Procedure for the assessment of aging and, in particular, the monitoring of the condition, computer program and computer-readable media

Abstract

The invention relates to a method for evaluating aging and, in particular, for monitoring the condition of an electrical device (1) comprising a fixed insulation system (I) and a liquid and / or gaseous insulating medium (13) in contact with said system (I). In this method, a thermohydraulic aging model (16) is provided, and a simulation of the electrical device (1) is performed. Local temperatures (T1-T5) are calculated for different areas of the electrical device within the simulation, and the quantities of at least one aging product, which forms due to the aging of the fixed insulation system (I) and passes into the insulating medium (13), are calculated for the different areas (8a-8f). Furthermore, the invention relates to a computer program and a computer-readable storage medium.

Description

[0001] The invention relates to a method for assessing the aging and, in particular, monitoring the condition of an electrical device, preferably a transformer or a choke, comprising a fixed insulation arrangement and a liquid and / or gaseous insulating medium in contact with the fixed insulation arrangement. Furthermore, the invention relates to a computer program and a computer-readable medium.

[0002] Insulation systems in electrical devices often consist of a combination of an insulating medium, which can be liquid or gaseous, and a solid insulation layer, typically cellulose-based. Substances that form due to the aging of the solid insulation layer, migrate into the insulating medium, and can be detected there are also considered "aging markers." However, a quantitative assessment of the stress on the most heavily used areas is lacking, as the measured information only provides an integral value. Therefore, its significance is limited to, for example, a traffic light system with threshold values. In particular, a reliable classification of whether a measured quantity of "aging markers" is consistent with normal operation or whether it indicates a suspected malfunction of the electrical device is missing.

[0003] The applicant is aware that attempts are sometimes made to adjust integral measurement results using empirical factors in order to infer an aging state in the warmest or most critical area (so-called "hotspot"). However, this approach can only provide very rough indications at best, since the temperature distribution is subject to various parameters that depend on the design of the electrical device, e.g., a transformer, and which can change during operation.

[0004] CN 112 257 232 A discloses a method for creating a lifetime model based on the ethanol content in oil and the degree of polymerization of transformer paper. Ethanol is used as a characteristic index for the lifetime assessment of the transformer insulating paper. A theoretical basis for representing the transformer's overheating failure and assessing the insulation condition at the winding's hot point, due to the particular sensitivity of ethanol in the oil to high temperatures, is presented.

[0005] The article "Simulation of long-term transformer operation with a dynamic thermal, moisture and aging model," presented at the 5th International Colloquium on Transformer Research and Asset Management, October 2018, Opatija, Croatia (further publication of the article: 5th International Colloquium on Transformer Research and Asset Management, Lecture Notes in Electrical Engineering, vol. 671. Springer, Singapore, 2020, https: / / doi.org / 10.1007 / 978-981-15-5600-5_17), introduces a thermo-hydraulic aging model (THAM) for a transformer. This model is used in a case study to simulate the long-term operation of a transformer whose windings, core, and cooling system are located in a tank filled with oil as a liquid insulating medium. The core, windings, and other transformer components are insulated with a cellulose-based solid insulation system.This assembly comprises pressboard and paper with which conductors are wrapped. The components are modeled with various branches possessing specific properties, connected by nodes, to create a thermo-hydraulic network model. Temperatures and the resulting local humidity values ​​are calculated at different locations within the solid insulation assembly, and local DP values ​​are derived from these calculations by applying an aging formula.

[0006] WO 99 / 60682 A1 specifies a method and arrangement for determining state variables. To determine temperatures in an oil-cooled transformer, the transformer terminal voltages, winding currents, and ambient temperature are measured, and the status of fans and pumps, as well as the position of a tap changer, are recorded. The measured and recorded quantities are fed into a thermohydraulic model. Using auxiliary variables such as transformer losses, heat transfer parameters, flow resistances, and oil flow, and a hydraulic network of the oil circuit comprising branches and nodes, state variables are calculated. These state variables are the mean temperatures and hotspot temperatures in loss-generating transformer components, and the mean oil temperatures in branches and nodes of the hydraulic network of the oil circuit.

[0007] It is an object of the present invention to provide a method of the type mentioned at the outset which enables a more reliable assessment of aging and, in particular, a distinction between normal and faulty operation of an electrical device with an insulation system consisting of a solid insulation arrangement and liquid or gaseous insulating medium, and thus enables particularly safe operation.

