Leak detection method

Abstract

The present invention relates to a method for detecting leaks in an object through which a medium flows, in particular a pipe or pipeline, by which patterns are identified in ascertained values ​​of changes in the flow rate and pressure of the medium, as well as in the temperature change, and based on the identified patterns, a probability of the presence of a leak is determined due to a self-learning system.

Description

Technical Field

[0001] The present invention relates to a leakage detection method according to the preamble of claim 1 and a computer program product according to the preamble of claim 12.

Background Art

[0002] When a leak occurs in a pipeline, it is often extremely important economically to detect, discover, and seal this leak quickly and safely. In particular, in the case of a pipeline system that is typically divided into pipeline segments but has portions extending between continents and transports large quantities of potentially environmentally harmful products (such as crude oil), generally, quickly sealing the leak that has occurred is also extremely important from the perspective of environmental protection.

[0003] Based on the principle of mass conservation, known methods of leak detection generally basically involve forming a mass flow balance. In principle, the amount of transport fluid entering a pipe section should also completely reappear again from the end of the said pipe section assuming there is no leak in the relevant section. Therefore, under ideal conditions, an imbalance between the inflowing mass flow rate and the outflowing mass flow rate indicates the presence of a leak.

[0004] However, with this method, the exact location of the leak along the relevant pipe section may not be easily confirmed. Furthermore, this idealized principle may not be fully applicable to an actual pipeline system. In particular, the influence of various environmental factors becomes a problem. In some cases, as a result of the fairly long lengths of pipelines or pipeline segments spanning thousands of kilometers, for example, sections of large pipelines often pass through multiple climate zones. In particular, the temperature along the pipeline, which varies from region to region and changes over time, can be a significant amount of interference for checking the mass flow balance, depending on the product being transported.

[0005] As a result of the thermal expansion of the transported fluid, which depends on the individual products in each case, the mass flow balance can be negative or positive even in the absence of leakage. In this case, the natural volume of the pipeline provides a certain buffering effect. When the pipeline passes through cold regions, the transported fluid will contract in these areas. For example, as a result of rainfall, or in the case of agricultural use on the pipeline, when the ground is shaded at various levels by tillage and harvesting areas, a similar effect occurs due to the local change in the ambient temperature of the pipeline in a short period. If this effect is ignored when checking the mass flow balance, the loss of the transported fluid will be recorded even though there is no leakage.

[0006] From an economic perspective, for example, it is almost impossible to fully monitor a pipeline network with a large total line length and a complex branch structure with respect to all relevant factors. Therefore, leaks that occur are rarely detected immediately by sensors. Furthermore, environmental factors and the thermodynamic properties of the medium usually cannot be detected to an appropriate extent to enable an accurate statement regarding the correction of the mass flow balance confirmed for the measurement section. For this reason, various methods are known for allowing or correcting the variations that occur through the statistical processing of the confirmed data. This is aimed at improving the identification of leaks under actual conditions.

[0007] To take into account the thermodynamic changes along the pipeline or the transported fluid, for example, a method of modeling the processes that occur and the related influencing factors using a real-time model is required. The corresponding method is known, for example, by the name "Real-Time Transient Model" (RTTM). In some cases, the known method also makes it possible to locate leaks in a specific area, for example, by detecting the propagation pressure wave that appears when a leak occurs.

[0008] However, in this case, there is always the drawback that the reliability of the results confirmed by statistical means is often low. This is especially true when only a relatively small amount of data can be used, or when it is necessary to evaluate by a single measurement. In the case of a leak that has not been erroneously detected, the high risks related to economy, environment, and safety usually mean that a decision is made to conduct a manual inspection of the relevant pipeline section when in doubt. This often requires a team of service engineers to take risks and travel long distances in difficult terrain, for example, to inspect an onshore pipeline. Naturally, it is desirable to avoid the related risks to the human body and the environment, and in some cases, the significant costs. Summary of the Invention Problems to be Solved by the Invention

[0009] Against such a background, an object of the present invention is to improve the reliability of leak detection based on the confirmed data.

[0010] The above object is achieved by the method according to claim 1 and the computer program product according to claim 12. Advantageous developments are, in each case, the subject matter of the dependent claims. Means for Solving the Problems

[0011] The method according to the present proposal first includes that a series of values that form the basis for subsequent evaluation are confirmed. These values include at least the flow rate change, pressure change, and temperature value change of the product or medium transported by the flow transfer object. Structurally, the flow transfer object (especially a pipe or pipeline) is divided into one or more measurement sections. This method includes that a plurality of measurement points are defined in each measurement section. Preferably, one measurement point is arranged in the initial region and the final region of the measurement section, respectively. Here, the values of the above physical quantities and, if necessary, the values of further physical quantities are confirmed at each measurement point.

[0012] The desired value can be confirmed by direct and / or indirect measurement of a physical quantity. In this case, it is advantageous to make the resolution of the recording, particularly the time resolution, as high as possible. However, alternatively or additionally, data generated in another way, particularly simulated data, can also be applied as the value of the method according to the invention or assigned to the measurement points.

