MX2011008941A - Monitoring system and device for detecting faults in transformers. - Google Patents

Abstract

The invention refers to a system for monitoring the behaviour of the operation of distribution transformers, with the purpose of detecting early faults, where the use of electronic devices coupled to the equipment for performing a diagnosis may help to anticipate flaws before they cause a catastrophe and risk the underground electric installation and users. The invention is a non-intrusive detection system, which may be used in both new transformers and transformers in current use, without performing important changes in the installation, the invention comprising a sensor for monitoring the behaviour of the installation, an inductive-type sensor for monitoring the behaviour of the installation, a device for conditioning the signal so as to lead it to a suitable measurement level so that, by means of a wireless transmission device, the data would be transmitted to a master hub and therefore to a master module for receiving the data in a wireless manner. A methodology specially designed for this purpose is used for interpreting the data and handling the information, this verifying that the data is within the normal operation ranges of the underground transmission line, these data being compared to the data contained in a database, allowing the type of fault to be identified in the installation as well as the location thereof within the transformer, an alarm being sent according to the fault level detected by the operator, who will take action for avoiding a catastrophic fault in the installation as well as users risks.

Description

System and Monitoring De for the Detection of Transformer Failures DESCRIPTION OBJECT OF THE INVENTION The object of the invention is a remote monitoring system and the des that comprise it for the on-site detection of faults in the electrical installations of distribution transformers, in order to detect incipient faults in the installation that may arise and by means of electronic des and inductive sensors coupled to the installation, make a diagnosis and anticipate the failures before they are catastrophic and endanger the installation, the electricity supply and the users.

It is a non-intrusive detection system, which can be used in both new and existing equipment, without making major changes in the installation, comprising a sensor de for monitoring the conditions of the electrical installation, a de for conditioning and signal processing, a wireless transmission de for the remote transmission of data to a master hub, which in turn is transmitted to a master station, where a server is located together with a personal computer, which by means of a method specifically developed for the interpretation of data and information management, you can analyze and see the state of the electrical equipment, anticipating a possible failure in the system.

BACKGROUND The electrical installations are comprised of power cables for medium and high voltage, transformers to supply the required voltage levels and the load supplied to the users. In the case of transformers for the distribution of electrical energy, their installation is carried out in different ways, between which they are fastened on the poles or in registers specially designed to contain them and / or on the sidewalk with special containers to avoid exposure to the users.

Nowadays, the detection of faults in electrical equipment is an already widely studied objective and with implemented solutions, since there are the necessary equipment and instruments to carry out this activity, highlighting that the detection of problems is currently carried out by means of equipment that requires that the installations are out of ser, which implies the interruption of the ser of the electrical supply and a preventive maintenance, however at present the world tendency focuses to the anticipated analysis of the behavior and operation of the installation in order to avoid the catastrophic failures , through the application of diagnostic techniques based mainly on the measurement of partial discharges and tests at reduced voltage at low frequency, incorporating the analysis study of the frequency response for the detection of mechanical faults, all this, as mentioned, performed with the installation out of ser.

In electrical equipment for medium and high voltage, are located mainly three characteristic points of failure: the insulation and the windings, the terminals or nozzles and the tap changer, where it has been detected that in more than 75% of the points of failure in the transformer, are generated in principle by the start of partial discharges (DP) long before becoming a real failure, so that the activity of the DPs in the system, is an important indicator that allows to estimate the level of degradation in equipment and installations of medium and high voltage.

This problem has been tried to solve in different ways; as in patent US 6,470,283, where a contactless monitoring system is proposed by means of a capacitive system where the behavior of an electrical system is detected by detecting power consumption, by means of sensors connected to a microprocessor for the analysis of the data collected, although the system is not focused on the detection of faults. In WO 2009/137902, a system for the monitoring of electrical energy is mentioned avoiding the theft of energy by means of a system connected directly to a transformer, comparing the energy supplied against the energy consumed in a certain point of consumption, this system is of direct measurement and there is no wireless data transmission. On the other hand in the patent WO 201 1/056057, a system of remote control and measurement of the electric power consumption of users connected to an electrical distribution network is mentioned, focused mainly to supervise and regulate the consumption of electrical energy at a distance. of users connected to a distribution transformer, this system has the characteristic that it is intrusive which can only be placed in new transformers, since they are immersed in the dielectric oil, a detection of a possible risk of failure is mentioned according to which the energy consumed by users. On the other hand in the article by D.J. McArthur, et al "The design of a multi-agent transforming condition monitoring system", IEEE Transaction on power systems, Vol 19, No. 4 Nov. 2004, mentions the detection of the start of failure in transformers by analyzing the dischargespartial (DP) present in a transformer using the detection methodology by means of ultra high frequency (UHF for its acronym in English).