[0008] This problem is solved by a method for aging assessment and, in particular, condition monitoring of an electrical device comprising a fixed insulation arrangement and a liquid and / or gaseous insulating medium in contact with the fixed insulation arrangement, preferably a transformer or a choke, in which S1) a thermo-hydraulic aging model of the electrical device is provided and / or created, and a simulation of the electrical device is performed using the aging model; S2) within the framework of the simulation, local temperatures are calculated for different areas of the electrical device, in particular for different areas of the solid insulation arrangement and / or for different areas of the volume occupied by the insulating medium, preferably, taking into account the local temperatures and optionally other aging-determining influencing factors, in particular the local humidity and / or the oxygen content of the insulating medium, local aging parameters, in particular local DP numbers, are calculated for different areas of the solid insulation arrangement; S3) for the different areas of the solid insulation arrangement, taking into account the calculated local temperatures and / or, if calculated in step S2,The quantities of at least one aging product that arises due to the aging of the solid insulation arrangement and migrates into the insulating medium are calculated based on the local aging parameters and preferably taking into account the mass of the solid insulation arrangement, preferably wherein the calculated quantities of aging products are used to distinguish between a normal and a defective state of the electrical device.

[0009] The calculated amounts of aging products are usefully used to distinguish between a normal and a faulty condition of the electrical device.

[0010] It has proven particularly useful to compare simulation results with measured values. In an advantageous embodiment, a comparison is made between the aging product quantities calculated in step S3, in particular a sum of calculated aging product quantities, and one or more aging product quantities measured on the electrical device. Based on the comparison result, a distinction is then made between a normal and a defective condition of the electrical device.

[0011] In other words, the core idea of ​​the present invention is to calculate, using a simulation model, the (location-dependent) generation and distribution of one or more aging products that arise during operation in an insulation system consisting of solid insulation, e.g., paper, and a liquid and / or gaseous insulating medium, e.g., oil, and which, in particular, migrate from the solid insulation into the insulating medium and / or form within the insulating medium itself ("aging markers"). Advantageously, the simulation result(s), or a part thereof, can then be used to determine whether the electrical device is operating incorrectly or normally. The aging product(s) can be measured on a real electrical device in a relatively simple manner, for example, by taking a sample of the insulating medium, e.g., oil.This allows for a comparison between simulation and measured values, and more meaningful and reliable conclusions can be drawn about the condition of the electrical device.

[0012] It is advisable to simulate internal processes within the electrical device. The operation of the electrical device can also be simulated.

[0013] Within the scope of the invention, local temperatures are calculated for various areas or locations of the electrical device, particularly for different parts thereof. Preferably, a network calculation of the temperatures within an electrical device, such as a transformer or an inductor, is performed. The local temperatures are expediently taken into account when optionally determining local aging parameters, for example, by being incorporated into equations used for calculating these parameters.

[0014] In a particularly advantageous embodiment of the method according to the invention, the thermo-hydraulic aging model is designed accordingly as a network model, in particular as a network model for a network calculation of local temperatures at different locations / areas of an electrical device, such as a transformer.

[0015] It is further preferred that the network model is designed to calculate, using or within it, the redistribution of local aging product generation, such as CO2 contributions, to all areas or locations or parts of the fixed insulation arrangement in order to achieve equilibrium in the overall system.

[0016] For example, the local insulation with the highest CO2 production rate does not necessarily have the highest specific CO2 content. In other words, using a network model makes these balancing processes calculable by connecting the local positions of the solid insulation structure with the flowing medium, e.g., insulating oil, which acts as a transport medium for, e.g., CO2. This involves an exchange of, e.g., CO2, to achieve equilibrium between the local insulation structure and the surrounding insulating medium. This exchange can occur in both directions—e.g., from the oil to the cellulose and vice versa—since the storage capacity of the aging product(s), such as CO2, in the solid insulation structure, e.g., cellulose, is generally not negligibly small. From the local values ​​of the at least one aging product dissolved in the insulating medium, e.g., CO2, the local insulation structure can be determined.By using CO2 dissolved in oil, an aging product content can be obtained for the total insulating medium, a value that is measurable. The network model is appropriately designed accordingly.

[0017] Instead of an empirical evaluation, a more systematic assessment of the aging state is carried out, and in particular, the aging state is monitored. The performance of simulation models is enhanced by evaluating individual components instead of making integral statements. The application of a thermo-hydraulic temperature model is expanded. In particular, through aging simulation via the calculation of the local DP number, which generally cannot be verified by direct measurement, the calculation of aging products according to the invention provides a link to the measured values ​​of these aging products in the insulating medium, such as oil.