[0013] According to the invention, it is self-evident that, as an alternative to or in addition to confirming the absolute value of the relevant physical quantity, the relative value and the change in the applicable physical quantity can also be confirmed in each case. Changes, particularly changes over time, provide information about the dynamics of the process occurring and are therefore more important for the method than merely the absolute value of the physical quantity.

[0014] Preferably, the evaluation of whether there is an undesirable loss of the volume of the medium is not essentially based on static considerations regarding the actual state. Instead, the method according to the invention involves, in particular, the use of dynamic models. For this reason, mainly the change in the confirmed quantity, particularly the temporal change, is extremely important.

[0015] The change in the flow rate of the medium, i.e., the change in the flow rate of the fluid transported within the flow transfer object, is particularly mass-based but can also be understood on a volume basis. Furthermore, the pressure change can be related to the hydrostatic pressure and / or the dynamic pressure. The underlying temperature value or its change relates in particular to the ambient temperature around the measurement point, but alternatively or additionally, it can also directly reflect the temperature change of the medium at the measurement point. Since the flow of the medium generally changes greatly depending on the existing temperature gradient and / or pressure gradient, the recording of the pressure and temperature changes is relatively very important.

[0016] In addition to the physical quantities cited, it is also possible to confirm the values of further quantities, such as the flow rate of the medium or its density, or the external ambient pressure in the region of the measurement point.

[0017] If the measured value is not available or the amount available is too small, the value for the desired amount can be interpolated based on the measured values from adjacent or nearby measurement points.

[0018] As an alternative to or in addition to actual measurements, the values at the measurement points can also be determined, in particular by means of modeling. In particular, the flowing transfer object is preferably modeled including the medium through which it is flowing in this case. This makes it possible, for example, to simulate the occurrence of specific values of the physical quantity of interest and means that the effects on the flowing transfer object and / or the medium can be determined based on the underlying model.

[0019] Normally, both the measured values and the values determined by means of modeling or simulation are basically subject to the influence of uncertainties, i.e., stochastic errors and / or systematic errors. For this reason, the method according to the invention can be used, in particular, to express it in the form of probabilities.

[0020] Instead of or in addition to the real-time model, values can also be generated by means of forward modeling. This particularly includes the use of an iterative method, by means of which the values available at the measurement points and the influence of such values are predicted. The number of iterative steps can basically be selected according to the requirements for the accuracy of the calculation in the individual case.

[0021] In a preferred configuration of the modeling employed, the possible trends for the overall system and / or the individual parameters can be approximated, in particular, by determining conditional probability values. In this context, in particular, estimation methods such as Bayesian statistics methods and / or maximum likelihood methods can be included.

[0022] In particular, in a one-dimensional model of a flow-transported object, preferably a pipeline or a pipeline system, it is possible to model the verified, modeled, and / or simulated data in a simple way. This may involve, in particular, a selective reduction of the data up to a certain range and / or a targeted combination of the data. This can furthermore be based on weighting in order to define the extent to which the data used affects the degree adopted in the model or the modeled result. In general, the corresponding simplification for the one-dimensional model enables a considerably simplified and thus more reliable detection of the important effects to be observed.

[0023] It is obvious that high-dimensional models and / or combinations of a plurality of one-dimensional and / or high-dimensional models can also be employed in an equivalent way. In principle, the reproduction or use of usually extensive available data for highly simplified models is advantageous with respect to the method according to the invention. A reduced representation according to the current or likely future trends therein enables, in particular for the user, a very reliable detection or evaluation of anomalies regarding the state and / or operation of the pipeline system. In this case, ultimately, the level of simplification required can be determined, in particular, based on the specific application situation in each individual case.

[0024] A real-time model and / or forward modeling of the state of the flow transfer object under consideration can preferably be used to determine the optimal value for the spatial and / or temporal density of the measurement points for capturing the data to be considered. The measurement point density is distributed non-uniformly, in particular, over the entire flow transfer object under consideration or over a specific measurement section. By verifying the optimal density, as a result of unnecessarily redundant capture, an excess of data volume occurs, and it is possible to locally and / or temporally collect a sufficient amount of data for evaluating the current and / or future state without the transmission, storage, and processing thereof requiring time and cost. For example, if a special need for a reliable assessment of a potential critical situation in a high-risk area is a local provision for a higher density of measurement points, the modeling enables the determination of the respective economically optimal degree at which appropriate data collection is performed in this regard.

[0025] When the values of the underlying physical quantities are verified at different measurement points of the measurement section, the method includes forming at least one numerical group from these values. The numerical group can ultimately include the total set of values recorded or otherwise verified, or can be formed by these lower-order numerical groups (subgroups).

[0026] The verified values are supplied to a data processing device for evaluation. The data processing device can be a computer locally present in the vicinity of the measurement section. However, particularly preferably, it is to centrally process the data from different measurement points and / or measurement sections by a common data processing device. The data processing device can further be a network including a plurality of interacting computers. In particular, the data processing device is preferably provided at a location physically remote from the measurement section to be monitored, for example, a central computer center. Therefore, data processing can also be performed, for example, based on the principle of cloud services.

