1. Field of the Invention
The present invention relates to a method and an apparatus for detecting leaks in an area of a supply network and to implementing the method in a supply network.
2. Description of the Related Art
Drinking water has by now become one of the most important goods in the 21st century. However, sometimes considerable water losses occur in water distribution networks.
The conservation of this goods item is a great challenge for detecting and locating leaks in water networks. The practice of detecting water losses via a mass balance, in which all feed-in quantities and, in particular, the water consumption quantities of all individual water consumers are measured, exceeds the limits of feasibility in terms of effort. Furthermore, legal situations hamper precisely timed consumption measurements and consumption recordings of all customer data.
In order to detect leaks in water networks, only the entire quantity of water flowing into the network is conventionally measured at a particular time, such as the period between 2 and 4 in the morning. These values produce a time series that is analyzed to determine whether there is a sudden rise in consumption and therefore a possible leak. It is not conventionally possible to balance the consumptions of the system. There are empirical formulas for the typical water consumption of particular consumers. However, these formulas are very general and cannot take into account, in particular, time-limited special effects with sufficient accuracy.
The usually very large water supply networks are commonly subdivided into water supply zones. These zones are in turn subdivided into subzones which are referred to district meter areas (DMA) because they are coined by British engineers. The DMAs are created such that they have only one inflow, the flow rate of which is measured. Irregularities in the water consumption and therefore leaks are inferred from the observation of this flow measurement. Specifically, a so-called “night flow analysis” is conventionally performed. In this case, detailed recording of the nightly inflow values into a DMA (such as every 5 seconds between 2:00 and 4:00) is used to determine a minimum inflow value (also referred to as background consumption here) which comprises the normal nightly (minimum) consumption and existing (in particular also small) leaks.
A time series is created over days and weeks based on these minimum inflow values into a DMA during low-consumption night-time hours, such as between 2 and 4 a.m., in which case only one value per night is then provided. The, in particular sudden, rise in these minimum consumption values, which can be detected by a threshold value being exceeded, for example, may be caused by a new leak.
In order to locate a leak, a step test is usually performed. For this purpose, small regions are gradually separated from the DMA at low-consumption times and the change in consumption is observed. Regions that result in a severe inexplicable decrease in consumption are then examined further for leaks.
Alternatively, noise meters can be used to listen to the water system in situ for leaks and the leak point can be calculated by considering the noise correlation.
Both of these conventional methods are not suitable for permanent monitoring. Step tests are associated with a large amount of effort because the affected households must be informed of the disconnection and a backup supply must be ensured. Noise measurement requires a large amount of expenditure because the measurements can be performed only by specialists in situ. In addition, these investigations are always only locally possible. In addition, both conventional methods can be used only in low-consumption times so that the measurements are not overly disrupted by consumption fluctuations.
It is therefore an object of the present invention to provide a method and an apparatus for detecting leaks in an area of a supply network, where leaks in the supply network are detected with a high degree of probability.
This and other objects and advantages are achieved in accordance with the invention via a computer-aided method for detecting leaks in an area (DMA) of a supply network comprising:
In accordance with the method of the invention, a distinction is made between flow sensors at the inflows and outflows and sensors within the area. Whereas the measured values from the inflow and outflow sensors describe the water consumption of the area and are therefore used in the hydraulic model, the measured values from the internal sensors are used to determine differences between the model and reality and are therefore used to detect leaks. The consideration of a plurality of measuring intervals compensates for a random atypical consumer behavior so that the latter is not overrated. The method can be easily automated (by using computers and corresponding software) and can be operated together with other methods which consider each area separately (such as camera monitoring or pressure sensors). Dynamic, non-periodic special effects (such as higher water consumption during sports events on the television) can also be taken into account, as a result of which the false alarm rate also continues to be reduced. The method can also be used for areas in which the night flow analysis cannot be used because high consumptions also occur at night, such as in megacities.
In a first advantageous embodiment of the invention the areas (DMA) may also be virtual district meter areas. Virtual zones or virtual district meter areas (DMA) are subareas of a network, the inflows and outflows of which are measured via flowmeters, in which case there is no requirement for the areas to be disjoint. The time series for all areas are gradually evaluated and leaks in the areas are detected. The location of the leak is then narrowed down using the leak information for the individual areas. An item of leak information is an item of information relating to whether or not a leak has been detected in the area. Virtual district meter areas (virtual DMAs) differ from conventional areas (DMAs) as follows. When subdividing zones into DMAs, an attempt was always conventionally made to form these DMAs such that only one inflow or inflow pipe resulted and can be monitored using a single sensor. In the supply zones, additional flow sensors are installed at selected points such that parts of the network result, the inflows and outflows of which can be measured. These parts should have a common element. The parts are intended to be superimposed and to have common flowmeters. Such parts are referred to as virtual zones or virtual DMAs.
