Condition Monitoring of an Object

Abstract

Disclosed is a method for estimating a condition of an object, in particular a cable or a pipeline, and/or material, in particular soil, surrounding the object. The method includes measuring a time course of values of the temperature of an optical fiber using a distributed temperature sensing system employing the optical fiber in proximity to the object, adjusting a value of at least one model parameter of plural model parameters of a thermal model, in particular relating to thermal properties of the object and the surrounding material, based on the measured temperature time course, comparing the adjusted value of at least one model parameter with a value determined theoretically and/or based on physical modeling and/or measured from another information source and/or predetermined and/or based on a value of a load and/or temperature the object is subjected to, to estimate the condition of the object and/or the surrounding material.

Claims

1. A method for estimating a condition of at least one of an object and material surrounding the object, the method comprising: measuring a time course of values of a temperature of an optical fiber using a distributed temperature sensing system employing the optical fiber in proximity to the object; adjusting a value of at least one model parameter of plural model parameters of a thermal model based on the measured temperature time course; comparing the adjusted value of at least one model parameter with a value determined based on at least one of: on theory, on physical modeling, on a measured value from another information source, on a predetermined value, on a value of a load, on a temperature the object is subjected to, to estimate the condition of at least one of the object and the surrounding material.

2. A method for estimating a condition of at least one of an object and material surrounding the object, the method comprising: measuring a time course of values of the temperature of an optical fiber using a distributed temperature sensing system employing the optical fiber in proximity to the object; deriving, using a value of at least one of a load and a temperature the object is subjected to, a time course of temperature values of the optical fiber based on a thermal model having plural model parameters; comparing the measured temperature time course with the derived temperature time course, to estimate the condition of at least one of the object and the surrounding material.

3. The method according to claim 1, wherein adjusting the at least one model parameter includes at least one of the following: performing a least-square optimization; applying a Kalman filter; applying a Wiener filter; applying a Bayes filter; applying a maximum likelihood estimation.

4. The method according to claim 1, wherein the plural model parameters include at least one of the following: temperature of at least a portion of the object; thermal capacity of at least a portion of the object; thermal resistance of at least a portion of the object; thermal capacity of the surrounding material between at least a portion of the object and the optical fiber; thermal resistance of the surrounding material between at least a portion of the object and the optical fiber; a parameter related to thermal loss of at least a portion of the object; heat flux away from or towards at least a portion of the object.

5. The method according to claim 1, wherein values of a first set of model parameters including at least one parameter are held fixed during the adjusting of the model parameters, wherein the values of a first set are determined by at least one of the following: physical modelling of at least a portion of at least one of the object and the environment and the surrounding material and installation properties of at least a portion of the object; weather data; a learning phase in which training time courses of measured temperatures of the fiber are acquired.

6. The method according to claim 5, wherein values of a second set of model parameters including at least one parameter are variable and are optimized during the adjusting of the model parameters.

7. The method according to claim 1, wherein the thermal model represents thermal properties of at least: a heat source; a heat sink; material between at least a portion of the object and the optical fiber; and at least a portion of the material beyond the fiber.

8. The method according to claim 1, wherein the object comprises one of an electric cable and a pipeline, wherein the load comprises one of an electric current and a medium being conveyed within the pipeline.

9. The method according to claim 1, wherein the thermal model applies to plural locations along the object, having individual values of the model parameters in dependence of the respective location.

10. The method according to claim 1, further comprising: repeating the method steps of measuring, adjusting and comparing plural times over time.

11. The method according to claim 1, further comprising: triggering an alarm if at least one model parameter indicates at least one of a deterioration of at least a portion of the object and a heat loss above a threshold and a heat flux above a threshold and an undesired variation in the surrounding material, wherein an alarm is triggered, if at least one of the following applies: a decrease or increase of the temperature of the object is not consistent with model prediction; at least one soil property indicates leakage of liquid or gas from the pipeline; one of release and consumption of latent heat by a product conveyed by the pipeline is indicated; one of release and consumption of heat by a thermodynamic process is indicated.

12. The method according to claim 1, further comprising: suppressing an alarm, if the temperature of the object is above a threshold, but the model parameters do not indicate a deterioration of the object.