[0027] A numerical group is inspected by a data processing device as to whether the values within the numerical group form a pattern or a pattern is formed within the numerical group. If a pattern is identified in the numerical group by the data processing device, the pattern, in particular its type and its characteristic strength, can be used to determine the likelihood of the presence of leakage in the measurement section of the flowing transfer object. This enables in particular heuristic leakage detection, so that even if the evaluation of the available data using known statistical methods does not yield reliable results, leakage occurring in the measurement section under consideration can be detected.

[0028] In particular, the method according to the invention can be used on the one hand to distinguish between patterns associated with changes in the flow rate and / or temperature of the medium as a result of environmental influences and, on the other hand, patterns associated with undesirable losses of flow due to leakage or unauthorized tapping (Entnahme (German), tapping (English), extraction). The aim in this case is to be able to respond as quickly as possible to each situation in order to keep the losses of the medium being transported as low as possible.

[0029] Preferably, a classification algorithm is applied to the numerical group or to the patterns identified in the numerical group. Thus, the existing patterns can not only be identified but also evaluated with respect to their classification into different pattern classes. The classes are in particular related to the relevance of the pattern with respect to the likelihood of the presence of leakage.

[0030] Alternatively or additionally, a pattern analysis algorithm may also be applied to the pattern, and the algorithm interprets the pattern based on its characteristics in a manner similar to an image recognition method, in particular to ascertain which event is represented by the pattern and with what likelihood.

[0031] The data processing device is preferably appropriately designed to be able to execute such a classification algorithm and / or a pattern analysis algorithm.

[0032] The confirmed values can be stored in a database as a dataset. Such a dataset can be formed, in particular, by a set of numerical values that are also used to perform an evaluation for pattern recognition. Alternatively or additionally, it is preferred that the identified pattern be stored in the database as a dataset and / or that such a pattern be assigned to a pre-stored or concurrently stored dataset. Thereby, such patterns and / or the underlying values can be accessed again for later analysis. In particular, further analysis can be verified thereby.

[0033] As a result of the evaluation of the values in the formed set of numerical values, if a pattern is identified and if one or more patterns are stored in or already stored in the database, the patterns can be compared with each other. The plurality of stored patterns form, in particular in this case, a kind of lookup table. The classification algorithm applied as necessary can preferably be used to determine which of the stored patterns is compared with the newly identified pattern. If the size of the database of stored patterns, each associated with a specific event, is sufficiently large, the current event can be quickly and reliably identified in this way according to the principle of fingerprint comparison.

[0034] In addition to an overall pattern comparison, alternatively or additionally, it is possible to compare merely the individual features of a specific pattern defined as characteristic with the newly identified pattern. In this case, the characteristic pattern is used as a criterion for the presence of leakage in the measurement section of the flowing transfer object. The characteristic pattern can in particular include the averaged measured values related to the presence of leakage. Furthermore, it is also possible to use the generated data, i.e., the data modeled and / or simulated by calculation, to generate the characteristic pattern. In this case, the characteristic pattern preferably corresponds to the pattern that ideally appears at the values confirmed in the case of leakage. Depending on the degree of coincidence between the newly identified pattern and the characteristic pattern, the data processing device can be used to indicate the possibility of the presence of leakage in the relevant measurement section. In this case, if it exceeds a defined threshold value, this can in particular be used as a hard criterion for the presence of leakage, enabling appropriate measures such as manual checking or emergency stop to be initiated.

[0035] A particularly preferred configuration of the method according to the invention provides a data processing device used to apply a learning algorithm to the confirmed values or the numerical groups formed therefrom. An algorithm with learning ability, unlike when using generally used statistical methods, in some cases not only makes the method more beneficial for current applications with each iteration. Rather, the learning algorithm is trained by any application and the processing of new data. Due to the evolutionary effect, the reliability of this type of self-learning system is enhanced over time. For this reason, the error rate for the identification, especially the interpretation, of the pattern of the confirmed values decreases.

[0036] General statistical methods for data analysis related to leakage detection are usually directed at compensating for the occurring variations so that the desired information can be read from the corresponding adjusted data. In particular, when an algorithm with learning capabilities is used for data analysis, the method according to the invention can detect leakage based on the occurrence of an appropriate pattern in the confirmed values, even under conditions where known methods fail. This can be the case, for example, when the values used have significant outliers, and as a result, the approximations made during statistical processing are off. In contrast, the method according to the invention involves systematically applying empirical data to newly confirmed values. In particular, by applying an algorithm with learning capabilities, events that are not detected by each strictly applied statistical algorithm can be identified based on the patterns that appear in the values.

[0037] In a particularly preferred configuration of the method, the confirmed values or the numerical groups formed are evaluated using an artificial neural network. The data processing device is preferably of an appropriate design for this purpose.

