Fault monitoring systems and methods for detecting connectivity faults

Abstract

A fault monitoring system detects connectivity faults in a device control centre. The fault monitoring system includes, for the or each control section, at least a first pair of temperature sensors adapted to detect the temperature of a first pair of terminals (T.sub.L1-IN), (T.sub.L1-OUT). The system includes a processor configured to receive the detected temperatures, calculate an IN-OUT difference, the IN-OUT difference being a difference between the temperatures of the first pair of terminals (T.sub.L1), compare the calculated IN-OUT difference with a predetermined threshold value (T*.sub.L1), whereby a calculated difference IN-OUT (T.sub.L1), greater than the predetermined threshold value (T*.sub.L1), is indicative of a connectivity fault at one of the first pair of terminals, and to generate at least one output signal based on the results of the comparison.

Claims

1. A fault monitoring system for detecting connectivity faults in a device control center, the device control center including at least one control section configured to control a supply of electrical power from a power source to a respective at least one device, the or each control section being electrically connected between the power source and the respective device at least at a first pair of terminals, one of the first pair of terminals connecting the control section to the power source side of a first power line and the other of the first pair of terminals connecting the control section to the device side of the first power line such that the control section completes the power circuit of the device, the fault monitoring system comprising: for the or each control section, at least a first pair of temperature sensors, one of the first pair of temperature sensors being adapted to detect a temperature of one of the first pair of terminals (T.sub.L1-IN), and the other of the first pair of temperature sensors being adapted to detect a temperature of the other of the first pair of terminals (T.sub.L1-OUT); a processor configured to receive the detected temperatures, calculate an IN-OUT difference, the IN-OUT difference being a difference between the temperatures of the first pair of terminals (T.sub.L1), compare the calculated IN-OUT difference with a predetermined threshold value (T*.sub.L1), whereby if the calculated IN-OUT difference (T.sub.L1) greater than the predetermined threshold value (T*.sub.L1), this is indicative of a connectivity fault at one of the first pair of terminals, and to generate at least one output signal based on results of the comparison; and an output module adapted to receive the output signal(s) generated by the processor and to communicate the output signal(s) externally.

2. The fault monitoring system according to claim 1, wherein the or each control section is additionally electrically connected between the power source and the respective device at a second pair of terminals, the second pair of terminals connecting the control section to the source side and device side of a respective second power line, the first and second power lines carrying power with different respective phases (L.sub.1, L.sub.2), and the fault monitoring system further comprises: for the or each control section, a second pair of temperature sensors, one of the second pair of temperature sensors being adapted to detect a temperature of one of the second pair of terminals (T.sub.L2-IN), and the other of the second pair of temperature sensors being adapted to detect a temperature of the other of the second pair of terminals (T.sub.L2-OUT); wherein the processor is further configured to calculate the IN-OUT difference between the temperatures of the terminals in the second terminals (T.sub.L2), compare the or each calculated IN-OUT difference with a predetermined respective threshold value (T*.sub.L2), whereby a calculated IN-OUT difference (T.sub.L2) greater than the corresponding predetermined threshold value (T*.sub.L2) is indicative of a connectivity fault at one of the respective pair of terminals, and to generate at least one output signal based on results of the comparison between the calculated IN-OUT difference (T.sub.L2) and the corresponding predetermined threshold value (T*.sub.L2).

3. The fault monitoring system according to claim 1, wherein the processor is further configured to compare the detected temperature of each of the first pair of terminals (T.sub.L1-IN, T.sub.L1-OUT) against a respective predetermined threshold value (T*.sub.L1-IN, T*.sub.L1-OUT), whereby a detected temperature greater than the corresponding predetermined threshold value is indicative of a connectivity fault at the respective terminal, and to generate at least one output signal based on results of comparisons between the detected temperature and the corresponding predetermined threshold value.

4. The fault monitoring system according to claim 3, wherein the fault monitoring system further comprises an ambient temperature sensor, adapted to detect an ambient temperature T.sub.AMB, and the processor is further configured to compare the detected temperature of each terminal (T.sub.L1-IN, T.sub.L1-OUT, . . . ) relative to the detected ambient temperature T.sub.AMB against a respective predetermined threshold value (T*.sub.L1-IN/AMB, T*.sub.L1-OUT/AMB, . . . ), whereby a detected relative temperature greater than the corresponding predetermined threshold value is indicative of a connectivity fault at the respective terminal, and to generate at least one output signal based on results of the comparison between the detected relative temperatures in the corresponding predetermined threshold values.

5. The fault monitoring system according to claim 2, wherein the processor is further configured to calculate an imbalance difference, the imbalance difference being a difference (T.sub.L1/L2-IN) between the temperatures of the power source side terminal of the first power line (T.sub.L1-IN) and the power source side terminal of the second power line (T.sub.L2-IN), and/or a difference (T.sub.L1/L2-OUT) between the temperatures of the device side terminal of the first power line (T.sub.L1-OUT) and the device side terminal of the second power line (T.sub.L2-OUT), and to compare the calculated imbalance difference with a predetermined respective threshold value (T*.sub.L1/L2-IN and/or T*.sub.L1/L2-OUT), whereby a calculated imbalance difference greater than the corresponding predetermined threshold value is indicative of an imbalance between the power supplied on the first power line and the power supplied on the second power line, and to generate at least one output signal based on results of the comparison between the calculated imbalance difference and the corresponding predetermined threshold value.

6. The fault monitoring system according to claim 1, wherein the processor is adapted such that the at least one output signal generated by the processor includes at least one status signal which is an alarm signal if a connectivity fault at one of the terminals or an imbalance between two of the power lines is indicated, and is a no-alarm signal otherwise.

7. The fault monitoring system according to claim 1, wherein the processor is adapted such that the at least one output signal generated by the processor includes detected temperature data which preferably comprises any of: the detected temperatures of one or more of the terminals (T.sub.L1-IN, T.sub.L1-OUT, . . . ), and/or one or more calculated temperature differences (T.sub.L1, T.sub.L2, T.sub.L1/L2-IN, T.sub.L1-IN/AMB, . . . ).

8. The fault monitoring system according to claim 1, wherein the output module comprises a communications device configured to convert the output signal(s) to a network data protocol and to communicate the output signal(s) to an external device via a network, the network data protocol preferably being a serial communications protocol, preferably being MODBUS.

9. The fault monitoring system according to claim 1, wherein the fault monitoring system comprises a fault monitoring module for each of the at least one control sections, each fault monitoring module comprising: the at least a first pair of temperature sensors for the respective control section; the processor, wherein the processor is a local processor configured to receive the detected temperatures from the at least a first pair of temperature sensors for the respective control section only; and the output module, wherein the output module is a local output module.

10. The fault monitoring system according to claim 9, wherein the device control center comprises a plurality of control sections, the fault monitoring system comprising a respective plurality of fault monitoring modules.