According to the above and in order to solve the problem of giving solution to the aforementioned, we propose the development of a detection system, which allows remote monitoring and monitoring of the behavior of a distribution transformer in a facility electrical, characterized by having a sensor of the inductive type, non-intrusive for each of the phases of the transformer, which can be placed in both new and existing equipment without making additional modifications to them. By means of this system it is possible to monitor load variations, voltage variations and possible interferences that occur in an installation before the failure occurs through remote monitoring, in addition to this there is a system that warns when the transformer is out of operation due to different causes such as short circuits, explosion and / or theft. This system is constituted by means of three main sections, the detection and monitoring system, the wireless data transmission system of the detected variables and the system of acquisition, analysis and interpretation of the information.

BRIEF DESCRIPTION OF THE FIGURES Figure 1 . General scheme of the monitoring system.

Figure 2. Schematic view of the transformer installed in the pole with the detection system installed.

Figure 3. General diagram of the intelligent electronic device installed in the transformer.

Figure 4. General scheme of the signal conditioning system and data transmission Figure 5. General scheme of the installed concentrator.

Figure 6. General view of the wireless hub.

Figure 7. Block diagram of the wireless monitoring and transmission process.

Figure 8. Schematic view of the master station and the personal computer.

Figure 9. Methodology for the analysis, reception process and information management. Figure 10. View of the information displayed on the results analysis.

DETAILED DESCRIPTION OF THE INVENTION This invention is focused on the design of a monitoring and monitoring system for transformers for distribution voltages from 15 kV to 25 kV, characterized by being of the non-intrusive type, the sensor has placed in each of the phases, these sensors can be place in new equipment or equipment already installed and have a signal conditioning system coupled to a wireless transmission system in each transformer, forming a global network of wireless communication.

Figure 1 shows a schematic of the monitoring and monitoring system, consisting of a network of transformers (1), which can be of any type, whether installed on pole (2) or any other installation system, these transformers have installed the monitoring system (3) in each of them with the different sensors (4A-4E) in each of the phases, being directly connected to the signal monitoring and conditioning system for the wireless transmission process to a central concentrator (5), which can be placed independently on a post (2) or directly on the installation of the equipment. In this central concentrator (5) all the signals of the different sensor devices (3) (4) are collected and in turn, these signals are sent wirelessly to a master station (6) for later analysis by means of a personal computer ( 7), where the information is processed so that the operator of the system can know the state in which the equipment is located, and in case of an incipient failure, it is detected and by means of an alarm signal, warning the operator of this condition. This system for monitoring and detecting faults, consists of three main sections, the detection and monitoring device (3) (4), the wireless transmission system (5) of the detected variables and the system of acquisition, analysis and interpretation of information (6) (7) ). The topology of the communication system that is implemented is a network of type ZigBee Mesh Networking, the main feature is that all its wireless members (3) serve as repeaters of information, looking for it to reach the central concentrator (5) by any means, developing an interconnected network where every element is a means of communication to the master station (6). The system seeks to establish a reliable communication between several stations with a personal computer (7) and a master module (6) through wireless signals, each station is mainly based on the serial communication system.

Monitoring consists of sending the signals coming from the sensor devices (3) (4) to the computer (7) for further analysis. Each channel receives a wireless signal from a sensor (3) (4), these can be up to a number of 16 per module and are distributed in the underground installation of the system that is to be monitored, there is a data acquisition system. type multiplexed and is connected to a personal computer (7).