[0018] It becomes possible to determine whether a measured quantity is consistent with normal operation or whether there is a suspicion of a malfunction in the device. The simulation according to the invention takes into account factors influencing aging processes and their spatial characteristics, in particular the temperature distribution, humidity distribution and the resulting local generation of aging products, as well as optionally the local DP number distribution, especially with associated insulating materials.

[0019] In a preferred embodiment, in step S3, quantities of 2-FAL and / or CO 2 +CO are calculated as aging product quantities using the thermo-hydraulic aging model.

[0020] These quantities represent particularly suitable "aging markers" that can be considered within the framework of the inventive method. Furthermore, these quantities can also be readily measured in an insulating medium. Generally, 2-FAL becomes visible primarily when significant aging has already occurred, while CO₂ + CO₂ is formed from the outset.

[0021] When calculating the quantities of aging products for different locations or areas of the solid insulation assembly, especially different parts thereof, it is expedient to take the masses of the different areas or parts of the insulation assembly into account. This is because the aging products migrating from the solid insulation assembly or parts thereof into the insulating medium will generally depend on the (respective) mass.

[0022] The insulating medium expediently has both insulating and cooling properties; in other words, it serves in particular for both insulation and cooling.

[0023] The thermo-hydraulic aging model used according to the invention is preferably designed as a multi-stage process or can comprise several sub-models. It includes, in particular: The calculation of electrical losses in the various parts of the electrical device (e.g., transformer, choke, etc.) based on the applied voltage and current; the calculation of dynamic local temperatures in different parts of the insulation system comprising the solid insulation arrangement and the liquid or gaseous insulating medium; (optionally) the calculation of the aging of the solid insulation arrangement, if cellulose-based, in particular the cellulose aging, which is typically expressed via the DP number, at different locations / parts of the solid insulation arrangement based on their local temperatures and preferably (further) aging-determining factors, such as oxygen and local humidity; the calculation of the aging products generated as a result of the insulation aging, in particular by the decrease in the DP numbers at different parts of the solid insulation, taking into account their masses.

[0024] The abbreviation "DP" in the DP number stands, as is well known, for "Degree of Polymerization." It is specifically a measure of the decreasing length of cellulose molecules due to thermal aging, which impairs mechanical strength and can lead to operational risks. The DP number indicates the number of basic molecular units in a cellulose chain formed by polymerization, the joining of the basic molecule. New cellulose has DP values ​​above 1000, while chemical decomposition causes the chains to break into shorter units, reducing their mechanical strength. At DP values ​​below 200, mechanical strength, such as tensile strength, is significantly reduced, often falling well below 50% of its initial value. This could jeopardize the safe operation of a transformer, especially in the event of short circuits in the network, which can lead to high currents and thus significant stresses or vibrations, at least briefly.

[0025] For the optional calculation of DP numbers as aging parameters, in particular the calculation of the decreasing DP number due to aging, at least one formula is preferably used, which includes as influencing factors the initial state DP Start, the time duration t, the temperature Θh of the solid insulation, especially cellulose, in the hotspot, and a material property A of the cellulose, which depends on H₂O and O₂ as well as the cellulose quality. For paper, a second cellulose quality is available: a thermostabilized quality with reduced thermal aging at high temperatures. In this regard, reference is also made to the standard IEC 60076-7:2018 "Loading guide for mineral-oil-immersed power transformers", 2018, and in particular to page 42 with equation A.1 and Table A.1 on page 43, whose parameters easily account for temperature, humidity, cellulose quality, and oxygen influence.The thermo-hydraulic aging model used according to the invention can comprise at least one such equation.

[0026] Alternatively or additionally, it may be provided that the following equation is used for the optional calculation of local DP numbers in step S2: D P t = D P t − 1 1 − p T + O t ⋅ M t ⋅ Δ t ⋅ p T − 1 1 1 − p T

[0027] In this equation, Mt is a humidity factor, p(T) is a temperature function, and Ot is an oxygen factor, which strongly influence the DP drop according to the equation.

[0028] The following is preferred for the temperature function: Non-thermostabilized paper and cellulose Thermostabilized paper p ( T ) = 1,182455 · e 0,005791· T < p ( T ) = 1,024508 · e 0,006021· T <

[0029] For the oxygen factor, the following equation is used in a further advantageous embodiment: O T ppm = 1 + O T − 1 ⋅ O ppm / O sat , with O sat = 37000 ppm (< 1,000 m sea level, in mineral oil) and O ppm equal to the oxygen concentration present in the transformer.