[0038] The learning algorithm is preferably trained using the stored values before being applied to the confirmed values or numerical groups, and the stored values are related to the events that actually occurred, in particular the actual presence of leakage or events recorded in this context. Alternatively or additionally, the learning algorithm can also be trained based on simulated values. Such simulated values are preferably determined by simulating leakage on the fluid transfer object. Training by the method described above teaches the learning algorithm to associate a specific combination of values or a pattern of numerical groups with a specific event. Thus, after appropriate training, it is possible to identify, using an algorithm with learning capabilities with an appropriate configuration of the query, a pattern of unknown or new values that indicates the presence of leakage in a particular type of event, in particular in the measurement section under consideration.

[0039] From a design perspective, it is preferred that the values used in the method according to the present invention, in particular the changes in the flow rate of the medium, the pressure of the medium and / or the temperature, are non-invasively confirmed in each case. Thereby, it is possible to avoid introducing a measuring device such as a sensor into the flow transfer object and affecting the internal medium flow. This would create a risk of distorting the measurement itself and thus the subsequent data evaluation. Appropriate measurement of the data is preferably carried out by a measuring device arranged on or within the shell of the flow transfer object, such as the wall of a pipeline. In the case of the flow rate, a so-called clamp-on flow meter is particularly suitable, which can detect changes in the flow of the medium inside the flow transfer object from the outside.

[0040] Particularly preferably, the change in the flow rate is measured by an acoustic method. This includes that the flow rate or its change is confirmed based on the propagation behavior of an externally introduced acoustic signal in the flowing medium. It has been found that an ultrasonic-based method in which the injected acoustic signal has a moderately high frequency is particularly suitable. In particular, the acoustic signal is injected non-contactly, i.e., without using a mechanical transducer that externally affects the wall of the flow transfer object.

[0041] The group of numerical values inspected according to the method for identifying the pattern is formed from the values confirmed at the measurement points, but is not necessarily limited to only these values. For example, it is further possible to include or add additional data, in particular generally available data, such as data regarding the current and / or predicted weather around the flow transfer object, to the group of numerical values. This can sometimes further enhance the importance of the results of the method according to the present invention.

[0042] In one preferred configuration of the method, the values confirmed from different measurement points are transmitted to a central data processing device. The transmission in this case is preferably carried out wirelessly.

[0043] The present invention further includes a computer program product for determining the possibility of the existence of leakage on a fluid-transferring object. The computer program product is designed in particular to execute a method for leakage detection according to the present invention or to be used in a method according to the present invention. Accordingly, the computer program product includes instructions for recognizing patterns in a set of numerical values, the set of numerical values being identified on a measurement section of the fluid-transferring object and formed by values related to at least a change in the flow rate of the fluid, a change in the pressure of the fluid, and / or a change in temperature.

[0044] The present invention will be described in more detail below based on exemplary embodiments. All features described and / or illustrated in the drawings form, each independently of the combinations in the exemplary embodiments or the dependent references in the claims, aspects of the present invention.

Brief Description of the Drawings

[0045]

Figure 1

Figure 2

Figure 3

Embodiments for Carrying Out the Invention

[0046] FIG. 1 shows a typical application scenario for the method according to the present invention. In the form of a pipeline or a pipeline section for transporting a product in the form of a specific fluid medium, the fluid-transferring object 1 is laid partly above ground, partly underground, and outdoors.

[0047] The details shown represent a measurement section 2 of a rather long fluid-transferring object 1. The measurement section 2 is monitored by a method for leakage detection according to the present invention. This is achieved by ascertaining the values of various physical parameters at each of two measurement points 3.

[0048] Starting from the two illustrated measurement points 3, the measurement section 2 can also have a greater number of associated measurement points 3. It is even more certainly preferred according to the present invention, but not essential, that the measurement points 3 for the various physical quantities are arranged at the same positions along the measurement section 2 of the flow transfer object 1.

[0049] Generally, the flow transfer object 1 can be understood to mean an object in which a medium is basically intended to flow. In this regard, in the present invention, it is basically also possible to check values regarding the measurement section 2 where the medium is not flowing continuously. The determination and / or prediction of environmental parameters such as changes in environmental temperature can be of interest, for example, regarding the upcoming transport of the medium through the measurement section 2.

[0050] In principle, for all relevant physical parameters, it is preferred that the applicable values are confirmed, if possible, non-invasively, that is, without the flowing medium being affected by the components introduced into the flow transfer object 1 or the flow being disturbed in another way.

[0051] Values regarding changes in the flow rate of the medium are confirmed. This is carried out in particular by the flow meter 4. In the embodiment shown in this case, the preferred configuration of the flow meter 4 is shown as a so-called clamp-on flow meter, which is applied externally to the flow transfer object 1. Thus, the flow rate of the medium, or the change in said flow rate, can be confirmed non-invasively. In principle, it is obvious that any other type of flow measurement is also useful for confirming the value. The flow rate can be understood with reference to mass and / or volume.