Figure 2 shows the arrangement of the monitoring system installed in a distribution transformer which consists of the non-intrusive detection system of the inductive type (4), which is composed of an inductive sensor (4a-4e) and a system signal conditioning and data transmission (3), which allows the monitoring and supervision of the operating conditions of the distribution transformer, characterized by being of the non-intrusive type that can be placed in both new and existing equipment without doing modifications in them. Through this system you can monitor load variations, voltage variations and possible interferences that occur in a facility before the failure occurs. The detection system is based on the response to the detection of partial discharges (DP); which is done by means of a sensor of the inductive type (4), which allows us to detect the variations in the current flowing through the terminals of the transformer as well as the DP present in the installation. The sensor device (4a-4e) is directly connected to a signal conditioning system (3) and performs wireless transmission to the master concentrator (5). The detection system (4) is preferably constituted by the construction of a coil, which is the current sensing device (4a-4e) and the selection of the circuit for conditioning and processing the signal (3), due to the nature of the the application, it is preferable that the sensor is constructed in a flexible form, containing a discontinuity that allows it to be opened to make measurements of the non-intrusive type and the diameter of the magnet wire used for the construction of the sensor, preferably having a diameter between 36 AWG to 28 AWG, in such a way as to ensure the robustness of the coil and the sensitivity ratio. The bandwidth should preferably allow the coil to perform well at low and high frequency. The signals obtained by means of the inductive sensor device (4), are processed by means of the second section, in such a way that by means of a wireless device (3) they facilitate the sending of the information and the operation in places where the computers can not remain on the site, as are the different registers and points of the installation that is intended to be monitored.

Figure 3 shows the detail of the installation of the detection and monitoring system in a distribution transformer (1) where the sensors (4a-4e) are placed in the terminals or nozzles of the transformer (la-le), which can be from 5 to 7, depending on the type and characteristics of the transformer to monitor, these sensors (4a-4d), are connected to a signal conditioning system and wireless transmission (3), which in turn transmits the data to a concentrator station (5), for subsequent transmission to a master module (6) and subsequent analysis on a personal computer (7).

Figure 4 shows a diagram of the electrical diagram of the signal conditioning system (3), shown in Figure 1, this system has four sections, is preferably formed by an acquisition section of the system variables (8). ), the integration section of the signal for its conditioning (9), an amplification section (10) to bring it to suitable values for analysis and a section for correction and filtering of the signal (1 1), must preferably have the following characteristics: be of low noise, have a bandwidth preferably greater than or equal to 25 MHz. For the case of DP detection, it must preferably have the capacity to handle capacitive loads connected to its output, this due to the use of coaxial cable for the connection of the conditioning circuit of the signal to the system wireless (3), shown in Figure 1, preferably must be powered by batteries, to avoid the phenomenon of beat frequency during the measurement of alternating electrical current, without neglecting any other source of power supply.

Figure 5 shows the schematic of the typical installation of the master concentrator (5), which can be installed in an independent pole (2) or directly in the same pole that has the detection system (4), as shown in figure 2, this in order to make a network for the transmission of data through an interconnected network where all elements are a means of communication to the master station (6), seeking to establish reliable communication between several stations with a personal computer (7) and a master module (6) through wireless signals, each station is mainly based on the serial communication system.

In figure 6 it shows the general scheme of the master concentrator station (5), which is constituted by the wireless reception system (5a), the antenna for sending the signal (5b) to the master module (6) and the electric power supply system (5c). Where the wireless reception system (5a) has the function of receiving data from the monitoring and monitoring system (4), sent by means of the signal conditioning and wireless transmission system (3), where the information is stored for later sending by means of packages to the master module (6) for later analysis by means of a personal computer (7). The antenna for sending the signal (5b) to the master module (6) is constituted in such a way that the information is sent by radiofrequency at 2.4 Ghz, by a serial protocol for data transmission at 250 kbps (12b), with connectivity line of sight up to 1.6 km., and used a decoder (13a) with mesh topology based on the reference model for the interconnection of open systems (13) as shown in figure 7. This module is powered by means of a system based on renewable energies either by means of solar panels or low capacity wind turbines, in order to make the The system is autonomous and independent of the power supply of the line, allowing it to continue sending the necessary information to make decisions about the control of the network and the conditions of the transformer.