In another advantageous embodiment of the invention the measuring period is, for example, from 2:00 to 4:00, from 0:00 to 24:00 and/or from 6:00 to 18:00. During a night flow analysis, all flowmeters or sensors are used to measure the flow rate within the scope of such an analysis at night, such as between 2:00 and 4:00, i.e., during times which usually have a low consumption. Instead of an analysis at night, other intervals of time, such as 24 hours or a plurality of measuring periods during a day, may be considered for integrated flow analyses.
In another advantageous embodiment of the invention step e) of determining the random consumptions of the consumers connected in the areas (DMA) is performed using the sequence of the following method steps:
This algorithm is used to describe the consumption distribution. A hydraulic simulation of the network section is performed using this consumption. For a hydraulic simulation, boundary conditions must be set such that the physical model is meaningfully described: if the zone has only one inflow/outflow, a constant pressure is set at this point, and, if the zone has a plurality of inflows and outflows, a constant pressure is set at one of them and the measured inflows and outflows are set at the others. The mass balancing in the model is therefore correct.
In yet another advantageous embodiment of the invention step e) of determining the random consumptions of the consumers connected in the areas (DMA) is performed using the sequence of the following method steps:
Alternatively, the random flows may also be distributed to the consumers using the algorithm illustrated, if it is known that the consumption profiles of the consumers in an area are the same. This is equivalent to the first embodiment of the method, but is distinguished by a lower delay time since fewer random numbers are picked.
In a further advantageous embodiment of the invention steps e) and f) are repeated for a fixed interval of time and the calculation results are averaged before the comparison according to step g) occurs. A fluctuation range for the calculation results is obtained as a result of the repetition. The influence of outliers is reduced in this case and the normal behavior of the area emerges.
In another advantageous embodiment of the invention, the method is implemented for an infrastructure network for transporting a fluid. The measured values for fluids can be easily and accurately determined by means of corresponding sensors (such as pressure or flow sensors) and can therefore be used for reliable predictions.
In still a further advantageous embodiment of the invention, the infrastructure network is a water supply or a gas supply or a district heating network. The disclosed embodiments of the invention can be used for all infrastructure networks in which fluids are transported and consumed. Examples of such infrastructure networks are gas supply and district heating networks.
It is also an object of the invention to provide a computer-aided apparatus for implementing the method for detecting leaks in an area of a supply network, where the apparatus comprises:
In another advantageous embodiment of the invention, output means are provided for presenting the comparison of the measured values determined by the Monte Carlo simulation with the measured values provided by the sensors and/or for presenting indicators of a leak. A graphical representation makes it possible to visually compare the results and to easily detect discrepancies as indicators of leaks.
It is also an object to provide a computer-aided simulation-based assistance system for detecting leaks in an area of a supply network, where measured values are recorded by sensors installed in a stationary manner for measuring a minimum inflow of the area within a stipulated measuring period and are transmitted to an evaluation device via remote data transmission,
The evaluation device includes:
Other objects and features of the present invention will become apparent from the following detailed description considered in conjunction with the accompanying drawings. It is to be understood, however, that the drawings are designed solely for purposes of illustration and not as a definition of the limits of the invention, for which reference should be made to the appended claims. It should be further understood that the drawings are not necessarily drawn to scale and that, unless otherwise indicated, they are merely intended to conceptually illustrate the structures and procedures described herein.
One exemplary embodiment of the invention is illustrated in the drawings and is explained below, in which
The reduction of water losses is a great challenge for detecting and locating leaks in water networks. In subareas of a network, water losses can be detected via flow rate measurements at all inflows and outflows of the area and formation of a mass balance.
Such measurements exceed the limits of feasibility both in terms of effort and in terms of the legal situation (precisely timed consumption measurement and recording of all customers may result in problems in terms of data protection law). Nevertheless, new more favorable measuring devices make it possible to monitor physically relevant values, such as pressure or flow rate.
The present invention presents a stochastic model for consumers in a supply network, which model makes it possible to set up a network-wide mass balance using a hydraulic simulation to detect both new and already existing leaks.
The usually very large water supply networks are subdivided into water supply zones (areas). These areas may in turn be subdivided into subzones that are referred to as district meter areas (DMA) because they are coined by British engineers. The DMAs are created such that they each have only one inflow, the flow rate of which is measured. Alternatively, instead of physical DMAs, it is also possible to use virtual zones which may have a plurality of inflows and outflows. Irregularities in the water consumption and therefore leaks are inferred from the observation of the flow measurement. Specifically, a night flow analysis is performed. A DMA is performed using the minimum inflow values during low-consumption night-time hours, such as between 02:00 and 04:00. With one value per night, a time series is created over days and weeks. The (sudden) rise in these minimum consumption values, which can be detected by a threshold value being exceeded, for example, may be caused by a new leak.
