The purpose of the present invention relates to the monitoring activity performed on water distribution network pipelines to detect and prevent real leaks. The present invention makes use of different techniques, in a single system, such as the balance of the volumetric flow and the usage of energy-related indices formulated on mathematical and engineering approaches.

The above-mentioned techniques are implemented integrating different technologies in a single system, such as the measurement of pipe's vibrational profiles and altitude, wireless communication technology, and an innovative postprocessing algorithm.

BACKGROUND OF THE INVENTION

Leaks and ruptures in water supply pipelines, blockages and overflow events in sewer collectors cost millions of dollars a year, and monitoring and repairing this underground infrastructure presents a severe challenge. Available statistics show that the amount of real water losses may exceed 40% of the input volume. The inlet and lost water in distribution network is yet drinkable, i.e., it is equipped with the hygienic and organoleptic requirements suitable for human consumption.

Accordingly, the main objective of the present invention is a system not only for the automatic early detection of leaks in service pipes and following breaks repair, but also for evaluation and checking of water network conditions. This invention also allows scheduling appropriate works of reconstruction and renovation.

In order to reduce water waste and limit leaks in a generic water network pipeline, the water balance describes the flow of water in and out of a system, and it can be used to help manage water supply and predict where there may be water shortages or leaks.

The water balance consists of several components, one of them is the non-revenue water, that is water that has been produced and is lost before it reaches the customer. Losses can be:

- Apparent losses: are the non-physical losses that occur in utility operations due to customer meter inaccuracies, systematic data handling errors in customer billing systems and unauthorized consumption. In other words, this is water that is consumed but is not properly measured, accounted or paid for;

- Real losses: are the physical losses of water from the distribution system, including leakage and storage overflows. These losses inflate the water utility's production costs and stress water resources since they represent water that is extracted and treated, yet never reaches final users.

Real losses are given by both background leaks (very small leaks occurring at storage tanks or pipe joints and fittings) and burst leaks (resulting from pipe holes and damages) as shown in FIG. 1 . While the former ones are associated with the normal operational conditions and cannot be reduced under the limit commonly known as Unavoidable Annual Real Losses, the latter ones are considered potentially recoverable losses. Hence adopting proper policies for managing burst leaks appears essential. If a water distribution system has a flow rate very variable, analyzing the Minimum Night Flow (MNF) it is possible determine if there is a leak or not. Controlling leakage effectively relies upon a proactive leakage management program that includes a means to identify hidden leaks, optimize repair functions, manage excessive water pressure levels, and upgrade piping infrastructure before its useful life ends.

As for Active Leakage Control, several strategies and technologies have been proposed for achieving leak detection and localization. Most of them are based on measurements of vibro-acoustic phenomena as well. Several studies deal with the detection of low-frequency noise and/or vibrations by means of hydrophones and/or accelerometers, or with the acquisition of high-frequency acoustic signals by means of acoustic emission sensors. In acoustic methods, Piezo-electric vibration detectors, or accelerometers, are placed at one or more locations in the water system on pipes, on the ground or on walls.

Relevant patents and patent applications regarding the present technical field are the following: U.S. Pat. No. 7,360,413 discloses a wireless water flow monitoring and leak detection system that includes a base station, a memory, and a central processing unit configured to control the operation of the system and to analyze stored data. The wireless flow sensor nodes can periodically read and store a data point corresponding to either a flow condition or a no flow condition occurring at the water fixture.

U.S. Pat. No. 6,789,41 1 discloses an apparatus for detecting leaks in an underground water pipe. A hydrophone monitors water flowing along the pipe and an alarm signal is generated when a flow parameter is above a maximum value or below a minimum value. A radio transmitter transmits the alarm signal to a remote receiver.

U.S. Pat. No. 8,820,143 discloses a leak detection system including a sensor disposed within the at last one tube to detect a pressure gradient or fluid movement within the tube. Such pressure gradient or fluid movement indicates a leak in the pipe adjacent to the tube location.