13. The method according to claim 1, further comprising at least one of the following: considering measurement of temperature at at least one of an inlet and an outlet of a pipeline; estimating a temperature change along a pipeline; applying a real time transient model; acquiring measurement data from a further fiber installed within a pipeline; acquiring ambient temperature values by further sensors or looping out the fiber into the atmosphere; receiving data from a weather database.

14. The method according to claim 2, wherein the plural model parameters include at least one of the following: temperature of at least a portion of the object; thermal capacity of at least a portion of the object; thermal resistance of at least a portion of the object; thermal capacity of the surrounding material between at least a portion of the object and the optical fiber; thermal resistance of the surrounding material between at least a portion of the object and the optical fiber; a parameter related to thermal loss of at least a portion of the object; heat flux away from or towards at least a portion of the object.

15. The method according to claim 2, wherein values of a first set of model parameters including at least one parameter are held fixed during the adjusting of the model parameters, wherein the values of a first set are determined by at least one of the following: physical modelling of at least a portion of at least one of the object and the environment and the surrounding material and installation properties of at least a portion of the object; weather data; a learning phase in which training time courses of measured temperatures of the fiber are acquired.

16. The method according to claim 15, wherein values of a second set of model parameters including at least one parameter are variable and are optimized during the adjusting of the model parameters.

17. The method according to claim 2, wherein the object comprises one of an electric cable and a pipeline, wherein the load comprises one of an electric current and a medium being conveyed within the pipeline.

18. The method according to claim 2, further comprising: triggering an alarm if at least one model parameter indicates at least one of a deterioration of at least a portion of the object and a heat loss above a threshold and a heat flux above a threshold and an undesired variation in the surrounding material, wherein an alarm is triggered, if at least one of the following applies: a decrease or increase of the temperature of the object is not consistent with model prediction at least one soil property indicates leakage of liquid or gas from the pipeline; one of release and consumption of latent heat by a product conveyed by the pipeline is indicated; one of release and consumption of heat by a thermodynamic process is indicated.

19. An arrangement for estimating a condition of at least one of an object and material surrounding the object, the arrangement comprising: a distributed temperature sensing system employing an optical fiber adapted to measure a time course of values of a temperature of the optical fiber in proximity of the object; a processor adapted to: adjust a value of at least one model parameter of plural model parameters of a thermal model based on the measured temperature time course, and compare the adjusted value of at least one model parameter with a value determined based on at least one of: on theory, on physical modeling, on a measured value from another information source on a predetermined value, on a value of a load, on a temperature the object is subjected to, estimate the condition of at least one the object and the surrounding material.

20. An arrangement for estimating a condition of at least one of an object and material surrounding the object, the arrangement comprising: a distributed temperature sensing system employing an optical fiber adapted to measure a time course of values of a temperature of the optical fiber in proximity of the object; a processor adapted to: derive, using a value of at least one of a load and a temperature the object is subjected to, a time course of temperature values of the optical fiber based on a thermal model having plural model parameters, and compare the measured temperature time course with the derived temperature time course, to estimate the condition of at least one of the object and the surrounding material.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Embodiments of the present invention are now described with reference to the accompanying drawings. The invention is not restricted to the illustrated or described embodiments.

[0040] FIG. 1 illustrates a graph with a number of time courses as considered in embodiments of the present invention;

[0041] FIG. 2 illustrates an equivalent circuit diagram representing a cable and soil portions as employed in embodiments of the present invention;

[0042] FIG. 3 illustrates an equivalent circuit diagram representing a pipeline and the environment as considered in embodiments of the present invention; and

[0043] FIG. 4 schematically illustrates an arrangement for estimating a condition of an object according to an embodiment of the present invention which is configured to carry out a method for estimating a condition of an object according to embodiments of the present invention.

DETAILED DESCRIPTION OF ILLUSTRATED EMBODIMENTS

[0044] The illustrations in the drawings are schematically presented. In different drawings, similar or identical elements are provided with the same reference signs.