11. The fault monitoring system according to claim 10, comprising a network of fault monitoring modules and a controller, wherein each fault monitoring module comprises an output module for communication with the controller across the network, wherein the output module comprises a communication device configured to convert the output signal(s) to a network data protocol and to communicate the output signal(s) to an external device via a network, the network data protocol preferably being a serial communications protocol, preferably being MODBUS.

12. A device control center, comprising at least one control section configured to control the supply of electrical power from a power source to a respective at least one device, the or each control section being electrically between the power source and the respective device at least a first pair of terminals, one of the first pair of terminals connecting the control section to of the power source side of a first power line and the other of the first pair of terminals connecting the control section to the device side of the first power line such that the control section completes the power circuit of the device, and a fault monitoring system in accordance with claim 1.

13. A method for detecting connectivity faults in a device control center, the device control center comprising at least one control section configured to control the supply of electrical power from a power source to a respective at least one device, the or each control section being electrically connected between the power source and the respective device at least a first pair of terminals, one of the first pair of terminals connecting the control section to of the power source side of a first power line and the other of the first pair of terminals connecting the control section to the device side of the first power line such that the control section completes the power circuit of the device, the method comprising: for the or each control section, detecting a temperature of one of the first pair of terminals (T.sub.L1-IN), and the other of the first pair of terminals (T.sub.L1-OUT); in a processor, calculating an IN-OUT difference, the IN-OUT difference being a difference between the temperatures of the first pair of terminals (T.sub.L1), comparing the calculated IN-OUT difference with a predetermined threshold value (T*.sub.L1), whereby a calculated IN-OUT difference (T.sub.L1) greater than the predetermined threshold value (T*.sub.L1) is indicative of a connectivity fault at one of the first pair of terminals, and generating at least one output signal based on results of the comparison; and at an output module, receiving the output signal(s) generated by the processor and to communicate the output signal(s) externally.

14. The method according to claim 13, wherein the or each control section is additionally electrically connected between the power source and the respective device at a second pair of terminals, the second pair of terminals connecting the control section to a source side and device side of a respective second line, the first and second power lines carrying power with different respective phases L.sub.1, L.sub.2, and the method further comprises: for the or each control section, detecting a temperature of one of the second pair of terminals (T.sub.L2-IN), and the other of the second pair of terminals (T.sub.L2-OUT); and optionally in the processor, calculating the IN-OUT difference between the temperatures of the terminals in the second pair of terminals (T.sub.L2), comparing the or each calculated IN-OUT difference with a predetermined respective threshold value (T*.sub.L2), whereby a calculated IN-OUT difference (T.sub.L2) greater than a corresponding predetermined threshold value (T*.sub.L2) is indicative of a connectivity fault at one of a respective pair of terminals, and generating at least one output signal based on results of the comparison between the calculated IN-OUT difference (T.sub.L2) and the corresponding predetermined threshold value (T*.sub.L2).

15. The method according to claim 13, further comprising comparing the detected temperature of each terminal (T.sub.L1-IN, T.sub.L-OUT) against a respective predetermined threshold value (T*.sub.L1-IN, T*.sub.L1-OUT), whereby a detected temperature greater than the corresponding predetermined threshold value is indicative of a connectivity fault at the respective terminal, and generating at least one output signal based on results of the comparison between the detected temperature and the corresponding predetermined threshold value.

16. The method according to claim 15, further comprising detecting an ambient temperature T.sub.AMB, comparing the detected temperature of each terminal (T.sub.L1-IN, T.sub.L1-OUT, . . . ) relative to the detected ambient temperature T.sub.AMB against a respective predetermined threshold value (T*.sub.L1-IN/AMB, T*.sub.L1-OUT/AMB, . . . ), whereby a detected relative temperature greater than the corresponding predetermined threshold value is indicative of a connectivity fault at the respective terminal, and generating at least one output signal based on results of the comparison between the detected relative temperature and the corresponding predetermined threshold value.

17. The method according to claim 14, further comprising calculating an imbalance difference, the imbalance difference being a difference (T.sub.L1/L2-IN) between the temperatures of the power source side terminal of the first power line (T.sub.L1-IN) and the power source side terminal of the second power line (T.sub.L2-IN), and/or a difference (T.sub.L1/L2-OUT) between the temperatures of the device side terminal of the first power line (T.sub.L1-OUT) and the device side terminal of the second power line (T.sub.L2-OUT), comparing the calculated imbalance difference with a predetermined respective threshold value (T*.sub.L1/L2-IN and/or T*.sub.L1/L2-OUT), whereby a calculated imbalance difference greater than the corresponding predetermined threshold value is indicative of an imbalance between the power supplied on the first power line and the power supplied on the second power line, and generating at least one output signal based on results of the comparison between the calculated imbalance difference and the corresponding predetermined threshold.

18. The method according to claim 13, wherein the at least one output signal includes at least one status signal which is an alarm signal if a connectivity fault at one of the terminals or an imbalance between two of the power lines is indicated, and is a no-alarm signal otherwise.

19. The method according to claim 13, wherein the at least one output signal includes detected temperature data which preferably comprises any of: the detected temperatures of one or more of the terminals (T.sub.L1-IN, T.sub.L1-OUT, . . . ), and/or one or more calculated temperature differences (T.sub.L1, T.sub.L2, T.sub.L1/L2-IN, T.sub.L1-IN/AMB, . . . ).

20. The method according to claim 13, further comprising converting the output signal(s) to a network data protocol and communicating the output signal(s) to an external device via a network, the network data protocol preferably being a serial communications protocol, preferably being MODBUS.

21. A computer program product storing a sequence of steps configured to perform the method of claim 13.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 shows a first embodiment of a fault monitoring system in an exemplary device control centre;

(2) FIG. 2 shows a second embodiment of a fault monitoring system in another exemplary device control centre;

(3) FIG. 3 schematically depicts a further example of a device control centre with a fault monitoring system;

(4) FIG. 4 shows exemplary contents of one of the cases of the device control centre of FIG. 3 in one embodiment;

(5) FIG. 5 shows exemplary contents of one of the cases of the device control centre of FIG. 3 in another embodiment;

(6) FIG. 6 shows exemplary contents of one of the cases of the device control centre of FIG. 3 in a further embodiment;

(7) FIGS. 7(a) and 7(b) show two examples of fault monitoring modules in accordance with further embodiments;

(8) FIGS. 8(a) and 8(b) show portions of two fault monitoring modules in accordance with further embodiments;

(9) FIGS. 9(a) and 9(b) illustrate two exemplary temperature sensor implementations as may be used in any of the embodiments; and

(10) FIGS. 10 and 11 schematically illustrate two exemplary networks forming fault monitoring systems in further embodiments of the invention.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

(11) As discussed above, the presently disclosed fault monitoring system and method is suitable for deployment in any device control centre. Typically the device control centre will be configured to control a plurality of devices via a corresponding plurality of control sections. However the present system and method can equally be used in device control centres controlling a single device. Thus the principles of operation of the fault monitoring system and method will initially be described with reference to control of a single device but as will be seen the same technique can be employed in device control centres (e.g. MCCs, switchboards, power distribution centres etc.) with any number of devices.