U.S. Pat. No. 2014/00228459 discloses a method for leak detection and localization in at least portion of fluid distribution system. The system includes a position locator and a vibration sensor which generates a signal indicative of vibrations detected at the location. A processor stores and processes the location of the device and the value of a calculated parameter depending by average power of the vibration signal over a time period.

Further relevant publications are the following:

Stoianov, et al. "PIPENET: A wireless sensor network for pipeline monitoring." 2007 6th International Symposium on Information Processing in Sensor Networks. IEEE, 2007.

Evans, Robert P., Jonathan D. Blotter, and Alan G. Stephens. "Flow rate measurements using flow-induced pipe vibration." Journal of fluids engineering 126.2 (2004): 280-285.

Dinardo, Fabbiano, and Vacca. "Fluid flow rate estimation using acceleration sensors." Sensing Technology (ICST), 2013 Seventh International Conference on. IEEE, 2013. Cross, Hardy. "Analysis of flow in networks of conduits or conductors." University of Illinois. Engineering Experiment Station. Bulletin; no. 286 (1936).

SUMMARY OF THE INVENTION

The present invention consists of an automated system based on wireless sensor networks which aims to detect, locate and quantify bursts, leaks and other anomalies in water distribution pipelines.

This invention provides a water flow monitoring system for:

- determining when a leak occurs and upon the determining of such a leak, take some action to stop further flow;

- implementing a pro-active approach to optimize management of the maintenance work for water distribution pipeline and schedule works of rehabilitation and replacement of pipes.

Operational pipelines are subject to complex, highly non-linear temporal and spatial processes that make it difficult to differentiate between faults and stochastic system behaviors. This makes detecting failures a challenging task, leading towards a solution based on integrating remotely captured data from several sources, basing on vibration signals. Starting by these signals through a specific algorithm, illustrated in the document "Fluid flow rate estimation using acceleration sensors." by Dinardo, Fabbiano, and Vacca (Journal of fluids engineering 126.2 (2004): 280-285), the proposed system calculates flow rates for each source. Vibration signals are used for detecting small leaks that might be precursors for catastrophic bursts, while the analysis of flow rates in network of pipes enables prompt detection and localization of larger leaks and malfunctioning.

Important advantages of the present invention are in terms of automation, communication and integration with information and management systems, in terms of structural simplicity (i.e., installation and use), energy consumption (i.e., maintenance), flexibility and modularity (i.e., optimization of the cost/benefits ratio).

In further detail, the main advantages of the system according to the proposed invention are, among others, the following:

- Multisensor-to-multipoint-to-point survey to increase the amount of data acquired from the deployed field instrumentation; - Flexibility of use in more or less vast metering areas, through implementation of elaboration algorithms and GIS applications; in order to minimize costs/benefits trade-off;

- Further facilitation in processes of installation through non-intrusive measuring technique and transducers with automated calibration methods;

- Further decrease of maintenance processes through ambient energy supply for local collector nodes;

- Possibility that the devices according to the present invention be fixed or removable according to different measuring scenarios.

Further features and advantages of the present invention will be apparent in the following description of a non-limitative embodiment with reference to the figures in the accompanying drawings, which are diagrammatic and show functional blocks which are adapted to be implemented with a hardware structure according to different circuitry solutions in practice or with a software structure, for example coded into firmware and executed by a suitable digital signal processor (DSP). BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a schematic illustration of a typical water flow composition for leak detection using the Minimum Night Flow analysis;

Figure 2 is a schematic illustration of the connection between wireless sensor for measurement automatization, communication and analysis to monitor the pipe status and detect and localize leaks, according to the present invention;

Figure 3 is a schematic illustration of data collecting and transmission for leak metering area between different devices, in one embodiment of the present invention;

Figure 4 is a schematic illustration of an example of non-intrusive flow measurement technique according to the present invention based on the acquisition of the transverse vibrations induced by the fluid motion on the pipe;

Figure 5 is a flow chart showing the algorithm defining flow directions and calculating estimated pressures for metering areas according to the present invention;

Figure 6 is a flow chart showing the general monitoring system algorithm according to the present invention;

Figure 7 is a table showing conditions to discriminate type and magnitude of alarm; Figure 8 is a schematic illustration of an embodiment of the invention showing an example of water distribution network and water management remote control center according to the present invention;

Figure 9 is a first schematic illustration of an example of sensor node employed in one embodiment of the present invention and

Figure 10 is a second schematic illustration of an example of sensor node employed in one embodiment of the present invention.