[0045] According to embodiments of the present invention, a distributed temperature sensing system may measure the temperature (of an optical fiber) spatially and temporarily resolved. Embodiments of the present invention may be applied for condition estimation of a cable or a pipeline system. The load on the cable system, for example electrical current conveyed through the cable system, may be known or given. This may allow for estimation of the thermal properties of the cable and the parameters involved using a suitable model. One embodiment of the present invention utilizes a thermal model of an electrical cable similar to a rating system and applies it to every location of the cable. This may allow for the optimization and tracking of several parameters of the cable installation using a suitable parameter estimation algorithm (for example one or more of a Kalman filter, a least square optimization, . . . ).

[0046] Embodiments of the present invention may allow for real-time monitoring of a cable state at every location along the cable. Embodiments of the present invention provide mechanisms for automatic alarming on increased losses, cable degradation and/or variations of the installation. Further, embodiments may indicate which of those factors is the cause of the alarm.

[0047] According to one embodiment of the present invention, a thermal model having model parameters is employed and the model parameters are (at least partly) adjusted based on measurement data from a distributed temperature sensing system. Thereby, parameter optimization techniques are used in thermal rating systems to determine a priori uncertain parameters of the system, for example the thermal soil properties. Here, an additional parameter (t) may be tracked which relates the produced heat of an electrical cable to the expected joule-losses:

W.sub.est(t)=(t)W.sub.th(t) Equation (1),

where W.sub.est(t) is the total loss estimated from the measurements and W.sub.th(t) is the estimated loss as proportional to the square of the load or from more advanced physical modeling of the original electrical cable, such as for example according to IEC standard 60287.

[0048] If the cable is working as expected with no damage, (t) may be set to 1. may be considered as a loss factor which, when deviating from 1, measures the deviation of the losses as derived from the temperature time course and the losses estimated from for example the electric current and the electric resistance of the cable, such as that the losses may be given as R.Math.I.sup.2, wherein R is the electrical resistance and I is the electric current. If deviates from 1 (for example after adjusting the model parameter according to the measured temperature time course), it may indicate that the electric cable is somehow damaged. Embodiments of the present invention allow setting an alarm threshold on that property which triggers on a significantly raised value of (t). An increase of a means that the cable has unexpectedly high loss. Alternatively, it may be monitored, that is gradually increasing which may indicate that there is a progressive degradation of the electric cable.

[0049] According to an embodiment of the present invention, the calculation of the model parameters or adjustment of the model parameters (such as the loss factor, ) may be performed for each temperature measurement location (along the optical fiber and thus along the object to be monitored) individually, thereby allowing to localize problematic portions of the electric cable.

[0050] In an analogous approach or manner, also the heat capacity and/or the heat resistance of the soil surrounding the electric cable may be considered as model parameters which may be adjusted during the condition estimating method. Thereby, the heat capacity and heat resistance of the soil may simultaneously be tracked. When also the condition of the surrounding environment or soil is monitored, it may allow discrimination if a hotspot (of a cable) is caused by internal defects of the electric cable (a change of (t), in particular an increase) or external variations (for example change of burial depth, wetting or drying out of soil). It may be advantageous to keep the model as simple and flexible as possible and define model parameters depending on the particular application. For example, if a joint of an electric cable is intended to be monitored, where the sensor fiber might be placed at a more remote location, this may be taken into account. Thus, the distance of the optical fiber from different object portions may be taken into account for condition monitoring. Accurate modeling of a full joint may be difficult due to an inaccurate placement or alignment of the sensor with features of the joint. In this case, a generic model, for example a part representing the heat source, the material up to the sensor and a part representing the soil beyond the sensor, may be used.

[0051] Generally, the model may be set up or defined by applying a learning phase where it may be assumed that the cable is unimpaired and works without any problem, for example directly after installation. During the learning phase, the relevant initial parameters of the model may be determined which may reflect the situation, where the cable is unimpaired and no damage is present. The relevant parameters may not necessarily represent physical values of the losses or the surrounding environment, but may be used as a reference state from which changes and deviations are tracked. In case of significant changes (for example loss or soil thermal resistance increase), the system may be able to detect and monitor the changes.

[0052] FIG. 1 illustrates derived temperature time courses for a number of model parameters as considered in embodiments of the present invention. Thereby, an abscissa 1 denotes the time, while an ordinate 3 denotes the normalized temperature.