(12) Therefore FIG. 1 illustrates a device control centre (DCC) 1 configured to control the supply of power from a power source PS to a single device D. In this case the device D operates on single-phase power and thus is supplied with power via a single power line L.sub.1. In this embodiment a return or neutral line L.sub.n is also present, but this is not required in all embodiments as discussed below. The device D could be a small pump circuit, for example. The DCC 1 includes a control section 5 connected between the power source side of the power line L.sub.1 at a first terminal L.sub.1-IN and the device side of the power line L.sub.1 at a second terminal L.sub.1-OUT. The first and second terminals are referred to as a first pair of terminals since they are both on the same power line which the control section 5 completes. The control section 5 includes switchgear 4, which here is represented schematically as including a simple switch but in practice may include any suitable control components including switches, circuit breakers, disconnectors, isolators, fuses, soft-starters etc. The control section 5 is typically contained within a case 2 such as the drawers or buckets mentioned previously, but this is optional.

(13) Provided locally to the control section 5 (and preferably within case 2) is a fault monitoring module 10 which in this embodiment constitutes the fault monitoring system but in other embodiments may form a part thereof as described further below. In this embodiment, the fault monitoring module 10 comprises a processor 11, first and second temperature sensors 12a, 12b and an output module here taking the form of an alarm device 16 and an optional remote output line 15/17 for outputting signals to a remote device or network R. Other optional features provided in the fault monitoring module 10 in this embodiment include a further temperature sensor 19 and a power supply 18 taken from the control section 5, both of which will be discussed below.

(14) The temperature sensors 12a, 12b are arranged to measure the temperatures of the first and second terminals on the power line L.sub.1. Thus, sensor 12a measures the temperature T.sub.L1-IN of the first terminal L.sub.1-IN and sensor 12b measures the temperature T.sub.L1-OUT of the second terminal L.sub.1-OUT. In practice, depending on the type of temperature sensors used, the temperatures may be measured from one of the cables joined at the respective terminals rather than directly on the terminals themselves, but in this case the measurements are taken from points as close to the terminals as possible. The processor 11 receives temperature measurements from the sensors which are updated at regular intervals so as to provide substantially real-time, continuous monitoring. For example, the temperatures from the sensors should be detected at least every 60 seconds.

(15) The processor 11 in this embodiment is pre-programmed to calculate the difference T.sub.L1 between the measured temperatures T.sub.L1-IN and T.sub.L1-OUT, and to compare the calculated difference T.sub.L1 against a predetermined threshold value T*.sub.L1. The predetermined threshold value T*.sub.L1 is set such that a calculated value T.sub.L1 greater than the predetermined threshold value T*.sub.L1 is deemed indicative of a connectivity fault at either one of the first and second terminals. Various methods for setting this and other predetermined threshold values for use by the processor are discussed below. Based on the comparison, the processor 11 generates at least one output signal which is communicated externally by an output module, here comprising a local alarm device 16, e.g. in the form of a light source such as an LED. Alternatively or in addition the output signal(s) may be transferred to a remote device or network R on line 15/17. The output signal(s) could for instance include a status signal which indicates the state of the terminals L.sub.1-IN, L.sub.1-OUT by transitioning from a non-alarm signal to an alarm signal when the processor calculates a temperature difference T.sub.L1 which is greater than the threshold T*.sub.L1. Preferably the alarm is latching meaning that should the calculated temperature difference T.sub.L1 fall back below the threshold T*.sub.L1, the status signal will remain in an alarm condition until reset by an operator. To reduce the possibility of false alarming, the processor 11 may be configured to wait until the calculated temperature difference T.sub.L1 has remained at a value greater than the threshold T*.sub.L1 for the duration of a predetermined time period before outputting an alarm signal, e.g. 60 seconds.

(16) The local alarm device 16 is controlled by the output signal from the processor 11, e.g. in the case of a light source becoming illuminated when the status symbol is an alarm signal and remaining off otherwise (or vice versa) or the light source changing colour. The alarm device can also be controlled by the processor 11 to provide additional information such as identifying to the operator which terminal L.sub.1-IN or L.sub.1-OUT has the fault, since this will be the terminal having the higher of the two measured temperatures. The alarm device 16 could be controlled to output this information e.g. by being actuated in accordance with a code sequencefor instance in the case of a light source it may be configured to flash or blink in a certain first sequence to indicate that terminal L.sub.1-IN has the fault and to flash or blink in a different second sequence to indicate that terminal L.sub.1-OUT has the fault.

(17) If a remote output line 15/17 is provided, this could simply output the same status signal, e.g. via a relay output. The remote device R could be a remote alarm device such as another light source or a siren, or could for instance be a computer, or a network of computers, having an appropriate graphical user interface configured so as to display the status of the terminals at control section 5 to a user, and to alert them to any alarm signal. Alternatively or in addition the output signals output on the remote line R could include the measured and/or calculated temperature data in which case this may be in the form of a serial data communications protocol such as MODBUS.

(18) Optionally, in addition to the above-described IN-OUT function, the fault monitoring system may also be responsive to the individual measured temperatures of the first and second terminals (T.sub.L1-IN and T.sub.L1-OUT), either absolute or referenced against the ambient temperature, which is referred to as the HIGH function. In the latter case the ambient temperature is measured by the further temperature sensor 19. The processor implements the HIGH function by comparing each measured temperature T.sub.L1-IN and T.sub.L1-OUT against respective predetermined thresholds T*.sub.L1-IN and T*.sub.L1-OUT (which may have the same value for each terminal, or could be different for each terminal), optionally referenced against the measured ambient temperature. If one (or more) of the measured temperatures T.sub.L1-IN and/or T.sub.L1-OUT is greater than the corresponding threshold, this is indicative of a connectivity fault at the respective terminal. Since the HIGH function is intrinsically vulnerable to false alarming due to unavoidable fluctuations in the load on the power line L.sub.1 (and hence the temperatures of the terminals), the predetermined thresholds T*.sub.L1-IN and T*.sub.L1-OUT may be set relatively high in order to reduce the occurrence of false alarms and/or the processor 11 may be configured to output an alarm signal only after the measured temperature has remained above the corresponding threshold value for a certain time period, e.g. 60 seconds. The HIGH alarm may be communicated externally either by the same local alarm device 16 (preferably by actuating it in a different manner, e.g. causing it to illuminate in a different colour), or by another local alarm device such as a second light source (not shown), and/or on remote output line 15/17. The HIGH function provides a useful failsafe alarm generation in rare scenarios such as both terminals experiencing connectivity faults at the same time, in which case their temperatures would each rise simultaneously and potentially the IN-OUT function could fail to alarm.