DETAILED DESCRIPTION OF THE INVENTION

The present invention is intended to detect, localize and quantify bursts, leaks and other anomalies in water distribution pipelines, and comprises the following features: A non-intrusive measuring technique based on evaluation of both structural vibrations and altimetry;

A Wireless Sensor Network (WSN) for automation and communication of measured values;

A software program including a dashboard platform comprising an elaboration module adapted to implement algorithms to define, manage and process appropriate functional parameters, said appropriate parameters being, for instance, filters chosen as required by the application context.

The WSN technology allows multi-sensor measurements and automated calibration in order to facilitate correlation between input data and measured data obtained by sensor nodes 10. This process guarantees that decisional parameters are normalized before extrapolation.

FIG 2 illustrates an example of WSN according to the present invention; the employment of a WSN increases system efficacy (productive, economical, and of maintenance) compared to traditional systems.

The preferred topology of the WSN according to the present invention is a star network (multipoint-to-point), consisting of a central node, to which all other nodes are connected.

FIG 3 illustrates another example of an embodiment of monitoring system according to the present invention; this network consists of:

Sensor Nodes 10 being the measuring points; they should be installed underground in contact with pipe in correspondence of the access points to the water network. Sensor nodes 10 are connected to a Local Collector Node 1 1 , they acquire signals and transmit them to an associated collector node.

Local collector nodes 1 1 must be installed externally to the water network because they provide bidirectional wireless propagation between Sensor Nodes 10 and a Central Server 12.

Central cloud connected Server 12 receives information from Local Collector Nodes, it stores and elaborates received data in order to monitor system and individuate critical areas.

FIG 4 shows an example of flow measurement method according to the present invention, that is based on the acquisition of the transverse vibrations induced by the fluid motion on a pipe. For a given pipe (in terms of width, diameter, material, and structural constraints), the first harmonic amplitude of the vibration signal transmitted from the flow to the pipe walls is linearly proportional to the flow rate. Therefore, by means of calibration coefficients which depend on the physical properties of the pipe, this measuring technique makes use of existing correlation between pipe transverse vibrations and flow rates.. It is demonstrated that such vibration components arise from the transfer of the fluid momentum variation due to the fluctuations of its velocity because of the turbulence. Depending on the elasticity properties of the pipe inside which the fluid runs, the vibrations can be more or less amplified.

The measuring device automatically performs measurements, preferably during night time, to reduce both the presence of possible perturbations affecting the vibration signals and the incidence of non-zero flow rate conditions induced by consumers. The proposed measurement technique is repeatable, accurate, reliable and non-intrusive. It makes use, effectively, of parameterized transverse vibrations based on physical characteristics of measured pipes and it is adaptable to any occurrence and independent of the environmental conditions and position in the leak metering area wherein measure takes place.

In order to monitor pipeline status, initial assumptions concerning the water distribution network and load balancing are needed. For this reason, the present invention uses the following data as input for each measuring point:

Input data during installation: -Georeferencing tag;

-Duct nominal diameter;

-Duct material.

Field input data measured by Sensor Nodes 10:

-Duct radial vibrations (VI BM (t));

-Punctual Altimetry.

Reference input data for the Server, the monitoring and control center:

-Piezometric head at the origin of distribution network;

-Water network topology;

-Minimum Night Flow Threshold (MN FTH) for each area: MNF _TH = AQ _ERR + CNU + BL where

- AQERR: measuring error [l/s];

- CNU: Consumer Night Use [l/s];

- BL: Background Leakage [l/s].

The proposed algorithm uses those input data to elaborate three functional parameters, which evaluation allows the monitoring system to discriminate and generate warnings and to localize anomalies.

The three parameters are: Water Flow Balancing (BQ), Duct Damage Index (DDI) and, Duct Wearing Index (DWI).