[0052] In the illustrated embodiment, the flow meter 4 is based on an acoustic principle for measuring changes in the flow rate of the medium. This includes, in particular, that an acoustic signal in the ultrasonic range is introduced into the medium through the wall of the flow transfer object 1, and its propagation speed is measured in order to draw conclusions about the flow characteristics of the medium. Preferably, the acoustic signal is a propagated signal that is injected and / or read non - contact, i.e., without mechanically coupling the transducer of the flow meter 4 to the wall of the flow transfer object 1.

[0053] Furthermore, the value of the pressure change in the medium is ascertained at each measurement point 3. The pressure change is measured, in particular, by means of a suitable pressure sensor 5.

[0054] Furthermore, the temperature change is ascertained, in particular, by means of a temperature sensor 6. This is, in particular, a value about the ambient temperature or its change at the location of the measurement point 3. Alternatively or additionally, values can also be recorded at a location remote from the measurement point 3, for example between two measurement points 3 of the measurement section 2. In this context, such values can be ascertained for the air temperature, the ground temperature, the temperature of the flow transfer object 1, or the temperature of the flowing medium itself. Thus, in particular, the influence of the ambient temperature on the medium within the flow transfer object 1 along the stretch can be taken into account.

[0055] The values ascertained in particular by measurement for the above - mentioned parameters and, if necessary, further physical parameters are then transmitted to a data processing device 7 for further evaluation. The transmission is preferably carried out wirelessly. It goes without saying that it can alternatively or additionally be carried out by means of wired transmission.

[0056] The data processing device 7 can be a central data processing device 7 arranged at a location remote from the measurement section 2, as shown in the representation of FIG. 1. The data processing device 7 can be in the form of a single computer, but can also be in the form of a network of a plurality of interacting computers. Furthermore, it can also be defined to configure the data processing device 7 as a composite system having a plurality of computers operating in parallel and / or to link them hierarchically.

[0057] Data transmission from measurement point 3 to data processing device 7 can be performed, in particular, in accordance with common transmission standards such as Bluetooth or WiFi and / or via a mobile wireless network. Furthermore, there is also the possibility of satellite-based communication between measurement point 3 and / or a device for verifying the values provided at measurement point 3 and data processing device 7.

[0058] Furthermore, at measurement point 3, communication can be carried out between applicable communication devices. As an example, this enables the provision of high-performance transmission facilities at only one measurement point 3 or at least some of the measurement points 3 in order to transmit the values verified at data processing device 7. The measured values at the individual measurement points 3 of measurement section 2 are first transmitted over a relatively short distance to this type of central measurement point 3 and from there transferred to data processing device 7. A suitable design, although not associated with a specific measurement point 3, can rather be located around the relevant measurement section 2 and thus can also be realized by another relay station 10 within the range of all the communication devices of the relevant measurement points 3.

[0059] One particular configuration of the method includes at least substantially exclusive data related to the flow rate of the medium or a change in said flow rate. These data are preferably distributed by flow meter 4 and / or verified by modeling.

[0060] Particularly preferably, a network of measurement points 3 or flow meters 4 extending over at least a part of flow transfer object 1 or a part of measurement section 2 is further used. In this case, the individual measurement points 3 or flow meters 4 preferably communicate wirelessly with each other and / or with data processing device 7 using relay stations 10 arranged therebetween. Alternatively or additionally, as with other configurations of the method, reliance can be placed on standard mobile wireless technologies and / or satellite-based communication can be provided.

[0061] As will be explained in more detail below, the transmitted data is evaluated by the data processing device 7 as part of the method according to the invention and examined for the presence of a pattern indicating the presence of a leak 8 in the inspected measurement section 2. If such a leak 8 is detected, or if a sufficient likelihood of the presence of a leak 8 is confirmed, appropriate measures can be taken in a short time to provide remedies.

[0062] In the representation of FIG. 1, such a leak 8 is shown in the part of the flow transfer object 1 extending underground. For example, a transport medium that can be crude oil has entered the soil 9 in a manner that is not controlled at the location of the leak 8 and may, for example, contaminate the groundwater there. In addition to the economic significance of the loss of the transport medium, such a leak 8 can have serious ecological consequences. Considerable damage to the environment is not only in the case of a catastrophic leak 8 where the transport medium leaks out in large quantities in a short time. Rather, small leaks 8 that only result in a slow leakage of the medium over time can also already pose significant ecological risks.

[0063] FIG. 2 shows a further application scenario of the method according to the invention by way of illustration. The flow transfer object 1 is formed there by a relatively complex pipe network. The details shown are intended to represent purely symbolically, in some cases, a network branched over a wide area of a very long pipeline. Apart from the branched network of supply lines over long distances between different regions of the earth, large industrial facilities such as refineries, for example, can also be composed of a relatively complex pipeline network in some cases. In such a highly branched flow transfer object 1, various measurement sections 2 can be defined. The measurement section 2 is not necessarily defined only by the section of the flow transfer object 1 between two measurement points 3 and can in particular include further areas where three or more measurement points 3 are provided. The definition of the measurement section 2 ultimately depends on which values received from which measurement points 3, or which values confirmed for which measurement points 3, are used for evaluation by the data processing device 7.