(19) It should be noted that whilst in this embodiment the local processor 11 performs the above-described IN-OUT calculation (and optionally the HIGH calculation) to determine whether a fault exists and to generate an output signal, in other embodiments these functions may be performed at a remote processor (not shown in FIG. 1), in which case the local processor 11 may have lesser functionality, e.g. simply conveying the measured temperatures T.sub.L1-IN and T.sub.L1-OUT to a remote device via remote output line 15/17 for further processing including performing the IN-OUT calculation (and optional HIGH calculation), and generating output signals. This may be preferred where there is a plurality of control sections 5 and corresponding fault monitoring modules 10 as described further below.

(20) Power is preferably provided to the fault monitoring module 10 by a connection 18 to a power supply commonly available on the control section 5, e.g. a 12 or 24V DC supply. If the required voltage is not available directly the fault monitoring module may include a suitable converter. In still further embodiments the fault monitoring module could instead include an onboard power supply such as a battery.

(21) In the FIG. 1 embodiment, the device D uses single phase power and hence there is a single power line L.sub.1 and a neutral line L.sub.n to complete the circuit. It will be appreciated that the neutral line could be connected across the control section 5 in much the same manner as power line L.sub.1, in which case corresponding switchgear 4 may or may not be provided to connect and disconnect the neutral line L.sub.n. However, this is not essential since power supply to device D can be controlled by controlling the switchgear on line L.sub.1 only (given that this will open/close the circuit as a whole). Equivalently, the switchgear 4 could be located on the neutral line L.sub.n only and not on line L.sub.1 with the same effect. Thus, the neutral line L.sub.n can if desired bypass the control section 5 (and compartment 5) entirely, and may be hardwired. If the neutral line L.sub.n is connected across the control section 5, it is preferred that the pair of terminals at which it connects to the control section 5 have their temperatures monitored by additional temperature sensors 12 (not shown) provided in the fault monitoring module 10 so that any connection fault on the neutral line L.sub.n can also be detected using the same principles as discussed above.

(22) Many other devices such as motors require multi-phase power supply and hence multiple power lines each carrying a different power phase. FIG. 2 shows a second embodiment in which the DCC 1 provides first, second and third power lines of different respective phases L.sub.1, L.sub.2 and L.sub.3. It will be noted that no neutral line L.sub.n is present and this is because a neutral line is not essential where the power supply is multi-phase. The control section 5 is connected between each of the three power lines at three respective pairs of terminals: L.sub.1-IN and L.sub.1-OUT, L.sub.2-IN and L.sub.2-OUT, and L.sub.3-IN and L.sub.3-OUT. The fault monitoring module 10 now includes three corresponding pairs of temperature sensors: 12a and 12b, 12c and 12d, and 12e and 12f, arranged to measure the temperatures of the six terminals (in FIG. 2, the cables connecting the temperature sensors 12 to the processor are not shown, for clarity).

(23) The processor 11 performs the above-described IN-OUT calculation for each of the three power lines L.sub.1, L.sub.2 and L.sub.3. Thus, the difference T.sub.L1 between the temperatures of the IN and OUT terminals on the first power line L.sub.1 is calculated and compared with a corresponding threshold T*.sub.L1 the difference T.sub.L2 between the temperatures of the IN and OUT terminals on the second power line L.sub.2 is calculated and compared with a corresponding threshold T*.sub.L2; and the difference T.sub.L3 between the temperatures of the IN and OUT terminals on the third power line L.sub.3 is calculated and compared with a corresponding threshold T*.sub.L3. Again, these predetermined threshold values may be different from one another but typically will be the same (with the result that only one threshold value need be stored and can be used in all three calculations). The processor 11 is configured to generate one or more output signals based on the calculations and to output the signal(s) to an local output module 16 and/or on remote output line 15/17 (each as described in relation to the first embodiment) so as to communicate the status of the terminals externally. The output signal(s) could include a status signal indicating the overall health of the control section 5 which transitions to an alert signal if any one or more of the three IN-OUT calculations indicates a fault. Alternatively the processor may generate three signals, one relating to each power line L.sub.1, L.sub.2, L.sub.3, which may be utilised to control three different alarm devices or to control one alarm device to output different alarms (e.g. colours or flash sequence) depending on which power line is indicating a fault.

(24) Optionally, the processor 11 may again be further programmed to perform the HIGH function described above in relation to the first embodiment. Hence in this case the absolute or ambient-referenced temperatures of each of the six terminals (T.sub.L1-IN, T.sub.L1-OUT, T.sub.L2-IN, T.sub.L2-OUT, . . . etc.) will be compared against corresponding predetermined threshold values (T*.sub.L1-IN, T*.sub.L1-OUT, T*.sub.L2-IN, T*.sub.L2-OUT, . . . etc.) which again may be the same as each other or different. The presence of a fault will be indicated by the measured temperature exceeding the threshold, preferably for a preset time period. The processor 11 will generate one or more output signals accordingly for control of alarm device(s) 16 and/or communication on remote output line 15/17.

(25) Multi-phase power supplies are prone to suffering so-called phase imbalance, in which the power levels on the multiple power lines L.sub.1, L.sub.2, L.sub.3 differ from one another. In devices such as multi-phase motors this can cause substantial damage and significantly shorten the lifetime of the device. In an advantageous embodiment the fault monitoring system can optionally be configured to additionally detect incidents of phase imbalance. Thus in a phase imbalance routine (PHASE), the processor 11 calculates the difference between the measured temperature of each IN terminal and each of the other IN terminals, i.e. calculating T.sub.L1/L2-IN (=T.sub.L1-INT.sub.L2-IN), T.sub.L2/L3-IN (=T.sub.L2-INT.sub.L3-IN) and T.sub.L3/L1-IN (=T.sub.L3-INT.sub.L1-IN), and/or the difference between the measured temperature of each OUT terminal and each of the other OUT terminals, i.e. calculating T.sub.L1/L2-OUT (=T.sub.L1-OUTT.sub.L2-OUT), T.sub.L2/L3-OUT (=T.sub.L2-OUTT.sub.L3-OUT) and T.sub.L3/L1-OUT (=T.sub.L3-OUTT.sub.L1-OUT). Each of the calculated differences is compared against a corresponding pre-determined threshold value T*.sub.L1/L2-IN, etc., which again may be different for each pair of terminals but will more typically all be the same. A calculated difference between the temperature of two IN terminals or two OUT terminals greater than the corresponding threshold is indicative of a phase imbalance existing between the two power lines in question. The processor is configured to generate one or more output signals based on the calculation which are communicated externally via local alarm devices such as 16 and/or on remote output line 15/17. As with the previous functions, the processor may be configured to generate an alarm signal only when the phase imbalance has been indicated for the duration of a preset period of time, e.g. 60 seconds, to avoid false alarming.

(26) The results of the phase imbalance routine could be taken into account in a status signal generated by the processor 11 indicating the overall health of the terminals as mentioned above. In this case the status signal could transition to an alarm signal if any of the IN-OUT, HIGH or PHASE routines identifies a connection fault or phase imbalance. However in preferred examples, multiple output signals are generated such that the output module can indicate to the use not only that there is a fault or phase imbalance but also which of the routines has identified the fault or phase imbalance, and preferably which terminal or power line is indicating the fault or phase imbalance.