Water Flow Balancing (BQ)

In the proposed system, the water flow balancing is calculated from radial vibrational measurements that are punctually measured on the pipe external surface as explained in documents "Flow rate measurements using flow-induced pipe vibration." by Evans, Robert P., Jonathan D. Blotter, and Alan G. Stephens, in the Journal of fluids engineering 126.2 (2004): 280-285, and "Fluid flow rate estimation using acceleration sensors." by Dinardo, Fabbiano, and Vacca, in the Sensing Technology (ICST), 2013 Seventh International Conference on. IEEE, 2013.

The BQ parameter concerns night measurements in order to minimize noise due to consumers and non-steady-state network operation. The BQ parameter is compared to MNFTH value so that possible losses can be quantified.

BQ = WATER FLOW BALANCING =∑ _n Q _n ≤ MNF _TH

Where: Q _N flow rates calculated by VI BM (t);

BQ-MNF _TH

BQ% = percentage variation of BQ compared to MNFTH = ^■ 100;

MNF _TH

THBQ% = Tolerance threshold of BQ% Qn values are measured unsigned by sensors positioned on each Sensor Node 10; in this way, it is not possible to know the water flow direction. Evaluation of Water Flow Balancing needs both amplitude and direction of flow rates. For this reason, this invention presents an algorithm, illustrated in FIG. 5, to establish flow direction for each pipe with sensor, calculate piezometric head and, consequently, estimate water pressure for each sensor position, and evaluate potential leakage in each monitored area.

Starting from the initial point of the water distribution network, for convenience called ODU (acronym for the Italian term Origine Distribuzione Urbana' that means origin of the urban distribution), Sensor Nodes 10 are positioned so as to create closed areas. They have to be located in every intersection between own area and outside; so same device can belong to several areas, adjacent to each other's. Levels of those areas are numbered according to their distance from the ODU.

In every area the followin

= 0

Where:

∑ q _in sum of all flow rates incoming in the area;

∑ q _out sum of all flow rates outgoing by the area;

Qarea - consists of flows dispersed within the area in the form of utilities or leaks.

The last variable should be equal to zero or negligible ideally, because in every area the amount of incoming water should be the same as the outgoing one. For this reason, Q _area is a threshold, it defines an acceptable bound of dispersed water in the area; exceeded this threshold, the leak detection system is alarmed and a maintenance work will be needed on that area. Its value must be appropriately evaluated according to information provided by the owner of water system. In order to define sign of Q _n values, knowing areas' topology and absolute value of flow rate for each sensor, this invention uses an algorithm for pre-localization of leaks. The global monitored area is divided into many smaller areas. The level of these sub-areas is defined by the distance from the ODU starting from level 0 if areas are directly connected to the ODU. Therefore, the general N-th area level is directly connected to the previous ones, i.e. the areas of (N-1 )-th level. So, their incoming flow is caused by pipes connected to previous areas (smaller than N-th level); outgoing flow comes from pipes connected to successive areas (higher than N-th level).

To define the direction of flow between areas with the same distance from the ODU, this algorithm uses flow balancing formulas and calculations to verify water distribution network, particularly, the Hardy Cross method. This method is an iterative method for determining the flow in pipe network systems where the inputs and outputs are known, but the flow inside the network is unknown. The Hardy Cross method is an application of continuity of flow (or mass) and continuity of potential (or energy) to iteratively solve for flows in a pipe network.

For a control volume with multiple inlets and outlets, the principle of Conservation of Mass requires that the sum of the mass flow rates into the control volume equal the sum of the mass flow rates out of the control volume.

The principle of Conservation of Energy implies that in each closed circuit the sum of all changes in head is zero.

This principle can be applied to every sub-areas and is proportional to

s which is the sign positive or negative according to the direction of flow in area;

r which is the loss of head in the pipe for unit quantity of flow; this quantity depends on the length and diameter of pipe and on its roughness;

q which is the flow rate along the incoming branches of the area;

and the flow exponent is n.