Figure 7 shows the system of processing and transmission of the signal, which comprises the wireless sensor devices (12a), is responsible for measuring electromagnetic induction incipient faults in a distribution transformer, in turn this information is sent by radio frequency at 2.4 Ghz, for a serial protocol for data transmission at 250 kbps (12b), with line of sight connectivity up to 1.6 km., and using a decoder (13a) with mesh topology based on the reference model for interconnection of open systems (13). For the connectivity with the sensors, it can be exchanged autonomously to a multicast network (Multicast), where each communication channel or sub-channel, through a central node or a peripheral node, working in a bidirectional way, sharing the available bandwidth . The connectivity with the central computer (7) is based on the centralization of the network by a star topology, the events are sent by several communication protocols from the data concentrator (13). Which contains a monitoring control which is responsible for the remote acquisition of data and identifications of the faults, coming from the sensor module (12), for sending them to the communications control (14). The protocol exchange and communication control device (13b) allows to accommodate the information sent by the sensors, replicating it to the expert diagnostic and assistance system (SEDA), using Radio Frequency connectivity at 900 MHz. (13f), it is connected at a distance with a point of view up to 64 km. with a 256-bit encryption, and a transmit power of preferably 4 watts using a 6 dB antenna, at a data transmission rate of up to 15,200 bps., the receiving modem after its encoder. It generates outputs with serial protocol, Wi-Fi, Satelital, BlueTooth and Ethernet, on the other hand, the protocol exchanger generates alarms (13c), which are sent as a text message (SMS), through the cellular network (13e) ) in quad band mode, operating preferably within the range of 824 - 849 MHz to 869 - 894 MHz, with 124 for cellular and 1850 - 1910 MHz to 1930 - 1990 MHz, with 299 computer channels, without ruling out the following ranges: from 890 - 915 MHz to 935 - 960 MHz with 124 channels for mobile phones and from 1710 - 1785 MHz to 1805 - 1880 MHz with 374 channels for computers. This communication is established for personal mobile devices (14a) or remote devices, by means of text messages (SMS). It has different ways of communicating and connecting to any device, machine or industrial process. The protocol exchanger (14b) of the expert diagnostic and assistance system (14) contains a Multiplexer SGW1 -MMP, with which it is connected to different Modbus masters with the same slave or vice versa. In addition, you can combine different serial port speeds and different types of Modbus serial (ASCII or RTU, RS232 or RS485).On the other hand, autocontiene the possibility of generating a wireless RS232 serial port device, using wireless technology by RF, Wi-Fi, satellite and cellular. The IPC of the system allows packaging the DNP protocol, in TCP / IP data packets within a GSM / GPRS cellular communication, in such a way that any equipment with said protocol can be connected remotely, at a very low cost. By using the DNP and ModBus protocols, which are industrial protocols for communications between intelligent equipment (IED) and substations of SCADA systems, which are widely used in the electrical sector. The Monitoring Control module (15) is composed of four elements: Expert system for underground line diagnosis (15a), Information Management module (15c) and the global database (15b).

The Information Management System (14) contains algorithms to identify characteristic line elements and performs the monitoring of the underground electrical line. The control of communications is responsible for prioritizing the signals that are generated in the database and is called Information Management System (15c), which must be sent to the Graphic Interface module. The traffic of information in a normal process concentrates in the monitoring control, in emergency cases alarm signals are also sent (13c), due to incidents in the transmission line and / or in the monitoring control. Depending on the incident that occurs, one or more warning signals may be activated. In the expert system module (15a), the diagnosis of the status of the transmission line is made and allows the user to be assisted, with the possibility of performing the functions of remote monitoring of point faults manually or autonomously, allowing an estimate to be made of failures and preventive maintenance. This wireless transmission system facilitates the remote and online operation of the reclosers, considerably improving maintenance logistics, allows early detection of power failures, end users get a better quality of supply with very low communication cost and only It pays for transferred data, it also allows the detection by loss of signal when a transformer is out of operation, sending an alert to the operator warning of a possible catastrophic failure or theft of the equipment.

The protocol exchanger and the communication exchanger connect up to 65,535 Masters with 1 Slave or connect up to 65,535 slaves with 1 master, compatibilize different bytes per second (bps) of each port and adapt different types of communication protocols in odbus (ASCII or RTU ) of each port, acts as a Modbus ACII to Modbus RTU converter and vice versa.