(27) In one example, the nature and location of the fault and/or phase imbalance can be indicated locally by implementing the alarm device(s) 16 as a multi-coloured light source (e.g. LED) or as an array of light sources of different colours. For Example, the alarm device 16 could be a three-colour light source capable of emitting e.g. green, yellow and red light (sequentially) under the control of processor 11. The processor may be configured to control the light source 16 to emit green light when there is no alarm condition (i.e. when no fault or phase imbalance has been identified by any of the routines), and to transition to yellow or red light when an alarm condition is active in accordance with the following table:

(28) TABLE-US-00001 Light Source Light Source Colour State Meaning Green Solid All OK (no alarm) Yellow Solid PHASE Alarm Yellow Solid then T Sensor failure coded flashing and location Red Solid then coded IN-OUT fault and flashing location Red Coded flashing HIGH alarm and location Unlit Loss of power to Device

(29) In an example, when the IN-OUT routine identifies a temperature difference (e.g. T.sub.L1) between a pair of sensors 12 on the same cable which exceeds the warning threshold (T*.sub.L1) for more than 60 seconds, the LED 16 changes colour from green to red. The LED will be solid red for say 5 seconds, followed by a number of flashes that will indicate the location of the highest temperature sensor in the relevant sensor pair. For instance, a single flash repeated at uniform intervals may designate terminal L.sub.1-IN, a double flash repeated at uniform intervals may designate terminal L.sub.1-OUT, and so on. When the HIGH route indicates that the temperature of any sensor 12 exceeds the relevant threshold then the LED 16 changes colour from green to red and flashes continuously, e.g. at a different frequency from any of the above-mentioned coded sequences. If the PHASE routine identifies a temperature difference between two IN or two OUT sensors which is greater than the corresponding threshold for more than 60 seconds, the LED 16 is illuminated solid yellow. Optionally this could be followed by a flashing code to indicate which of the power lines the phase imbalance exists between.

(30) As shown in the above table, the LED 16 can optionally also be used to indicate if any of the temperature sensors 12 have failed, which will be detected by the processor 11 as either an open-circuit or a short-circuit. Again, the open-circuit or short-circuit condition may need to be detected for a preset time period e.g. 60 seconds, before an alarm signal will be generated. In this example, the LED 16 is lit solid yellow for 5 seconds followed by a number of coded flashes that will indicate the location of the faulty sensor.

(31) All of the alarms are preferably latching so that if an alarm event disappears (e.g. a measured temperature or calculated temperature differences falls back below the relevant threshold), the alarm will not clear automatically but remains in place until reset by an operator.

(32) It should be noted that both the HIGH and PHASE routines are optional and can be implemented independently of one another. That is, the processor 11 may be configured to perform either, both or neither of the HIGH and PHASE routines alongside the IN-OUT routine.

(33) Other features of the second embodiment indicated using the same reference numerals in FIG. 2 as used in FIG. 1 are the same as in the first embodiment and will therefore not be described again.

(34) The device control centres (DCCs) 1 described so far have been configured for the control of a single device D. However in practice the DCC 1 will typically be configured to control a plurality of devices D via a corresponding plurality of control sections 5, each optionally being housed in a separate case 2 such as a drawer or bucket. FIG. 3 shows a schematic power circuit diagram for such a DCC 1 controlling the supply of power from a power source PS to multiple devices D. The DCC 1 could be, for instance, a motor control centre (MCC), a power distribution unit (PDU), a Power Distribution Cabinet (PDC), a Power Control Cabinet (PCC), a Remote Power Panel (RPP), a fuseboard, or a switchboard. Whilst all the devices D are designated with the same identifier in the Figure, in practice they are likely to include a mixture of different device types, such as motors, lighting circuits, pumps, compressors etc. Each device (or type of device) D may have different power supply requirements and/or control requirements. For example, some of the devices may require single-phase power, others multi-phase, and some may require a return (neutral) conductor line L.sub.n whereas others may not. Thus, whilst all of the devices D are depicted in FIG. 3 as being connected to four power lines L.sub.1, L.sub.2, L.sub.3 and L.sub.4, this may not be the case in practice. Similarly, whilst all of the control sections 5 are identically depicted, in practice the functionality of the control sections may differ depending on the requirements of the device D to which the control section is connected. Some exemplary control sections 5 will be described in more detail below.

(35) Of the fault monitoring system, only the locations of the fault monitoring modules 10 are indicated in FIG. 3 and the other components of the system such as the temperature sensors, processors or output modules (which may form part of the fault monitoring modules) are not shown, for clarity. These can be implemented as discussed above in the first and/or second embodiments in respect of each control section 5, further examples of which will be described below. FIG. 3 also does not depict any communication lines which may optionally be provided between the fault monitoring modules 10 and/or to a remote device or network. Examples of this will be provided below in relation to FIGS. 10 and 11.

(36) Thus, the DCC 1 receives input power from a power supply PS on first, second and third power lines L.sub.1, L.sub.2, L.sub.3 at three different respective phases. A neutral or return line L.sub.n is also provided. The input power is received from the power source PS via an optional main incomer 8 which may include, for example, a switch or circuit breaker. The incoming power is distributed across the DCC along respective power lines provided in a bus 6, and branches 7a, 7b and 7c supply power from the bus to respective columns of control sections 5. For instance, each branch 7a, 7b, 7c may comprise a set of vertical conductors, one for each power line L.sub.1, L.sub.2, L.sub.3, L.sub.n, to which each of the control sections 5 in the column connects at its IN terminals.

(37) Each of branches 7b and 7c is depicted as supplying power to a respective column of control sections 5 which here are contained in individual cases 2, which could be removable drawers or non-removable buckets. Exemplary implementations of each are described below with reference to FIGS. 4 and 5 respectively. On branch 7a, different control sections 5 in cases 3 are depicted, which comprise outgoing feeders such as circuit breakers suitable for supplying devices of varying loads. An example of this is provided below in relation to FIG. 6.

(38) Turning to FIG. 4, a schematic diagram of the contents of a drawer 2 in one embodiment is shown. At the rear of the drawer, a series of connection terminals is provided: four IN terminals for connection to the power supply side of the respective power lines L.sub.1, L.sub.2, L.sub.3 and L.sub.n; and four OUT terminals for connection to the device side of the respective power lines L.sub.1, L.sub.2, L.sub.3 and L.sub.n. Physically, the terminals each comprise a set of spring loaded jaws or similar removable connection solution forming a clamp which can be fitted onto the DCC conductors (IN terminals) and onto conductors connected to the respective device D (OUT terminals). Between the IN and OUT terminals, the four power lines are routed between various switchgear components 5a, 5b and 5c forming control section 5. In this example, component 5a is the main isolator and is mounted to the front surface of the drawer 2. Component 5b is a control device such as a contactor, a Variable Frequency Drive (VFD) or a soft-starter. Component 5c is a unit providing thermal/magnetic overload protection. It will be appreciated that the switchgear components illustrated are merely exemplary and in other cases the control section 5 could comprise any other suitable switchgear, or only one or some of the examples indicated. In this example, the control section 5 also comprises an optional metering device 5d, e.g. for measuring the voltage and/or current carried by at least some of the power lines.