Solving complex pipe systems for water distribution is extremely difficult due to the nonlinear relationship between head loss and flow; Hardy Cross developed two methods for solving flow networks. Each method starts by maintaining either continuity of flow or potential, and then iteratively solves for the other. The proposed algorithm uses the method of balancing piezometric heads; it is directly applicable where the quantities flowing at inlets and outlets are known. In this one the flows in the pipes or conductors of the network satisfy the condition that the total flow into and out of each junction is zero, and these flows are successively corrected to satisfy the condition of zero total change of piezometric head around each circuit. The Hardy Cross method - explained in document "Analysis of flow in networks of conduits or conductors." by Cross, Hardy (University of Illinois. Engineering Experiment Station. Bulletin; no. 286 - 1936) - is adapted to, in a network of pipes which consists of connected and closed areas, verify a pipe network for distribution by means iterative steps:

1 . Assume any distribution of flow (its value is measured by sensors).

2. Compute in each pipe the loss of head /i _£ . With due attention to sign, compute the total head loss around each elementary closed circuit∑ h _t =

∑s _i r _i \q _t \ ⁿ .

3. Compute also in each such closed circuit the sum of the quantities R = n r _t |(? _ί ^_1 without reference to sign.

4. Set up in each circuit a counterbalancing flow to balance the head in that circuit (to make∑s _£ r _£ | qr _£ | ⁿ = 0) equal to

this is only an approximation due to the terms that are ignored from the Taylor expansion.

5. Compute the revised flows \q- 1 = | | qr _£ | + s _£ Aq\ and repeat the procedure. Continue to any desired precision.

Of course, if Aq is relatively large compared with q _t and n> 1, the approximation is not very good, but this is less important than it might at first seem, because in any case we must correct for the unbalanced head produced in one circuit by corrections in the adjacent circuits, which in general requires a recomputation of all circuits. The convergence is, for practical purposes, sufficiently rapid. Now, the computation of piezometric heads for each Sensor Node 10 position is immediate knowing its head loss; starting from a completely known node, such as the ODU, heads for directly connected nodes will be calculated and so on moving away.

Duct Damage Index (DDI)

In water networks, pressurized pipes are characterized by radial vibration signals, their overall energy content characterizes the operating conditions of the pipes themselves. In order to define an index monitoring those operating conditions (e.g., loose joints, cracks, openings) the overall energy content of vibrational profiles can be expressed by its root mean square (rms) value: where:

VIBM-i (t) and VI BM2(t) are vibration signals acquired with a duration of TMEAS and sampling frequencies Fsi and Fs2, respectively;

VIBc(t) is a characteristic vibration signal acquired with a duration of TMEAS and sampling frequency Fs∑.

The Duct Damage Index (DDI) is proportional to the deviation between RMS of characteristic signals and RSM of acquired signals, as shown in the following equations.

VIB _{M1 RMS} VI B _{C RMS}

DDL = DUCT DAMAGE INDEX1 100

VIBcrms

VIB _M2 _rms ^~ VIBc_rms

DDI, = DUCT DAMAGE INDEX! = 100

VIBc_rms

where VIBc_rms is obtained through characterization and modeling activities on ducts.

The choice of values for Fs Fs2, and TMEAS depends on various causes, such as: -Minimizing the sampling interval and the data transmission in order to reduce energy consumption;

-Response time of the accelerometer; -Bandwidth of the transverse vibration (related to Fsi and Fs2) and the lowest component of the frequency spectrum of the transverse vibration (related to TMEAS) -Spatial resolution for localization of anomalies; the proposed system uses specific algorithms for performing leak detection and localization via cross correlation of vibration signals.

The percentage deviations of previous two indexes are:

I VIBMl rms ^~ VIBc_rms I

100

VIBc_rms

V1B} 2 rms ^~ VIB _{C rms}

DDI ₂ % 100

VIBc_rms

Those values have to be compared to a known threshold THDDI%, i.e., the tolerance threshold for different types of damages (e.g., loose joints, cracks, openings). The threshold is defined by engineered experiment on pipes in typical operational condition of flow and pressure.

Duct Wearing Index (DWI)

Geometrical characteristics of pipes can be connected to energy content of acquired radial signals, in terms of frequency components and signal amplitudes.