[0030] Without measurement, two possibilities can be assumed in particular: open system with oxygen supplied from air, closed system with little oxygen.

[0031] Preferably, the electrical device is designed to be enclosed, in particular by a system or device with an airtight seal.

[0032] The thermo-hydraulic aging model, in its appropriate configuration, is designed to calculate both the steady-state and the transient behavior of the following quantities, particularly based on the load: The heat flow and temperature of components or areas of the electrical device (in the case of a transformer, especially the insulating medium, such as oil, in the upper and lower regions of the tank or coolers, core temperatures, temperatures of winding sections, as well as local hot spots and local oil temperatures in the windings). The oil flow in the electrical device or its components (in the case of a transformer, in the core, the windings, the tank, etc.), which is determined by the hydraulic resistance, the buoyancy force, and optionally the pump pressure.

[0033] Furthermore, the model is preferably designed to calculate the following parameters relating to moisture and aging behavior: The moisture exchange between solid insulation and the liquid or gaseous insulating medium, preferably including any potential moisture exchange with the atmosphere through diffusion of moisture to the outside, in a transformer especially through the oil expansion vessel (conservator). Calculation of the local aging parameter, in particular the degree of polymerization (DP value) of the various areas or cellulose elements, taking into account: ∘ the influence of moisture and oxygen ∘ the quality of the solid insulation, in particular the cellulose quality (e.g.thermally unstabilized or thermally upgraded paper) ∘ the influence of moisture formation due to aging itself; the influence of aging (DP value) on the moisture absorption of solid insulation, especially cellulose, and the risk of blistering if the pressure of dissolved gases and moisture in the solid insulation or insulating medium exceeds the ambient pressure acting upon it. This can lead to the destruction of the device.

[0034] Within the scope of the present invention, an extended version of the thermo-hydraulic aging model disclosed in the article "Simulation of long-term transformer operation with a dynamic thermal, moisture and aging model", 5th International Colloquium on Transformer Research and Asset Management, October 2018, Opatija, Croatia (further publication of the article: 5th International Colloquium on Transformer Research and Asset Management, Lecture Notes in Electrical Engineering, vol. 671, Springer, Singapore, 2020, https: / / doi.org / 10.1007 / 978-981-15-5600-5_17) can be used. This simulation model can be extended to additionally enable the calculation of the aging product quantities according to step S3.includes substances that arise due to the aging of the solid insulation arrangement (2-FAL and / or CO2 + CO2) and pass into the insulating medium and, through propagation through the insulating medium, reach an equilibrium state in the overall system, resulting in a uniform value of the aging product dissolved in the insulating medium, the aging marker (2-FAL, CO2+CO2, ...).

[0035] Input variables for the thermo-hydraulic aging model can include, for example, the ambient temperatures during the operating period under consideration and the load. In a suitable embodiment, the switch position, cooling stage, and / or the applied voltage also constitute input variables for the thermo-hydraulic aging model.

[0036] It may be provided that in step S3, when calculating the quantities of aging products for the various areas of the solid insulation arrangement, the equilibrium state of the respective aging product(s) between the solid insulation arrangement and the insulating medium is taken into account, preferably considering the increase of the respective aging product(s) due to aging and the mixing of the respective aging product(s) in the entire insulating medium. This is particularly important to determine the absorption of the at least one aging product into the insulating medium.

[0037] In other words, the calculation of the aging product quantities can be carried out in local areas or parts of the solid insulation arrangement, where the equilibrium state of the substance or aging product between the local solid insulation arrangement and the local insulating medium is taken into account, considering the increase of the respective aging product(s) due to aging and the mixing of the respective aging product(s) in the entire insulating medium.

[0038] A balancing process can therefore be taken into account, in particular the mixing / redistribution due to transport in the insulating medium. The calculation of the local balancing between the local solid insulation arrangement or areas / locations thereof and the local insulating medium is determined primarily by two effects: the increase of the respective aging product due to aging and the change due to transport in the liquid or gaseous insulating medium. Instead of the iterative change due to transport, a simplification results from an immediate transfer to the entire insulating medium. This means, for example, that the specific aging product value absorbed in the hotspot does not necessarily correspond to the high specific production rate of the aging product. Conversely, areas / locations / parts with low aging receive elevated aging values, more so than their actual aging would suggest.

[0039] An integral value can be derived from the calculation of local aging product equilibria. The aging model is preferably designed accordingly.