(39) The fault monitoring module in this example comprises a processor 11 which receives measured temperatures from eight temperature sensors 12a, . . . 12h, each of which is arranged to detect the temperature of one of the IN or OUT terminals as shown. The fault monitoring module here also includes a local alarm device 16 in the form of a multi-coloured LED positioned on the front panel of the drawer or bucket 2 for alerting an operator to an alarm event or otherwise communicating the status of the terminals. The processor 11 receives power via a DC connection 18 to the control section 5 (here provided by control device 5b) and optionally outputs output signal(s) via a remote output line 15/17 connected to a communications port also available on the control section 5. This and other remote communication options will be discussed further below with reference to FIGS. 7, 10 and 11. An ambient temperature sensor 19 may optionally be provided as discussed previously.

(40) The processor 11 is configured to perform the IN-OUT function, and optionally the HIGH and/or PHASE functions, on the temperatures measured by sensors 12a to 12h, in the same manner already discussed above with respect to FIGS. 1 and 2. If a fault or phase imbalance event is identified, the LED 16 is controlled by the processor 11 in the manner previously described, and optionally output signals are also conveyed externally on communications line 15/17.

(41) FIG. 5 is a schematic diagram of the contents of a bucked 2 in another embodiment. Here, the bucket 2 is not intended to be removable in everyday use and so the IN and OUT terminals are configured as bolted or terminal style connections onto the vertical conductors (IN) and device conductors (OUT), rather than spring-loaded jaws. This embodiment also differs from that of FIG. 4 in that only three phase power is provided on three lines L.sub.1, L.sub.2 and L.sub.3, and no neutral line is provided (although this may be provided in other bucket style implementations depending on device requirements). The remaining features of the embodiment are the same as already described and the processor 11 is configured to perform the IN-OUT routine and optionally the HIGH and/or PHASE routines as already described.

(42) FIG. 6 schematically shows a circuit breaker 3 such as may be deployed on branch 7a of the DCC shown in FIG. 3. Typically this too will be contained in a case (not shown in FIG. 6). The control section 5 is hard wired between the power supply side of the three power lines L.sub.1, L.sub.2 and L.sub.3 (IN terminals) and the device side of the same three lines (OUT terminals). A fault monitoring module equipped with temperature sensors 12 is provided as before to detect the temperatures of the six terminals and to perform the IN-OUT (and optionally HIGH and/or PHASE) routines as already described to identify the presence of connectivity faults and/or phase imbalance. In this example, only a local output device 16 is provided and no remote output line, but in other implementations such a remote line 15/17 can also be added.

(43) Two exemplary implementations of fault monitoring modules 10 which can be used in any of the above embodiments will now be described with reference to FIGS. 7(a) and 7(b). Both of these exemplary implementations include a local processor 11 which is pre-programmed to perform at least the IN-OUT routine described above and optionally the HIGH and/or PHASE routines. However as mentioned above in other embodiments this level of processing may be performed centrally, e.g. at a computer to which all of the fault monitoring modules 10 are connected. Such implementations will be discussed further below.

(44) In both the FIGS. 7(a) and 7(b) embodiments, the processor 11 is provided on a circuit board or the like, preferably encased in a housing 10a which is equipped with appropriate access points for making connections with each of the components described below. The housing 10a is preferably sized so as to occupy minimal space within the DCC case, e.g. having maximum dimensions of 806020 mm. The housing 10a is preferably designed with a DIN rail mount on its narrow face.

(45) The aforementioned temperature sensors 12 (of which typically there may be 6) are connected to the processor 11 via cables 13 and a connector 20 which will be described further below with reference to FIG. 8. In both embodiments, power supply input terminals 18a/b are provided for connecting the processor to an alternating or direct current power source, e.g. a 12 to 48V DC power source which may typically be available on control section 5 as mentioned above. Alternatively the terminals 18a/b could be replaced by an onboard power source such as a battery. Also provided in both embodiments are output terminals 16a/b for connection to a local alarm device 16 such as the light source, e.g. LED, mentioned in previous embodiments. In practice, the output terminals 16a/b can be controlled by the processor 11 via a relay output.

(46) Both of the embodiments shown in FIGS. 7(a) and (b) provide for a remote output line via which at least one output signal can be communicated to a remote device. However, the nature and extent of this remote communication differs between the embodiments. In the FIG. 7(a) embodiment, the remote output line 15 is another relay output (preferably a dry relay output contact) which can be used to control a remote alarm device (not shown). For instance, this remote alarm device could be a siren or an indicator on a panel, positioned outside the DCC, which is activated by an alarm signal from processor 11 to indicate the occurrence of a connectivity fault or phase imbalance in the particular control section to which fault monitoring module 10 is attached. In another example, the relay output 15 could provide an input to a remote computer which is configured to alert a user to the presence of a fault or phase imbalance by reporting the receipt of an alarm signal on line 15, e.g. by sending a message to a user and/or by indicating a fault on a graphical user interface representative of the DCC. The relay output 15 can be controlled by the processor 11 in the same manner as has been described above in relation to its control of the local alarm device 16, e.g. so as to provide information in the signal as to which of the IN-OUT, HIGH or PHASE routines has indicated the fault or phase imbalance and, optionally as to which terminal has the fault. However, no further information such as the measured or calculated temperature data can be output via relay line 15.

(47) The embodiment of fault monitoring module 10 shown in FIG. 7(b) is equipped with more sophisticated communications functionality, in addition to local alarm line 16 and relay output 15 (or as an alternative). The processor 11 is provided with an output module including a communications device 11a which is configured to convert data output by the processor into a suitable communications protocol, preferably a serial communications protocol such as MODBUS. In practice this functionality may be subsumed into that of the processor 11 itself. One or more data lines 17 are provided for connecting the module 10 to a remote device or network operating on the said data protocol. For instance, in the FIG. 7(b) embodiment, items 17a and 17b are two screw terminals for connection of a MODBUS RS485 daisy-chained network. Of course, any other data protocol could be used but a non-proprietary serial data protocol such as MODBUS is preferred because it will be supported by a wide range of DCC manufacturers across many industries.

(48) The data output on the data lines 17a, 17b and/or 17c could be one or more of the output signal(s) generated by the processor 11 already discussed above, e.g. one or more status signal(s) indicating either an alarm or non-alarm event. Alternatively or in addition, the outputs on lines 17a, 17b and/or 17c could comprise temperature data such as any of: the measured temperatures from sensors 12 (optionally referenced against the ambient temperature), or any temperature differences calculated by the IN-OUT or optional HIGH or PHASE routines. In this way additional information as to the state of the terminals, e.g. the magnitude of any temperature increase, can be obtained thereby allowing more accurate monitoring. Examples of networks to which the module 10 can be connected via lines 17a, 17b and/or 17c will be described below.