[0064] If the complexity of the dendritic structure of the flow-transported object 1 is correspondingly high, it may be difficult to directly monitor the object only by applicable sensors. As in the case of extremely long pipelines, the complete monitoring of the system ultimately poses a problem in terms of the costs that will arise for an appropriate number of sensors. Furthermore, in the various branches of the flow-transported object 1, the partial volumes that are fluid-connected to each other in each case cause interaction and buffering effects when the transported medium propagates through the pipe network. This also hinders the evaluation of the mass flow balance.

[0065] The method according to the invention provides an advantageous effect here by detecting interference events such as the occurrence of a leak 8 in a specific measurement section 2 by identifying a pattern of confirmed values.

[0066] The influence of different temperatures on the behavior of the transport medium occurs not only when passing through various climate zones or due to different weather conditions along the pipeline. Also in the example of industrial facilities, pipelines usually extend along structures of different temperatures. For this reason, the temperature of the medium usually changes as it flows through the pipeline or pipeline network. The associated expansion or contraction of the medium significantly hinders the verification of the mass flow balance and, for example, prevents the detection of actual mass losses due to leaks 8 or unauthorized tapping (Entnahme (German), tapping (English), extraction) on the transport route.

[0067] In this regard, the method according to the invention particularly allows for the fact that various influencing factors usually affect the transport medium, in particular the dominant pressure and / or flow conditions, on different time scales. Changes in climate- or weather-related influences generally affect the medium in the pipeline, especially those extending underground, with a time delay, which is associated with a certain inertia in the reaction of the system. In contrast, for example, the desired tapping of the medium by the end consumer causes fluctuations that occur particularly in the short term and particularly locally, and these also need to be considered in an appropriate manner.

[0068] Particularly preferably, unplanned tapping of the medium in measurement section 2, for example by the end consumer, can be modeled by appropriate local consumption measurements and included in the method according to the invention. For this purpose, it is possible to define an appropriate positioning of one or more measurement points 3, in particular including the flow meter 4, in the vicinity of known tapping points.

[0069] The object of the method according to the invention is to distinguish a pattern of confirmed values occurring based on temperature and volume fluctuations in the medium due to external and internal influences from a pattern related to the actual losses of the medium from the flow transfer object 1 on the transport path. The natural influences on the medium are diverse and thus it is difficult to fully account for them with only general statistical methods.

[0070] The occurring fluctuations are mainly related to changes in the temperature of the medium transported temporally and spatially, in particular along the flow transfer object 1. This depends strongly on the ambient temperature but is also influenced by a number of other factors. The air and ground temperatures depend to different degrees on the solar radiation and accordingly affect the temperature of the medium. In contrast, rain and clouds have a short-term cooling effect. Furthermore, especially in the case of pipelines extending underground, the surface biomass can influence the temperature of the medium in the line, for example in the form of an insulating effect or shading of sunlight by the ground. Also, this factor can sometimes undergo short-term changes as a result of cultivation and harvesting in areas used for agriculture, for example.

[0071] If the flow transfer object 1 is of a sufficiently large extent or has a correspondingly complex branched structure, such as a pipeline or a pipeline network, thermodynamic changes in the flow characteristics of the transported medium generally always occur for reasons of internal effects. This reason is, for example, fluctuations or changes in the flow resistance resulting from the shape of the pipeline. In particular, if the transport medium is composed of various substances, changes in its composition may additionally occur. This can also potentially affect the flow behavior of the medium.

[0072] Large pipelines or pipe networks can further have a rather large natural volume that is initially filled, while so-called "line packing", i.e., filling the pipeline with the medium to increase the operating pressure before the medium exits again or is tapped at a specific point. The sufficiently large internal volume of the flow transfer object 1 also provides a buffering effect that can only indirectly record changes related to the volume of the medium even during operation. Thus, without further considering internal and / or external parameters, it is almost impossible to draw meaningful conclusions from the comparative measurement of the flow rate or its changes at the input and output of the measurement section 2 of the flow transfer object 1.

[0073] Extensive tests have surprisingly shown that the confirmed values can form different types of patterns. There are natural fluctuations that cannot be completely removed by general statistical methods even when environmental parameters are included, but patterns occur in the data. These are distinguished from the patterns that can be observed in the case of actual mass loss, for example, as a result of leakage points 8, line breaks, or unauthorized tapping of the medium on the transport route.

[0074] This is the starting point of the present invention in that these two types of patterns are identified and distinguished from each other. As already mentioned, the method involves using the data processing device 7 to search for patterns in these values during or after the evaluation of the confirmed values regarding changes in flow rate, pressure, and temperature, and optionally changes in further physical quantities.

[0075] The representation shown in Figure 3 shows the basic sequence for the evaluation of the confirmed values by the data processing device 7 for leakage detection. First, a numerical group 11 is formed from the confirmed values, and the data processing device 7 analyzes the presence of patterns. The numerical group 11 can be composed of all the confirmed values at the measurement points 3 of the measurement section 2 or can be a subset thereof.

