined prior time period. For example, the predefined prior time period can be at least 30 days like 30 days, at least 60 days like 60 days or at least 90 days like 90 days. To pro vide a very reliable linear fitting line it is preferred for typical embodiments that the linear fitting line is calculat ed for a period of at least 60 days, more preferred at least 90 days. The linear fitting line is calculated as known to the person skilled in the art. For example, the mean error square is minimized to find the most suitable linear fitting line possible.

The third action typically preferably contains sending a no tification of an operator, sending a third action dataset to a database and/or triggering a security mechanism, more pre ferred sending a notification of an operator and/or sending a third action dataset to a database. This allows to note, rec ord and/or react to long term changes like in case the seal ing of the circuit breaker beginning to worsen or the gas filling unit becoming less secure. This especially allows to easily monitor a complete fleet of circuit breakers and to statistically identify circuit breakers to be serviced or even replaced to prevent a loss of the gas contained therein.

Surprisingly, it was noted that including an automatic for warding of the dataset significantly increased the overall reliability of the process. According to further embodiments it is preferred that the circuit breaker is adapted to for ward the collected dataset automatically. This allows to re duce the amount of mistakes being especially problematic for handling a gas like SF 6 .

According to another aspect the present invention refers to a circuit breaker as specified above.

Herein, it was noted that it is typically beneficial to in clude an automatic mechanism collecting the required da tasets. For example, implementing a corresponding hardware or software stored on an integrated processing unit allows to prevent a data loss and to optimize the complete method. Ac cording to further embodiments it is preferred circuit break er is adapted to automatically collect the datasets.

While it is preferable to replace existing circuit breakers providing a retrofit for existing systems is also a highly interesting possibility. It was noted that, for example, a corresponding monitoring element being adapted to collect the dataset can be easily attached to the gas filling device like a valve of such circuit breaker being adapted to refill the circuit breaker. Preferably, such monitoring element is adapted to send the collected dataset to a processing unit or data storage or contains a processing unit or data storage. This allows to directly perform the corresponding evaluation steps of the inventive method. Such retrofit especially al lows to provide the inventive monitoring capability for an existing fleet of circuit breakers.

The term "processing unit" as used herein refers to data pro cessing units as used for processing data. Herein, for exam ple, calculations, checksums and cryptographic checksums are generated, measured and predefined values are compared, a re action to a specific situation a determined, an output is generated, a part of a data set is reconstructed, a checksum, preferably cryptographic checksum, is validated, new blocks for a blockchain are generated, new blocks are integrated in the blockchain, and so on. Such processing unit can, for ex ample, be found in computers, clients, smart phones, and servers. For example, such processing unit can also be found in knots of the distributed database like a blockchain.

According to a further aspect the present invention refers to a monitoring system as specified above.

According to further embodiments it is preferred that the monitoring system contains an inventive circuit breaker.

It was noted that it is beneficial to provide the inventive monitoring system with a database allowing to store the col lected datasets. According to further embodiments it is pre ferred that the monitoring system contains a database, pref erably a distributed database, wherein the database is adapted to store the datasets.

Furthermore, it was noted that it is beneficial to already include a corresponding processing unit in the monitoring system to calculate the required standard deviation and/or perform the further evaluations. According to further embodi ments it is preferred that the monitoring system contains a processing unit being adapted to calculate the standard devi ation.

Simultaneously, it was further noted that it is also benefi cial to include a corresponding processing unit in the moni toring system to calculate the linear fitting line to provide a more reliable system not relying on external resources.

This allows to reduce the data transfer and also prevents that the data can easily be tampered with.

Also, it was surprisingly noted that further applications can be provided by improving the reliability and security of the data forwarded. For example, it was surprisingly noted that the inventive methods allow to be specifically adapted for utilization in maintenance applications by providing secured data that allows a reliable assessment of the current situa- tion. Additionally, this allows to store reliable data the that might be utilized in case of later failures to provide evidence of the state of the circuit breaker if required. Ac cording to further embodiments it is preferred that the moni toring system contains a hardware being adapted to provide a secure data connection between the circuit breaker and the database. For example, such system can be realized using an end-to-end encryption between the circuit breaker and the da tabase.

According to another aspect the present invention refers to a use as described above.

According to one aspect the present invention refers to a computer program as specified above.

According to one aspect the present invention refers to a de vice for providing an inventive computer program product as specified above.

The present invention was only described in further detail for explanatory purposes. However, the invention is not to be understood being limited to these embodiments as they repre sent embodiments providing benefits to solve specific prob lems or fulfilling specific needs. The scope of the protec tion should be understood to be only limited by the claims attached.

Figure 1 shows a scheme of an inventive method. In the fol lowing the features of an embodiment utilizing such scheme are included for illustrative purposes. However, other embod iments as specified above may be utilized for such scheme.