(49) The data connection enabled by lines 17a, 17b and/or 17c can optionally also be used to upload data to module 10, e.g. to update the software on processor 11 and/or to set certain parameters such as the various thresholds mentioned above such as T*.sub.L1, etc., used in the IN-OUT routine, T*.sub.L1-IN, etc., used in the HIGH routine, and/or T*.sub.L1/L2-IN, etc., used in the PHASE routine, as well as the optional preset time periods for which an alarm event must be detected before an alarm signal will be output. In such embodiments, these values can be stored in memory in processor 11 and written thereto using known techniques by sending appropriate commands to processor 11 via the selected network protocol.

(50) However, in preferred examples, values such as the above-mentioned thresholds may either be factory-set, or user-selectable from a list of factory-set options. This is particularly advantageous in embodiments such as that of FIG. 7(a) in which there is no data input available, but is still the preferred implementation in embodiments such as that of FIG. 7(b) where such communication is possible, since installation is more straight forward and less prone to user error.

(51) Therefore, in all embodiments, the module 10 is preferably provided with a suitable user input such as a series of switches (e.g. either rotary or DIP switches) 14 which can be used to set the value(s) of selected thresholds for use in the above-mentioned calculations. This enables the user to tailor the fault monitoring system to the particular circumstances of the installation, e.g. to select a higher or lower level of fault sensitivity and/or to account for unusually high or low ambient temperatures. In the examples shown in FIGS. 7(a) and (b), a set of eight DIP switches 14 is provided which can be used by the installer to set the values of the IN-OUT threshold T*.sub.Lm (the same threshold value being used in relation to each power line L.sub.m should more than one be provided), the HIGH threshold T*L.sub.m (the same threshold value being used for every terminal), and the PHASE threshold T*.sub.Lx/Ly-IN (the same threshold value being used for each pair of power lines).

(52) For instance, the first three switches (each of which can be set to a value of either 1 or 0) can be used to select the IN-OUT threshold value from a list of pre-stored options, e.g.:

(53) TABLE-US-00002 Switch Positions T*.sub.Lm 0, 0, 0 10 C. 0, 0, 1 14 C. 0, 1, 0 18 C. 0, 1, 1* 22 C. 1, 0, 0 26 C. 1, 0, 1 30 C. 1, 1, 0 34 C. 1, 1, 1 38 C.

(54) The starred entry 0, 1, 1 (=22 degrees C.) indicates the preferred default setting with which the module 10 may leave the factory.

(55) Similarly, the fourth, fifth and sixth switches of block 14 can be used to select the HIGH threshold value from a list of pre-stored options, e.g.:

(56) TABLE-US-00003 Switch Position T*.sub.Lm 0, 0, 0 55 C. 0, 0, 1 60 C. 0, 1, 0 65 C. 0, 1, 1* 70 C. 1, 0, 0 75 C. 1, 0, 1 80 C. 1, 1, 0 85 C. 1, 1, 1 90 C.

(57) Again, the starred entry 0, 1, 1 (=70 degrees C.) indicates the preferred default factory setting.

(58) The last two switches (seventh and eighth) can be used to select the PHASE threshold value from four prestored options, e.g.:

(59) TABLE-US-00004 Switch Position T*.sub.Lx/Ly-IN 0, 0 10 C. 0.1* 14 C. 1, 0 18 C. 1, 1 22 C.

(60) Where the module 10 is provided with a data output 17 for connection to a network, the network address of the module 10 will also need to be set. If one or more of the above threshold values is to be set by other means (e.g. writing directly to the processor 11 via the communications channels), some of the switches in block 14 could be used for this purpose. Alternatively, another set of switches could be provided for the installer to set the network address. For example, a set of 5 switches can be used to select a network address between 1 and 31, by assigning each switch a value as shown below. The assigned address will be the sum of the switches which the installer sets to 1:

(61) TABLE-US-00005 Switch Modbus slave values 1 1 2 2 3 4 4 8 5 16

(62) In embodiments of this sort equipped with a network communications functionality, the processor 11 will typically be programmed in accordance with the selected protocol (e.g. MODBUS) to hold each item of data in a register assigned accordingly. In one example, the following MODBUS registers could be provided:

(63) TABLE-US-00006 Data Register Type Access Size Description 1 Coil Boolean Read only Bit IN-OUT Alarm 2 Coil Boolean Read only Bit HIGH Alarm 3 Coil Boolean Read only Bit PHASE Alarm 4 Coil Boolean Read only Bit Sensor failure Alarm 5 Coil Boolean Read only Bit Aggregate Alarm 6 Coil Boolean Read only Bit On/Off watchdog 7 Coil Boolean Read/Write Bit Drawer out of service register 30001 Input Integer Read only 16 bit Firmware version 30002 Input Integer Read only 16 bit Serial number - high bits 30003 Input Integer Read only 16 bit Serial number - low bits 30010 Input Read only 16 bit T.sub.L1-IN ( C. 10) Signed Integer 30011 Input Read only 16 bit T.sub.L2-IN ( C. 10) Signed Integer 30012 Input Read only 16 bit T.sub.L3-IN ( C. 10) Signed Integer 30013 Input Read only 16 bit T.sub.L1-OUT ( C. 10) Signed Integer 30014 Input Read only 16 bit T.sub.L2-OUT ( C. 10) Signed Integer 30015 Input Read only 16 bit T.sub.L3-OUT ( C. 10) Signed Integer 30016 Input Read only 16 bit T.sub.AMB ( C. 10) Signed Integer 40001 Holding Read/write 16 bit T*.sub.Lm ( C. 10). Integer 40002 Holding Read/write 16 bit T*.sub.Lm ( C. 10). Integer 40003 Holding Read/write 16 bit T*.sub.Lx/Ly-IN ( C. 10) Integer

(64) Thus, registers 1, 2 and 3 each indicate 1 (alarm) or 0 (no alarm) according to whether the processor has identified an alarm event based on the IN-OUT function, the HIGH function or the PHASE function respectively. Register 4 indicates 1 if a temperature sensor 12 has been detected as faulty (e.g. due to a short circuit or open circuit), and 0 if not. Register 5 is a common alarm register which will set to 0 if no alarm condition exists and to 1 if any one or more of the alarms in registers 1 to 4 exists. Register 6 is a watchdog register which toggles continuously between 1 and 0 at a preset interval to enable an external system to confirm that the processor 11 is operational. Register 7 is provided to enable the user to inform the processor whether the control section 5 to which the module 10 containing processor 11 is fitted is in service (1) or out of service (0). By setting the register to out of service, the functions of the fault monitoring module 10 are disabled such that the control section 5 can be worked on, e.g. disconnected, without triggering an alarm.

(65) Registers 30001 to 30003 are multi-bit registers for storage of data such as the version number of the firmware or software with which the processor 11 is programmed, and a serial number of the particular processor.