[0076] When a pattern is identified in the numerical group 11, the data processing device 7 can use this pattern as a basis for determining the possibility of the presence of the leakage part 8 in the measurement section 2 of the flowing and transferring object 1. Such a pattern in the data of the numerical group 11 is identified by an appropriate algorithm of the detection routine, particularly in the same way as in the case of digital image recognition.

[0077] The data processing device 7 is preferably designed to execute a classification algorithm and apply such an algorithm to the numerical group 11. Therefore, the identified pattern is classified in terms of its type, nature, and / or characteristics.

[0078] As an alternative to or in addition to such a classification algorithm, a pattern analysis algorithm can also be applied to the numerical group 11 by the data processing device 7. Such a pattern analysis algorithm can interpret the importance of the identified pattern. Thereby, it is possible to create a statement about what real event the pattern occurring in the confirmed values represents.

[0079] In a preferred configuration, the data processing device 7 accesses a database where the identified pattern may be stored as a data set 12. The same applies to the confirmed values, that is, the numerical group 11. In particular, the confirmed pattern, the numerical group 11, and / or a specific real event that occurred, such as the presence of the leakage part 8, can be linked to each other and stored in the database as a data set 12 or as a joint data set 12.

[0080] By comparing the pattern of the analyzed numerical value group 11 with one or more patterns stored in the data set 12 of the database, in the simplest case, the identified pattern can be quickly assigned to the event group. Such a comparison by the fingerprint method is possible especially when the data processing device 7 can access the stored pattern and / or the data set 12 classified with respect to the related event, and the identified pattern can be uniquely assigned to one of these classes based on its characteristics.

[0081] Alternatively or additionally, the characteristic or idealized pattern can also be used as a criterion taken as a basis for determining the possibility of the presence of the leakage part 8 in the measurement section 2 under consideration by comparison with the pattern identified in the numerical value group 11. This type of characteristic pattern may be based on measured values from one or more measurements related to the actually occurring event, or otherwise based on simulation values.

[0082] If there is a sufficient degree of coincidence between the identified pattern and the characteristic pattern, that is, if it exceeds the defined threshold value, it may be evaluated that the criterion regarding the presence of the leakage part 8 is satisfied, and appropriate measures may be taken.

[0083] The configuration of the method according to the present invention in which the data processing device 7 applies a learning algorithm to the confirmed value or the numerical value group 11 in order to identify the pattern is particularly preferable. Alternatively or additionally, an algorithm having a learning ability can also be used to function as a classification algorithm and / or a pattern analysis algorithm. Compared with the above method of identifying and evaluating the pattern of the numerical value group 11 based on essentially firmly defined criteria, the algorithm having a learning ability has the advantage of becoming more powerful and more reliable over time as a result of appropriate training with appropriate data. Therefore, the sensitivity to errors regarding the misinterpretation (false determination) of the pattern as an indicator of the leakage part 8 and the failure to detect the existing leakage part 8 based on the confirmed value (detection leakage) will be reduced.

[0084] Such a learning algorithm is preferably trained by a data set 12 measured when an actual event, particularly an event related to or associated with the presence of the leakage portion 8, occurs. Since such data ultimately models reality in the best possible way, the trained learning algorithm is ultimately adjusted to the specific patterns that can occur in the confirmed values in individual cases under actual conditions.

[0085] Alternatively or additionally, the learning algorithm can also be trained using simulation values or model data. Thereby, components related to idealized conditions can be added to the algorithm.

[0086] For optimal detection performance regarding the identification, classification, and / or interpretation of patterns in the numerical group 11, training the algorithm with a combination of real data and simulated or ideal data can, in some cases, be particularly advantageous.

[0087] In a more preferred configuration, the data processing device 7 can use a particularly iterative method for modeling values. This includes using a method for forward modeling in particular to confirm the values that can be expected under specific operations and / or ambient conditions.

[0088] The values confirmed by such a modeling method can be employed in different ways for the method according to the present invention. By way of example, by applying such modeling methods in parallel, independent verification of the measured values and / or the patterns that appear in the measured values can be performed.

[0089] The data obtained by the method of forward modeling is also suitable for training the learning algorithm.

[0090] Preferably, the comparison of the evaluation of the real data with the modeling of the specific trends of the system enables the determination and in particular the correction of possible artifacts specific to the pattern. In this way, it is preferably possible to compensate for the drawbacks of the learning algorithms that appear in this context, and in particular, it is possible to condition on the suboptimal prioritization during the training of the algorithms. Therefore, the repeated use of this method continuously improves the reliability of pattern identification.

[0091] Furthermore, the pattern recognition-based method according to the invention can be continuously linked to a corresponding method for modeling data based on measured values. Thereby, for example, future trends can be modeled based on known or measured starting parameters, and in the results thus obtained, the risk of imminent structural failure of the flowing transfer object 1 can be evaluated by identifying the occurring patterns.