Special events, such as the use of lawn sprinklers, sports events or customs during the night measurement therefore cause an increase in the minimum inflow into different DMAs. An alarm is therefore triggered for all DMAs. Such a possible false alarm hitherto had to be detected by the user of the system and eliminated via further analysis. The present invention allows an automatic method for detecting network-wide events for reducing false alarms during leak analysis.
The procedure of creating virtual zones (virtual areas) presents a universal method of subdividing supply networks such that subareas, such as one or more line sections, can be repeatedly monitored with respect to leak detection. The monitoring of each virtual zone functions according to the same principle and can accordingly be repeatedly used in a technical solution. The subdivision of a network into virtual zones provides the advantage that, apart from the installation of flowmeters, there is no need to make any change to the existing network. Another advantage is that the leak detection process can run in an automated manner without disrupting the operation of the supply network or carrying out laborious, cost-intensive measurements in situ.
In particular, the evaluation device AE comprises means M1 for determining the water consumption for each area DMA within a stipulated measuring period via the installed sensors, means M2 for mapping the topology of the supply network in a hydraulic simulator and creating a hydraulic simulation model for each area DMA, means M3 for determining the consumption profiles of the consumers connected in the areas DMA, means M4 for determining the flow behavior in the supply network within the stipulated measuring period via Monte Carlo simulation, and means M5 for comparing the measured values determined by the Monte Carlo simulation with the measured values provided by the sensors to detect possible leaks in an area DMA in the event of discrepancies. In this case, the topology of the supply network VN is simulated in a hydraulic simulator. The pipes are parameterized using the known physical values. In contrast, the consumers at the nodes of the network are unknown. A stochastic equivalent model is set up for this purpose.
The computer-aided assistance system AS can be produced using commercially available means. For example, the corresponding sensors SE1-SE3 (for example flowmeters) are available in corresponding specialist shops, and the means for calculating, determining and comparing can be implemented on personal computers C and corresponding software (such as table calculation, mathematical, or simulation programs). In order to generate automatic countermeasures or to display measures for the operating personnel, the assistance system AS may be based, for example, on model-based techniques (for example, CBR, i.e., Case Based Reasoning). Discrepancies between the actual and expected values (simulation results) can be displayed on an output unit M (e.g., a screen) of a computer C. The computer C also comprises storage media, such as a database DB for storing or buffering the measured values from the sensors SE1-SE3 arriving via the remote data transmission line DFÜ.
Additional sensors that do not necessarily define further zones or (virtual) DMAs are placed in a zone/area (for example a virtual DMA) with a known inflow. As is customary in leak detection, the consumption in a low-consumption period of time, such as from 2 to 4 a.m., is again considered (night flow analysis). A hydraulic model is now set up for the zone/area (for example a virtual DMA) based the measured inflow into the zone (DMA). For this purpose, the topology of the network is simulated in a hydraulic simulator. The pipes are parameterized using the known values. In contrast, the consumers at the nodes of the network are unknown. A stochastic equivalent model is set up for this as follows:
The water consumption in the zone is randomly distributed to all consumers for each measuring period. The distribution of the consumption is described in algorithm 1. A hydraulic simulation of the network section is performed using this consumption. For a hydraulic simulation, boundary conditions must be set such that the physical model is meaningfully described: if the zone has only one inflow/outflow, a constant pressure is set at this point, and, if the zone has a plurality of inflows and outflows, a constant pressure is set at one of them and the measured inflows and outflows are set at the others. The mass balancing is therefore correct.
The Monte Carlo simulation using different events of the random distribution of the measured values is used to calculate the calculated sensor values from the internal sensors of the zones. The discrepancy between the measured values and the calculated values indicates possible leaks.
In step S1, the supply network is divided into areas (DMA) each with a known inflow. This can be effected in a computer-aided manner based on models of the network or based on empirical values.
In step S2, flowmeters are installed in a stationary manner at inflows and outflows of an area (DMA). When installing the sensors, it is possible to access sensors that have already been installed, or new sensors are installed depending on the intersection of the areas (DMA). The measured values from the sensors can be reported to a control room for further processing, such as via remote data transmission, radio or satellite link.
In step S3, the water consumption for each area (DMA) is determined within a stipulated measuring period via the flowmeters. This is also advantageously effected in a computer-aided manner.
In step S4, the topology of the supply network is mapped in a hydraulic simulator and a hydraulic simulation model is created for each area (DMA). This is also advantageously effected in an automatic and computer-aided manner.