(66) Registers 30010 to 30016 are multi-bit registers for each of the temperatures measured by the temperature sensors 12a to 12h and the optional ambient temperature sensor 19 (T.sub.AMB).

(67) Registers 40001 to 40003 are multi-bit registers containing the IN-OUT, HIGH and PHASE threshold values respectively. As described above these may be set via DIP switches 14 local to the processor, or by communication across the network.

(68) In all embodiments, the temperature sensors 12 could be hard wired to the processor 11 or connected via standard terminal blocks, e.g. via a circuit board such as a PCB on which the processor 11 is carried. However, in preferred embodiments a connection assembly 20 is provided via which the temperature sensors can be removably coupled to the rest of module 10. This not only increases the ease of installation but allows for straightforward replacement of the sensors should one become damaged or if a different type of sensor or sensor cable is required. FIGS. 8(a) and (b) show two preferred implementations of such a connection assembly 20. In the FIG. 8(a) embodiment, the temperature sensors 12 are hard wired to a multi-pin connection unit (e.g. plug) 22 and a corresponding multi-pin connection unit (e.g. socket) is provided to couple thereto. In this way, all of the sensors 12 can be connected to the processor in one action. This allows for simple testing and connection to the module 10, as well as: Simple, low cost testing of the sensors 12 prior to shipment; Reduced assembly time in the factory; Shortened installation time and reduced skills required by installer; Improved quality of the connections from the sensors to the processor 11; Straightforward in-service replacement of any damaged sensors 12; and Various sensor cable lengths to be supplied, and/or UL variations for the cable 13.

(69) A variant of this embodiment is shown in FIG. 8(b). Here the connection unit 22 to which the sensors are connected comprises a set of terminal block connectors 22b (one for each sensor), joined to the multi-pin connector 22a. This enables the installer to select e.g. different sensor cable lengths for each of the sensors.

(70) In all embodiments, it is preferred that the sensors be labelled or otherwise coded (e.g. via choice of cable colour) to indicate which terminal each one is to measure the temperature of, since it is critical that the location of each temperature input is known.

(71) In all embodiments, the temperature sensors 12 themselves (and ambient sensor 19 if provided) could take various different forms and could be deployed in different positions depending on the geometry of the installation. However in all cases the temperature of each terminal should be measured as close to the terminal itself as possible. Since it can be difficult to ensure good physical contact between a sensor and a termination (which may have an irregular shape), it may therefore be preferable to utilise non-contact temperature sensors such as radiation thermometers, e.g. infra-red detectors, sited so as to receive radiation emitted by the respective terminals. In other embodiments it is preferred to use contact temperature sensors such as thermistors or thermocouples (if thermocouples are used then an ambient temperature reference 19 is essential). For instance, thermistor sensors of type 10K3A1 resistance values are suitable.

(72) Contact sensors such as thermistors are preferably mounted as close as possible to the respective terminals, e.g. on one of the cables adjacent the termination. For improved measurement accuracy it is desirable to insulate each temperature sensor from the ambient temperature so that its temperature matches that of the terminal or cable to which it is affixed as closely as possible.

(73) Therefore in preferred embodiments each sensor 12 is provided with a cover formed of a thermally insulating material, arranged to reduce heat transfer between the temperature sensor and the ambient environment relative to heat transfer between the temperature sensor and the respective terminal. FIGS. 9(a) and (b) show two examples of suitable covers. In FIG. 9(a) the temperature sensor 12 (here a thermistor) is affixed to the surface of cable C using a thermally insulating adhesive tape 25 which is wrapped around the outside of the sensor 12 and cable C (and preferably does not pass between the cable C and the sensor 12). The material from which the tape 25 is formed preferably comprises a foam or three-dimensional fabric containing pockets of air or other gas to provide increased thermal insulation. FIG. 9(b) shows an alternative embodiment in which a cover 26 is provided in the form of a block formed of an insulating material such as plastic. The block 26 is sized so as to house the sensor 12 in a cavity therewithin. The insulating material from which the block 26 is made extends at least around the sides of the sensor 12 which would otherwise be exposed to the ambient conditions. The block 26 may be open on the underlying surface such that there is no insulating material between sensor 12 and cable C, or could have a thinner wall or membrane here for relatively high thermal transfer. The block 26 can be attached to the cable C either using a mechanical means such as a cable tie (not shown) or by a suitable adhesive 27.

(74) As already described, the DCC will typically include a plurality of control sections 5 and preferably each will be provided with a fault monitoring module 10 of one of the sorts described above. Each fault monitoring module could simply provide a local output such as the alarm device 16 mentioned above. However, more preferably the modules 10 are connected to one or more remote devices so that the occurrence of an alarm event can be notified to persons away from the DCC, or even off-site.

(75) If the modules 10 are provided with relay output lines 15 as described in relation to FIG. 7 above, these outputs can all be connected into a suitable remote device such as a display panel or an input device into a computer network for converting signals on the relay lines 15 into appropriate notifications for users.

(76) Where the modules 10 include communications modules and data outputs 17 (e.g. MODBUS), the modules are preferably each connected to a suitable communications network operating on the selected protocol. An example of such a network is shown in FIG. 10. It should be noted that the power circuitry of the DCC in which the fault monitoring system is deployed (e.g. as shown in FIG. 3) will of course also be present, but is not shown in FIG. 10. Only selected components of the fault monitoring modules 10 are depicted for clarity but in practice each of these can be implemented with any or all of the features described in the embodiments above. Thus, the plurality of fault monitoring modules 10 are connected via data lines 17 to form a network on which also resides at least a controller 31 and preferably one or more user input devices and/or displays such as 32 and 33. Each module 10 communicates with the controller 31 using the selected serial protocol (e.g. MODBUS), outputting data such as alarm status and/or temperature data. The network can also be used to input data to the modules 10 as described previously. Alarms and other data can then be indicated to the users via interfaces 32, 33 on the network.

(77) Networked systems such as that shown in FIG. 10 also lend themselves to alternative implementations in which the above-described processing functions are not performed locally within each module 10 but rather at a remote processor 11, which could for instance form part of controller 31. Thus the local processors 11 in each module 10 could simply convey the temperatures measured by the temperature sensors 12 to the central processor 11 which could then perform the IN-OUT routine (and optionally the HIGH and/or PHASE routines) for each module 10 and generate the necessary output signals for communication to the users via interfaces 32/33 and/or by remotely controlling the alarm devices 16 which may preferably still be located local to each module 10 (via the local processors 11).

(78) An alternative network implementation is shown in FIG. 11. Here, rather than connecting all the modules 10 to the network in series, network aggregators 35 are provided each of which accepts inputs from a subset of the modules 10 and provides communications between that subset and the rest of the network. For example, each unit 35 may be a MODBUS aggregator which accepts up to 32 inputs and provides a simple MODBUS slave device output. The use of aggregators 35 reduces the network traffic and simplifies the installation of the modules 10.