[0092] In addition to considering the data set 12 of the database by the method described above, it is also possible to include data from external sources, in particular generally available data, in different ways. These data are in particular added to the numerical group 11 and / or linked to the numerical group 11 in order to be considered for evaluation. However, this kind of external data can also be used for modeling and / or training of the learning algorithms. By way of example, the data may relate to weather, the geological composition of the ground, the agricultural use of the region in particular, etc.

[0093] The evaluation of the confirmed values, or the numerical group 11 formed therefrom, includes identifying the pattern in the numerical group 11 and, if necessary, interpreting the pattern or otherwise associating it with a specific event or event likelihood. Preferably, the data processing device 7 then generates an appropriate output 13 for communicating to the user the results of the previous analysis or the results of the method used.

[0094] Output 13 may be provided in different ways, preferably in visual, auditory and / or text form. In particular, as shown in Figure 3, output 13 may consist of a warning regarding the presence of leak 8. Further, for example, a status report may be generated.

[0095] Regarding a system that operates automatically, alternatively or additionally, improvement measures regarding output 13 may be taken immediately, for example, a warning may be distributed to maintenance and / or service personnel.

[0096] In this case, it is obvious that different outputs 13 or reactions can basically be combined with respect to the analysis results by this method.

Explanation of Signs

[0097] 1 Flow transfer object 2 Measurement section 3 Measurement point 4 Flow meter 5 Pressure sensor 6 Temperature sensor 7 Data processing device 8 Leak 9 Soil 10 Relay station 11 Numerical group 12 Dataset 13 Output

Claims

1. A method for detecting a leak in a fluid transfer object (1), i.e., a pipe or pipeline, wherein at each of a plurality of measurement points (3) on a measurement section (2) of the fluid transfer object (1), values relating to changes in the flow rate of the medium, pressure changes in the medium, and changes in temperature values are confirmed, at least two of the plurality of measurement points (3) being arranged upstream and downstream of the measurement section (2) respectively, and the confirmed values being recorded and statistically evaluated by a data processing device (7). In the method, a numerical group (11) is formed from the confirmed values, a pattern is identified in the numerical group (11) by the data processing device (7), and based on the identified pattern, the possibility of the presence of a leak portion (8) in the measurement section (2) of the fluid transfer object (1) is determined. A classification algorithm and / or a pattern analysis algorithm is applied to the identified pattern. The data processing device (7) applies a learning algorithm comprising the classification algorithm and / or the pattern analysis algorithm to the confirmed values and / or the numerical group (11). The learning algorithm is trained using stored values and simulated values before being applied to the confirmed values and / or the numerical group (11), and the stored values and simulated values are associated with the actual presence and / or simulated presence of a leak portion (8) on the fluid transfer object (1) of the medium. A method, characterized by the above.

2. The method according to claim 1, characterized in that the identified pattern is stored as a data set (12) in a database and / or assigned to a data set (12) stored in the database.

3. The method according to claim 1 or 2, characterized in that the identified pattern is compared with one or more stored patterns.

4. The method according to any one of claims 1 to 3, characterized in that a characteristic pattern functions as a criterion for the presence of a leak portion (8) in the measurement section (2) of the fluid transfer object (1).

5. The method according to any one of claims 1 to 4, characterized in that the values relating to changes in the flow rate of the medium, the values relating to pressure changes in the medium, and / or the changes in the temperature values are each confirmed non-invasively.

6. The method according to any one of claims 1 to 5, characterized in that the change in the flow rate is measured by an acoustic method.

7. The method according to claim 6, characterized in that the acoustic method is an ultrasonic-based method.

8. The method according to any one of claims 1 to 7, characterized in that data from an external source is processed by the data processing device (7).

9. The method according to claim 8, characterized in that the data from the external source is data regarding the weather around the measurement section (2) and / or the measurement point (3).

10. The method according to claim 8 or 9, characterized in that the data from the external source is linked to and / or added to the numerical group (11).

11. The method according to any one of claims 1 to 10, characterized in that the confirmed values from different measurement points (3) are transmitted to a central data processing device (7).

12. The method according to claim 11, characterized in that the confirmed values from the different measurement points (3) are wirelessly transmitted to the central data processing device (7).

13. A computer program for determining the possibility of the presence of a leakage part (8) on a medium flow transfer object (1), i.e., a pipe or pipeline, comprising instructions for recognizing a pattern of a numerical group (11), the numerical group (11) being formed by values confirmed for the measurement section (2) of the flow transfer object (1) regarding the change in the flow rate of the medium, the change in the pressure of the medium, and / or the change in temperature between the upstream and downstream of the measurement section (2), comprising instructions for applying a classification algorithm and / or a pattern analysis algorithm to the recognized pattern, comprising instructions for applying a learning algorithm comprising the classification algorithm and / or the pattern analysis algorithm to the confirmed values and / or the numerical group (11), comprising instructions for training the learning algorithm using stored values and simulated values before applying the learning algorithm to the confirmed values and / or the numerical group (11), the stored values and simulated values being associated with the actual presence and / or simulated presence of a leakage part (8) on the medium flow transfer object (1). A computer program, characterized in that...

Images