[0053] The curve 5 illustrates a measured or derived temperature time course for the situation that an electrical cable is unimpaired and exhibits no damage. The curve 7 represents the measured or derived temperature time course when the loss factor a has increased by 50%, thus, is set to 1.5. As can be observed, the measured or derived time courses 5 and 7 are considerably different. In turn, when the temperature time course 7 is measured, the loss factor a may be adjusted according to an embodiment of the present invention from 1 (curve 5) to the value of 1.5 (curve 7) so that the measured time course matches with a temperature time course derived using the thermal model. Thereby, when measuring the time course 7, it may be concluded that the electric cable is damaged at a particular location.

[0054] The curve 9 corresponds to the case where the environment or soil has 50% increased thermal resistance and the curve 11 corresponds to the case where the environment or the soil surrounding the electric cable has 50% increased thermal capacity. Thus, it can be appreciated from FIG. 1 that changes in different model parameters give rise to different temperature time courses as would be derived or measured using the distributed temperature sensing system.

[0055] In particular, the curves in FIG. 1 show how different variations of relevant model parameters affect the response of the temperature after a step in load (at time=0). One can clearly observe that each of the parameters of interest shows a different effect on the temperature time course. Loss and thermal resistance increase result in a higher steady-state temperature while thermal resistance increase additionally changes the time constant to reach that state. Increased thermal resistance of the installation changes at the time constant but leaves the steady-state temperature unaffected.

[0056] Each of the three model parameters considered above, i.e. the loss factor , the thermal resistance of the soil and the thermal capacity of the soil may have its own signature. A step load is chosen in FIG. 1 for simplicity of the argumentation but with a suitable algorithm (like the previously mentioned Kalman filter) and is not limited to neither a step-like load change nor a steady-state condition afterwards. Instead, also continuous load changes may be covered by embodiment of the present invention. To distinguish the above-mentioned effects, the data needed to be evaluated should cover approximately the time virtually needed to reach the steady-state. This should be around 5 times the time constant of the system which is usually the thermal time constant of the soil. The soil time constant may be typically around 1 week. To achieve a reasonable accuracy of all parameters, in particular the model parameters, one would need a month or longer of data to evaluate.

[0057] The previously described method may require an experienced person to generate a thermal model of the cable and the environment to start with.

[0058] It may be desirable that the effort during realization of a certain estimation is reduced. This may be achieved by a self-learning algorithm which may determine the relevant parameters in a learning phase or in many cases during full operation. This may be reached by extending the model to a non-local version.

[0059] The optical fiber of a distributed fiber optical sensing system allows the acquisition of data (in particular temperature) at multiple locations almost simultaneously. Thereby, embodiments of the present invention model the system under consideration, for example electrical cable or pipe system, in time and also in space. Some of the model parameters may vary with position others do not at all or may vary only very weekly.

[0060] FIG. 2 illustrates an equivalent circuit 13 which may be considered in embodiments of the present invention. The equivalent circuit 13 illustrated in FIG. 2 in particular applies to an electric cable and may be applied or defined for each location individually. Thus, for each location of the electrical cable, the model parameters of the equivalent circuit 13 may be adjusted, in particular optimized, or, when not adjusted, a deviation between a derived temperature time course and a measured temperature time course may indicate that one or more of the model parameters are not accurately reflecting the state or condition of the object. Thereby, locations of for example increased losses or increase soil resistivity, etc. may be found using the method. Thereby, the common knowledge on cable ratings may be used and fixed values for the resistance and capacitance of the cable may be used. It may be very reasonable that the thermal properties of the cable do not change with each location and the thermal properties of the soil do not vary significantly every sample point of the temperature measurement.

[0061] In the equivalent circuit 13 illustrated in FIG. 2, the terminal 15 (T.sub.cond) indicates the temperature of the object, i.e. the temperature of the electric cable. The resistor 17 (R.sub.cable) represents the thermal resistance of the cable. The terminal 19 (T.sub.sens) represents the temperature as measured by the distributed temperature sensing system. The resistors 21 and 23 represent the half of the thermal resistance of the soil between the electrical cable and the optical fiber delivering the DTS signals. The terminal 25 (T.sub.amb) represents the ambient temperature. The capacitor 27 (T.sub.cable) represents the thermal capacity of the electrical cable and the capacitor 39 (C.sub.soil) represents the thermal capacity of the soil. These capacitors are connected in parallel between the upper branch (between terminals 15 and 25) and a lower branch 29 between terminals 31 and 33 representing a lower temperature than the temperature of the object. Between the upper branch 16 and the lower branch 29, further, a thermal loss element 34 is connected which may be identified as W.sub.est (in Equation (1) above) being the estimated heat loss of the object under consideration.