The method is performed utilizing an inventive circuit break er being part of an inventive monitoring system. The calcula tions are performed using a device providing a processor and a data storage executing a computer program product to trig ger or execute the specified action. Herein, the dataset containing the numerical value of the gas density of SF 6 in a circuit breaker is determined and stored on a data storage during step 2. The determination is per formed each day at a specific time at 4 am. Utilizing the collected dataset and datasets stored in the dataset collec tion 1 a standard deviation is calculated in step 3. Herein, the standard deviation is calculated based on at least three datasets collected during the period of the last ten days, wherein the present day is deemed to be one of these ten days. All datasets collected during this time period are uti lized to calculate the standard deviation.

The standard deviation is evaluated in step 4 whether it ful fills the requirement of the predefined threshold being 0.5%. In case it exceeds the threshold step 5 takes place including triggering a first action. The first action includes a noti fication to the operator informing him/her about this devia tion and sending a first action dataset to a database. Here in, the first action dataset is either stored as it is or ac cording to the first action dataset a corresponding database dataset is changed. For example, the number of days providing a specified gas loss is increased by 1. Furthermore, the first action includes a determination step, wherein the da taset is evaluated with regard to the overall security taking into account the determined gas loss. In case certain limits specified by the operator are exceeded a replacement circuit breaker takes over and the monitored circuit breaker is taken off the grid. Simultaneously, a corresponding notification is send to the operator and a maintenance or replacement is scheduled. Before, during or after the first action the da taset is stored in the dataset collection 1.

In case the standard deviation does not exceed the predefined threshold the linear fitting line for the predefined prior time period being the last 60 days is calculated in step 6. Herein, the datasets available for the last 60 days, wherein the present day is deemed to be one of these days, are re trieved from the dataset collection 1 and including the da taset collected in step 2 are utilized to provide the linear fitting line and its upper limit. The upper limit specifies the highest value of the linear fitting line within the pre defined prior time period. As it is a linear line the upper limit can be either located at the first or last day of the linear fitting line so that it represents the value for ei ther day 1 or day 60. This also applies in case day 60 is the present day and day 1 has not been measured. In this case, the theoretical value resulting from the linear fitting line is providing the upper limit. The numerical value of the gas density collected in step 2 is compared to the upper limit in step 7 and in case it is lower than 99% of the upper limit step 8 is performed including triggering a second action. The second action includes a notification to the operator to in form him/her about the deviation and draw his attention to this observation. Furthermore, a second dataset is send to a database to update a gas loss protocol to eventually adapt the prediction of the future gas loss and required new SF 6 gas for refill actions.

Figure 2 shows a scheme of another embodiment of the in ventive method. In the following the features of an embodi ment utilizing such scheme are included for illustrative pur poses. However, other embodiments as specified in the de scription may be utilized for such scheme. The method is per formed utilizing an inventive circuit breaker being part of an inventive monitoring system. The calculations are per formed using a device providing a processor and a data stor age executing a computer program product to trigger or exe cute the specified action.

Herein, the dataset containing the numerical value of the gas pressure of SF 6 in a circuit breaker is determined and stored on a data storage during step 2'. The determination is per formed each day during a specific time period being from 1 am to 5 am. Utilizing the collected dataset and datasets stored in the dataset collection 1 a standard deviation is calcu lated in step 3'. Herein, the standard deviation is calculat ed based on at least four datasets collected during the peri od of the last eight days, wherein the present days is deemed to be one of these eight days. All datasets collected during this time period are utilized to calculate the standard devi ation.

The standard deviation is evaluated in step 4' whether it fulfills the requirement of the predefined threshold being 1%. In case it exceeds the threshold step 5' a review step takes place, wherein a database is reviewed whether some in teraction with the gas content of the circuit breaker has been noted down. For example, in case a gas refill has been noted down the process is either stopped or a second review step takes place to evaluate the deviation and whether it possibly results from this interaction. For example, this can be partially manually by automatically triggering a corre sponding request for the operator or fully automatic by per forming a logic check like whether the change is within the required and expected boundaries. In case no such interaction is noted down step 5' includes triggering a first action. The first action includes a notification to the operator inform ing him/her about this deviation and sending a first action dataset to a database. Herein, the first action dataset is either stored as it is or according to the first action da taset a corresponding database dataset is changed. For exam ple, the number of days providing a specified gas loss is in creased by 1. Furthermore, the first action includes a deter mination step, wherein the dataset is evaluated with regard to the overall security taking into account the determined gas loss. In case certain limits specified by the operator are exceeded a replacement circuit breaker takes over and the monitored circuit breaker is taken off the grid. Simultane ously, a corresponding notification is sent to the operator and a maintenance or replacement is scheduled. Before, during or after the first action the dataset is stored in the da taset collection 1 .