In step S5, the random consumptions of the consumers connected in the areas (DMA) are determined. This is advantageously effected via software programs.
In step S6, the flow behavior in the supply network is determined within the stipulated measuring period via Monte Carlo simulation. The Monte Carlo simulation is effected by means of a simulation program.
In step S7, it is determined whether there is a leak by comparing the measured values determined by the Monte Carlo simulation with measured values provided by the sensors. The comparison is effected in a computer-aided manner and discrepancies that may be indicators of a leak are advantageously graphically displayed.
Countermeasures (such as closing intake valves, or activating diversions) can be automatically initiated on the basis of the discrepancies, or possible countermeasures can be proposed to the operating personnel of a control room.
1) Classifying the consumers: different consumers have a different consumption profile. These profiles indicate how the daily total consumption of a consumer can be mapped to the day. Residential buildings therefore have a different consumption behavior to office buildings, schools or SMEs. Consumers who cannot be classified must be measured exactly and are disregarded in the further description.
2) A theoretical total consumption in the zone can be determined based on the average daily consumption of each consumer which is obtained, for example, using the year-end settlement. For this purpose, a theoretical consumption is calculated for all consumers during the nighttime measuring period based on their average daily consumption and their consumption profile. A theoretical total consumption is calculated therefrom.
3) For initialization, the consumption of all consumers is set to 0 for the entire observation period. In addition, a small quantity of water Q which is subsequently distributed (for example 31) is stipulated; this quantity should be considerably smaller than the minimum water consumption in the zone (DMA).
4) The total consumption of the zone is measured for each period of time of the measuring period (generally 1-3 minutes).
5) Consumers are now randomly selected: the probability of selecting a consumer is his proportion of the total consumption determined in 2). The consumption of the consumer is increased by Q for this period of time.
6) As long as the total water consumption distributed for the observation period is smaller than that determined in 4), go to 5).
7) Repeat steps from 4) for the next measuring period.
The flow behavior for the period of the night flow analysis is now simulated using the random consumption constructed above. A Monte Carlo simulation (as described, for example, in Kurt Binder [et al.], Monte Carlo methods in statistical physics, Springer, Berlin 1979) is performed for this purpose. For this purpose, the simulation is performed for a large selection of random consumptions constructed as described above, and the behavior in the zone is inferred from the large number of simulation values. For this purpose, the measured values from sensors inside the zone under consideration are depicted against the calculated values, as can be seen in the following image. The deviation of the best-fit line through the point cloud from identity is an indicator that the model does not fit the measurements, which indicates a leak (see
Alternatively, the random flows can also be distributed to the consumers using the following algorithm 2. This is equivalent to the method described above but is distinguished by a shorter delay time since fewer random numbers are picked.
Given: quantum quantity Q
Initialize: R=measured total consumption (as an integer multiple of the quantum quantity Q)
G=sum of the average consumptions V(n) of all nodes
This algorithm can be used in the case of identical consumption profiles of all consumers. The algorithm allows efficient calculation.
Method, apparatus and assistance system for detecting leaks in an area of a supply network are thus provided, in which case leaks are detected by comparing actual measured values provided by sensors with measured values determined via a Monte Carlo simulation. The method in accordance with disclosed embodiments of the invention for detecting leaks determines, in particular, irregularities, which can be attributed to leaks, for example, based on a hydraulic analysis. Already existing leaks can be detected as a result. In addition, the method in accordance with disclosed embodiments can also be applied to sensors temporarily fitted in the network, which provides the network operator with additional freedom when searching for leaks. The method in accordance with disclosed embodiments can also be used for other supply networks and infrastructures.
Thus, while there have shown, described and pointed out fundamental novel features of the invention as applied to a preferred embodiment thereof, it will be understood that various omissions and substitutions and changes in the form and details of the devices illustrated, and in their operation, may be made by those skilled in the art without departing from the spirit of the invention. For example, it is expressly intended that all combinations of those elements and/or method steps which perform substantially the same function in substantially the same way to achieve the same results are within the scope of the invention. Moreover, it should be recognized that structures and/or elements and/or method steps shown and/or described in connection with any disclosed form or embodiment of the invention may be incorporated in any other disclosed or described or suggested form or embodiment as a general matter of design choice. It is the intention, therefore, to be limited only as indicated by the scope of the claims appended hereto.
Number | Date | Country | Kind |
---|---|---|---|
DE 102011078240.0 | Jun 2011 | DE | national |
This is a U.S. national stage of application No. PCT/EP2012/060963 filed 11 Jun. 2012. Priority is claimed on German Application No. 10 2011 078 240.0 filed 28 Jun. 2011, the content of which is incorporated herein by reference in its entirety.
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/EP2012/060963 | 6/13/2012 | WO | 00 | 12/23/2013 |