Figure 8 shows the system arrangement of the master module (6), which is composed of an antenna for the reception of the signal preferably at 6 dB, coming from the inductive sensors (3) (4) and the central concentrator (5), shown in Figure 1, which converts and conditions the signal preferably according to those described in Figure 4, to be sent later to the master station (6) for further analysis and interpretation by means of a personal computer ( 7).

Figure 9 shows the flow diagram of the process of analysis and interpretation of the signals sent wirelessly to the master module (6), shown in figure 1, the implementation of this computational algorithm is preferably based on object-oriented programming language to facilitate interaction with the user. It is preferably done according to the following process: after the variables are detected by the sensors (3) (4), sent to the central concentrator (5) and subsequently to the master station (6), the process starts with the detection of the variables (91) of voltage and current in each of the phases of the transformer (1), depending on the detected variables are made the calculations of different parameters (92) such as the input power, the output power, the variations load and possible interference in the electrical system due to atmospheric discharges and switch maneuvers among others, The input signals are processed and analyzed (93) preferably using the Fast Fourier Transform, in order to detect incipient failures in the transformer (1). This process is compared (94) with a database (95) in which the different alarm levels specified by the user are configured, in case the specified alarm level is not exceeded, the process returns to evaluate a new alarm level. variable detection value (91. In the case of exceeding the alarm level, the automatic fault detection system when an anomaly is detected in the frequency spectrum formed preferably by the harmonics detected in the signal being analyzed. they are compared automatically (97) with the patterns stored in the database (96) and correspond to the previously characterized fault spectra., alarm signals (98) will be activated at different levels preferably according to what is described in figure 7, by means of SMS and sending signals to any DM equipment, without ruling out any other type of communication.

Figure 10 shows the method for analyzing the data that is being collected, as well as the information that can be obtained from it. To carry out the analysis of these signals, there is an algorithm, based on object-oriented programming, by means of which it is intended to alert the operator of possible anomalies that arise in the transmission line. This algorithm consists of several windows and screens that facilitate its operation. Being able to take actions based on the values of the signals that are in the interface is without a doubt one of the main advantages of this type of systems.

To carry out this task, the input signal is taken for each channel (16), and general information is obtained, such as frequency and amplitude (17). In addition, the fast Fourier transform is obtained, in order to obtain information as peak and phase magnitude of the harmonic components of the signal (18).

The analysis of the signals that enter the computer equipment is based on the values of the signal and its harmonics (19). The number of harmonics to analyze is determined by the user. The purpose is to find trends, associations and patterns that allow us to identify a potential failure (20). The method is based on the visualization of different waveforms from which different measurements are taken to detect anomalies in the transmission. The principle of operation is preferably based on the analysis of the signal (93), according to the flow diagram shown in Figure 9, by means of the fast Fourier transform to detect the harmonics present in the signal. This process is carried out for the most common faults that occur in the transformers (1), such as failures in the insulations and windings, terminals or nozzles and the tap changers shown in figure l. The idea is to associate each type of failure with a particular pattern (96) (97). These patterns (96) are associated with the magnitude of different harmonics present in the signal (93). In the interface, the first fourteen harmonics per channel can be displayed, however the other harmonics are analyzed, although their value is not displayed in the interface.

EXAMPLES In order to exemplify the operation of the system, Figure 10 shows the detection of a fault in a system of distribution of underground electrical energy (16), this presents a characteristic pattern to identify it (17), an abrupt increase is observed in the amplitude and a decrease in the frequency with respect to the operating conditions in steady state (18). Upon detecting a fault, the electrical characteristics identified (19) in the transformer (1) are compared with the previously identified patterns (20) and stored in a database of the computer equipment (96). The comparison is carried out by means of the process according to Figure 9.

Among the characteristics that identify each of these failures in the transformer monitoring systems, we have the following: Insulation and windings failures.- There is a decrease in the magnitude, accompanied by variations in the harmonic components present in the transmission line.

- Failures in the terminals or nozzles.- There is an abrupt increase in the amplitude of the transient spectrum and a decrease in the frequency of operation.

- Faults in tap changers. - There is a decrease in the magnitude of the signal, due to the difference in potential generated between the mismatched parts, facilitating losses and low energy efficiency.