In case the standard deviation does not exceed the predefined threshold the linear fitting line for the predefined prior time period being the last 90 days is calculated in step 6'. Herein, the datasets available for the last 90 days, wherein the present day is deemed to be one of these days, are re trieved from the dataset collection 1' and including the da taset collected in step 2' are utilized to provide the linear fitting line and its upper limit. The upper limit specifies the highest value of the linear fitting line within the pre defined prior time period. As it is a linear line the upper limit can be either located at the first or last day of the linear fitting line so that it represents the value for ei ther day 1 or day 90. This also applies in case day 90 is the present day and day 1 has not been measured. In this case, the theoretical value resulting from the linear fitting line is providing the upper limit. The numerical value of the gas pressure collected in step 2 is compared to the upper limit in step 7' and in case it is lower than 99% of the upper lim it step 8' is performed including triggering a second action. The second action includes a notification to the operator to inform him/her about the deviation and draw his/her attention to this observation. Furthermore, a second dataset is sent to a database to update a gas loss protocol to eventually adapt the prediction of the future gas loss and required new SF 6 gas for refill actions.

Furthermore, the gradient of the linear fitting line is com pared to a gradient limit in step 9'. It is determined wheth er the gradient is lower than the gradient limit being se lected from the range from -0.01 to -0.0005 like -0.001. In case the gradient is lower step 10' takes place including triggering a third action. The third action includes notify ing the operator. Furthermore, it includes sending a third action dataset to a database. Herein, the corresponding data is collected for the complete fleet of circuit breakers available at the site. Based on the data obtained maintenance and replacement plans are scheduled, wherein batches of cir cuit breakers are addressed. Herein, it is possible to deter mine the health and near future development of the circuit breakers with surprisingly high accuracy.

Figure 3 shows a scheme of another embodiment of the in ventive method. In the following the features of an embodi- ment utilizing such scheme are included for illustrative pur poses. However, other embodiments as specified in the de scription may be utilized for such scheme. The method is per formed utilizing an inventive circuit breaker being part of an inventive monitoring system. The calculations are per formed using a device providing a processor and a data stor age executing a computer program product to trigger or exe cute the specified action.

Herein, the dataset containing the numerical values of the gas pressure and gas density of SF 6 in a circuit breaker is determined and stored on a data storage during step 2''. The determination is performed each day during a specific condi tion like a predefined room temperature of the room contain ing the circuit breaker. Utilizing the collected dataset and datasets stored in the dataset collection 1'' a standard de viation is calculated in step 3''. Herein, the standard devi ation is calculated based on at least four datasets collected during the period of the last five days, wherein the present days is deemed to be one of these five days. All datasets collected during this time period are utilized to calculate the standard deviation.

The standard deviations of the numerical value of the gas pressure and gas density are evaluated in step 4'' whether they fulfill the requirements of the predefined thresholds being 1%. In case both exceed the threshold a first action is triggered in step 5''. The first action includes a notifica tion to the operator informing him/her about this deviation and sending a first action dataset to a database. Before, during or after the first action the dataset is stored in the dataset collection 1''.

In case the standard deviation does not exceed the predefined threshold the linear fitting lines for the predefined prior time period being the last 30 days are calculated in step 6'. Herein, the datasets available for the last 30 days, wherein the present day is deemed to be one of these days, are re trieved from the dataset collection 1' and including the da- taset collected in step 2' are utilized to provide the linear fitting line and its upper limit. The upper limit specifies the highest value of the linear fitting line within the pre defined prior time period. As it is a linear line the upper limit can be either located at the first or last day of the linear fitting line so that it represents the value for ei ther day 1 or day 30. This also applies in case day 30 is the present day and day 1 has not been measured. In this case, the theoretical value resulting from the linear fitting line is providing the upper limit. The numerical values of the gas density and gas pressure collected in step 2'' are compared to the upper limit in step 1'' and in case at least one is lower than 98.2% of the upper limit step 8'' is performed in cluding triggering a second action. The second action in cludes a notification to the operator to inform him/her about the deviation and draw his attention to this observation. Furthermore, a second dataset is sent to a database to update a gas loss protocol to eventually adapt the prediction of the future gas loss and required new S 6 gas for refill actions.

In case the requirement with regard to the upper limit is fulfilled the linear fitting lines are further utilized in step 9'' for a further evaluation. Herein, the gradients of the linear fitting lines are compared to a corresponding gra dient limit. It is determined whether the gradients are lower than the gradient limit being selected from the range from -0.02 to -0.001 like -0.01. In case both gradients are lower step 10'' takes place including triggering a third action. The third action includes notifying the operator and sending a third action dataset to a database.