In order to define an index for monitoring geometrical characteristics, the proposed algorithm uses wavelet technique that represents vibration signals (both characteristic and measured ones) by means Intrinsic Mode Function (IMF):

VlB _c {t) = MF _cj {t) + r _c (t) n

VlB _M2 (t) = MF _M2j (t) + r _M where r _x (t)\ residues which contains signal mean trend or is a constant;

The IMFxj(t) functions, originate from VIBc(t), are calculated starting from interpolation method, their content is connected to single frequency components of initial function. The energy content of vibration signal for j-th IMF is:

EIMF _j =

Now, it is possible define a Duct Wearing Index (DWI) that is proportional to the deviation of energy of IMF functions between characteristic vibration signal and measured one; as in the following equation:

EIMF _M2j - EIMF _cj \

DWh = DUCT WEARING INDEX

EIMF,

In this way, the proposed index can:

o keep track of the aging pipe through a comparison with the acquired historic data;

o adjust the characteristic function VIBc (t) to take into account variations of reference conditions compared to the initial status (e.g., constraints).

The percentage deviation of previous index is:

This value has to be compared to a known threshold THDWI%, i.e. the tolerance threshold of DWIj% for duct wearing (wear and corrosion). The threshold is defined by engineered experiment on pipes in typical operational condition of flow and pressure.

The processing and the discernment of the three functional parameters, BQ, DDI and DWI, is shown in FIG. 6. The proposed invention generates the following main outputs:

o Punctual duct data:

- Plano-altimetric georeferencing;

- Material;

- Nominal Diameter;

- Flow rate;

o Critically indication through alarm signaling;

o Critically type indication (leak, damaging, wearing) and quantification of the potential leak respect to MNF;

o Percentage intensity of alarm;

o Criticality localization. The three functional parameters allow system to discriminate intensity and type of alarm in metering area basing on FIG. 7. The system accuracy is lower than the MNF but, in the worst case, it is exactly equal to MNF.

Referring to FIG. 8, a preferred embodiment consists of a water distribution network, a wireless water monitoring network and, a water management remote control center. The system includes at least one adduction pipeline that is completely known; a plurality of sensor nodes 10 that are installed over pipes defining several leak metering areas. For each area, the proposed system includes a local collector node 1 1 which routes measured data to a cloud connected server 12 that collects, stores and analyzes data, as well as can transmit commands. The portion of pipe network in which leak detection is to be performed may be accessible or buried underground.

FIGS. 9 and 10 show an example of sensor device for monitoring and leak detection in at least a portion of fluid distribution system in accordance with one embodiment of the present invention. Such devices are highly-sensitive wireless sensor nodes 10 that can periodically detect, store and transmit measured data. The device is shown schematically in FIG. 10; it is positioned over the distribution pipe by means a magnetic fixing base. The device includes a vibration sensor, such as an accelerometer that continuously or periodically picks up the transverse vibrations induced by the fluid motion in the pipe and, also, an altimetry sensor and a georeferencing tag that determine the exact geographical position of the sensor node 10. The device is further provided with VHF (Very High Frequency) radio, for example 169MHz, that allows wireless communication with associated collector node.

Therefore, the system can also include a plurality of local collector nodes, FIG. 9 provides one of them for each area, for relaying data and commands, they can be installed at the center of the leak metering area above the ground. This device can be provided with two different transceivers: a VHF radio to communicate with sensor nodes 10 and receive their measurements and a GPRS radio to transmit received data towards a cloud connected server 12.

Each sensor node 10 periodically measures flow along the pipeline using a non- intrusive technique based on the acquisition of the vibrations induced by the fluid motion and punctual altimetry along the pipeline for pressure estimation. Other additional potential measurements can be pressure or quality of water. Each sensor node 10 transmits measured data to associated collector node which collects received measurements and transmits them to the server 12. In the water management remote control center, for each leak metering area, the system will perform a vibrational pattern check, to monitor the pipe status through the evaluation of predefined parameters (BQ, DDI and DWI), and the water volume balance to detect leaks. Starting from the three functional parameters, the system will generate a different type of alarm according with FIG. 7.