[0062] Some of the model parameters illustrated in FIG. 2 may be considered as variable and some may be considered as fixed, for example fixed may be considered to be C.sub.soil, C.sub.cable and R.sub.cable. Variable may be considered the thermal loss element 34 and the thermal resistance of the soil 21, 23 as well as the ambient temperature 25 and the temperature of the object 15. In other embodiments, other parameters are considered to be fixed and other are considered to be variable and are adjusted.

[0063] Having a model containing equivalent circuits for each location with no pre-calculated values for the parameters but the circuits sharing some of the parameters (like C.sub.cable, R.sub.cable . . . ) which are optimized globally, one can improve the stability and significance of the conclusions drawn from the deduced parameters. And as there is no pre-engineering of the algorithm required one could call this a self-learning system.

[0064] Also the learning phase described above might be ceased as the majority of locations will not have any error, and due to their weight in the calculation will provide good estimations even if there is some location with a fault.

[0065] A similar approach can be used for detection of atypical behavior of pipelines. This can be for example leakages or unwanted state changes of the product inside the pipe (e.g. waxing) which both under certain circumstances affect the temperatures of a fiber optic cable installed in the soil around the pipe. However, as there are other temperature effects caused by the pipeline itself (changing product temperature, variation in transport rate, maintenance periods) or the environment (diurnal/seasonal cycles, burial properties variations) these can mask the temperature effects of the event of interest. A suitable physical model describing the heat transfer from different known heat sources to and from the cable allow for detection of untypical behavior, which allow from the characteristics of the change to deduce information on the source of change.

[0066] FIG. 3 schematically illustrates an equivalent circuit 40 representing a thermal model for monitoring a pipeline. The equivalent circuit model 40 may be similar as set up in thermal rating calculations. Alternatively, a finite element method calculation may be performed. Depending on the design of the thermal model and the external temperature effects, additional measured input values may improve the results of the model. This may be necessary in the case of modeling pipelines:

[0067] Actual and historic local temperature values of the transported fluid inside the pipe may be mandatory which could be determined by [0068] Measurements of the temperature on the inlet and outlet of the pipe combined with a suitable model of the cool-down along the distance of the pipe. This may work well if the pipeline operates in a steady-state temperature condition. [0069] The above point may be improved by using the results of a real-time transient model (RTTM) of the pipeline. [0070] A secondary optical fiber may be installed inside the pipe to determine the actual temperature distribution in situ of the conveyed medium, such as oil or gas.

[0071] Actual and historic temperature of the ambient may be necessary, desirable or mandatory which could be determined by [0072] additional external sensors [0073] segments of the sensors fiber looped out and placed at suitable locations to be exposed to the environment or the ambient [0074] a local weather database may be accessed to obtain approximate values of the ambient temperature.

[0075] In FIG. 3, the terminal 41 (T.sub.product) denotes the temperature of the medium conveyed within the pipe. The resistor 43 (R.sub.pipe) denotes or models the thermal resistance of the pipe. The resistors 45, 47 each model half of the thermal resistance of the (internal) soil between the optical fiber and the pipeline. The terminal 49 (T.sub.sens) denotes the temperature as sensed by the optical fiber using the distributed temperature sensing system. The resistors 51, 53 each denote half of the thermal resistance of the outside soil. The terminal 55 (T.sub.amb) denotes the temperature of the ambient atmosphere.

[0076] The capacitors 57 (C.sub.pipe), 59 (C.sub.soil,in) and 61 (C.sub.soil,out) denote the thermal capacities of the pipeline, the inside and outside soil, respectively. Thereby, soil,in denotes soil portions of a cylinder having same center axis as the pipeline and having a radius as distance of the location of the fiber cable from pipeline center. Soil,out denotes all soil outside of the previously described cylinder.