During the process of fault detection, monitoring and detection is carried out in a preventive manner, since once an anomalous operation condition (94) is identified, warning signals (98) are activated. The warning signs are presented in three levels: visual, auditory and by means of text messages to mobile devices in order that the operator is aware of the start of failure that occurs in the transmission line, indicating both the type of it fails as the place where it presents itself to take the necessary measures.

Claims (18)

CLAIMS Having described the invention enough, I declare my property contained in the following clauses:

1. A system for monitoring the prediction of transformer faults for electrical voltages between 5 kV to 34.5 kV, comprising a sensor device for monitoring the operating conditions of equipment, a system for the treatment and conditioning of the signal, a system for transmitting the data wirelessly to a master concentrator and a reception system to a master station for the analysis and interpretation of the data collected with the previous devices by means of a personal computer.

2. A monitoring and monitoring device for distribution transformers, according to claim 1, characterized by allowing the monitoring of the behavior of the equipment by detecting the variables of voltage, current, load and partial discharges.

3. A monitoring and monitoring system for distribution transformers, according to claim 1, characterized by allowing early detection of faults in the transformer, improving maintenance logistics.

4. A sensing device according to claim 2, characterized by a current sensor of the non-intrusive inductive type in a flexible manner and with a discontinuity that allows the opening and closing of the sensor to be installed in each of the phases of the transformer, preferably with a conductor diameter that guarantees the robustness of the sensor and a bandwidth sensitivity ratio for good performance at high and low frequencies.

5. A signal conditioning and processing circuit according to claim 1, characterized by an integration section of the signal for conditioning, an amplification section and a correction section of the signal.

6. A circuit for conditioning and processing the signal, according to claim 1, characterized by being low noise, having a bandwidth preferably greater than or equal to 25 MHz, characterized by the ability to handle capacitive loads connected to its output.

7. A wireless transmission system, according to claim 1, characterized by having a wireless sensor system, a protocol exchange system and a diagnosis and assistance system for the acquisition and analysis of information by means of a personal computer.

8. A wireless system for data transmission, according to claim 1, characterized by connecting up to 65,535 masters with a slave.

9. A wireless system for data transmission, according to claim 1, characterized by connecting up to 65,535 slaves with a master.

10. A wireless system for data transmission, according to claim 1, characterized by combining different bytes per second (bps) of each port and different types of Modbus (ASCII or RTU) of each port.

1. A wireless system for data transmission, according to claim 1, characterized as acting as a Modbus ACII to Modbus RTU converter and vice versa.

12. A wireless system for data transmission, according to claim 1, characterized by generating outputs with serial protocol, Wi-Fi, satellite, BlueTooth and Ethernet.

13. A wireless system for data transmission, according to claim 1, characterized by generating alarms, which are sent as a text message (SMS), through the cellular network (GSM) in quad band mode, preferably operating within the range of 824 - 849 MHz at 869 - 894 MHz, with 124 for cellular and 1850 - 1910 MHz at 1930 - 1990 MHz, with 299 computer channels, without ruling out the following ranges: from 890 - 915 MHz to 935 - 960 MHz with 124 cellular channels and from 1710 - 1785 MHz to 1805 - 1880 MHz with 374 channels for computers.

14. A system for analyzing and interpreting the collected data, according to claim 1, characterized by a master receiver and a personal computer for the analysis of the data.

15. A system for analyzing and interpreting the collected data, according to claim 13, characterized by being of the multiplexed type or simultaneous acquisition of the signal wirelessly.

16. A method for analyzing and interpreting the data collected, according to claim 13, characterized by being based on object-oriented programming, allowing the operator to be alerted of possible anomalies that arise in the transmission line.

17. A method for the analysis and interpretation of the data collected, according to claim 15, characterized by being composed of several windows and screens, facilitating the operation by the user, allowing to take actions based on the values of the data collected by the sensor devices and sent to the master receiver.

18. A method for the analysis and interpretation of the data collected, according to claim 15, characterized by being based on the analysis of the values of the input signal, depending on their harmonics and the number to be analyzed, in order to to be able to find trends, associations and patterns that allow us to identify a potential failure.