[0077] The splitting of soil thermal capacity may enable representation of the temperature at the sensor location in the simplified discretized thermal model.

[0078] Embodiments of the present invention allow identification of monitored parameters and control to trigger or suppress alarms satisfying identified criteria.

[0079] For detection of leaks different approaches are possible:

[0080] Leakage of gases experiencing a Joule-Thomson cooling

[0081] If temperature drops are in conflict with the model prediction: Alarming could be achieved by a certain threshold exceeded [0082] 1. If this threshold is adapted according to the current quality of the model and the measured data false alarms can be reduced. A suitable distance measure is for example the Mahalanobis distance in case of usage of Kalman filter or other probabilistic algorithms. [0083] 2. Leakage of liquids having an initial temperature difference to the ambient.

[0084] If temperature drops or rises are in conflict with the model prediction; alarming could be performed as above [0085] 3. Leakage of liquids having no significant temperature difference to the ambient.

[0086] Leakage will not affect the temperature of the sensor directly, but the liquid saturating the pores will affect the properties of the soil. Such leakage will typically increase the thermal conductivity and capacity of the previously unsaturated soil. Thus, monitoring of the thermal parameters of the soil can reveal leakage in this case.

[0087] However, it should be noted that accurately tracking the soil parameters is a difficult task, and such a system may be prone to false alarms. [0088] 4. State changes of the transported product

[0089] States changes may involve releasing or consuming latent heat. If part of the model tracks the heat transfer to environment, deviations of that could be detected as the temperature rises or falls unexpectedly. [0090] 5. State changes affecting the thermal properties of the pipe wall

[0091] This will also effect the heat transfer from the transported product towards the sensor. As for example the heat resistance of the pipeline wall increases with deposed solid material.

[0092] FIG. 4 schematically illustrates an arrangement 70 for estimating a condition of an object 71, in the present case an electrical cable, and also for estimating the condition of material 73 surrounding the object 71, in particular being between the object 71 and an optical fiber 75 which is part of a distributed temperature sensing system 77. The distributed temperature sensing system 77 comprises a light source, a light detector and a processor in the DTS module 79 as is known in the art. The light source is configured to generate light pulses and supply them into the optical fiber 75. Depending on the condition, in particular temperature of the optical fiber 75, the optical properties along the optical fiber 75 may temporarily change which will affect the scattering, in particular Rayleigh scattering of the introduced light. The scattered or reflected light is, time-resolved, detected by the light detector and carries information about the change of the refractive index, spatially resolved. The received light signals are processed and analyzed to derive the temperature at every location along the fiber 75. Thereby, employing the optical fiber 75 and the DTS-module 79, the arrangement 70 is configured for measuring a time course or values of the temperature of the optical fiber in proximity to the object 71 having objection portions 72, 74, e.g. joints, and 76, e.g. cable sections. The optical fiber 75 may or may not be integrated into an insulation of the cable 71.

[0093] Different from a conventional DTS-module 79, the processor 81 is adapted, according to one embodiment, to adjust a value of at least one model parameter (for example of a model as is illustrated in FIG. 2), of plural model parameters of a thermal model and to compare the adjusted value of the at least one model parameter with a value determined theoretically and/or based on physical modeling and/or measured from another information source and/or predetermined and/or based on a value of a load and/or a temperature the object is subjected to, to estimate the condition of the object 71 or/and the surrounding material 73.

[0094] According to another embodiment of the present invention, the processor is adapted to derive, using a value of a load and/or a temperature the object 71 is subjected to, a time course of temperature values of the optical fiber 75 based on a thermal model (for example thermal model 13 illustrated in FIG. 2) having plural model parameters, and to compare the measured temperature time course (for example one of the time courses 5, 7, 9 or 11 illustrated in FIG. 1) with the derived temperature time course, to estimate the condition of the object and/or the surrounding material.

[0095] It should be noted that the term comprising does not exclude other elements or steps and the a or an does not exclude a plurality. Also elements described in association with different embodiments may be combined.

[0096] Implementation of the invention is not limited to the preferred embodiments shown in the figures and described above. Instead, multiple variations are possible which use the described principles according to the invention even in the case of fundamentally different embodiments.