[0040] It is advantageous if the equilibrium state is at least roughly known and taken into account accordingly. In the simplest approach, a constant, temperature-independent ratio can be used for the distribution between solid insulation and insulating medium, assuming that the vapor pressure in the solid insulation and insulating medium have the same temperature dependence, so that their ratio is temperature-independent. This simplified assumption does not hold true for the vapor pressure of water dissolved in cellulose and oil. In this case, the moisture shifts into the insulating medium as the temperature increases. The thermo-hydraulic aging model used is, or will be, expediently designed accordingly.

[0041] Alternatively or additionally, when calculating the quantities of aging products in step S3, a redistribution of the aging products from areas with a higher generation rate to areas with a lower generation rate can be taken into account, particularly via transport through the insulating medium. Here too, it is essential that the thermo-hydraulic aging model used is appropriately designed. This effect depends in particular on the specific solubility properties of the respective aging product(s) in the different areas of the insulation arrangement and the insulating medium.

[0042] In a further preferred embodiment, it is provided that in step S3 the calculation of the aging product quantities is carried out using at least one formula that is or was created on the basis of metrologically acquired data, in particular on the basis of metrologically acquired data that link a decreasing DP number with an increasing quantity of at least one aging product, preferably CO₂ + CO and / or 2-FAL. 2-FAL stands for 2-furfural, a compound from the family of cellulose decomposition products that is particularly informative in practice.

[0043] In other words, it can be, in particular, one or more formulas that are or were created using measurement data, which link the aging parameters of a solid insulation, especially DP numbers, with quantities of at least one aging product that migrates into an associated insulating medium. The DP number decreases with increasing age, and the quantity of the aging product(s) increases.

[0044] It should be noted that other aging products, such as methanol, also exist. However, these are generally unstable and decompose further. Using these unstable substances would therefore increase the complexity of the process. In all cases, however, CO₂ + CO₂ is ultimately found. Generally, the proportion of CO₂, which indicates oxygen deficiency during the decomposition process, depends on the decomposition conditions, such as temperature and the rate of temperature rise. Furthermore, CO₂ has a lower solubility in oil than CO₂. For measurement purposes, CO₂ and CO₂ are best measured separately.

[0045] As a purely illustrative example, on a real electrical device, such as a transformer with an oil-filled tank and cellulose-based solid insulation, or on a corresponding representative measurement setup, the DP values ​​of the solid insulation (arrangement) as well as the quantities of aging products transferred, such as CO₂ + CO and / or 2-FAL, in the oil are or have been measured at several intervals. Corresponding measurement data can, for example, be plotted in a graph, and at least one corresponding fit function can be determined, which can then be incorporated into the simulation model as at least one formula. Such measurement data can be acquired specifically for the creation of the simulation model. Alternatively or additionally, existing measurement data can be used.As a purely exemplary example, reference can be made to the article "Diagnosis of Thermal Degradation for Thermally Upgraded Paper in Mineral Oil" by Naoki Yamagata et al., 2008 International Conference on Condition Monitoring and Diagnosis, Beijing, China, April 21-24, 2008, which is e.g. in . Figure 8Figure 1 shows a graph in which the (mean) DP number (as a percentage of the initial value) for two different types of paper is plotted against the amount of CO₂ + CO₂ (ml / g paper) that passes into the oil. Corresponding fit lines are also shown. Thus, laboratory measurement results are available for a correlation between "CO₂ + CO₂ production per gram of paper" and the DP number. These measurements are based on a single mass at a single temperature (single-mass model). This is insufficient for an electrical device such as a transformer. In this embodiment of the present invention, these measurement results are incorporated into a network model to describe the complexity of an electrical device such as a transformer (multi-mass model, balancing processes).

[0046] To correlate the DP number with the 2-FAL recorded in the insulating medium, e.g., oil, a modified version, or at least a modified equation, of the so-called "De Pablo model" can be used. The main equation of this model is given by

[0047] De Pablo: 2 FAL μg / g Papier = 10 6 ∗ DP 0 / DP t − 1 / 162 * DP 0 * 96 * 0 , 3 * 1 , 2

[0048] In this, DP 0 is the initial DP number, DPt is the DP number at any given time during aging, "g paper" is the insulation mass, 162 is the molecular weight of glucose units, 96 is the molecular weight of furfural, 10 6< is the correction factor from g to µg, 0.3 is the reaction yield, and 1.2 is the furfural absorption correction from the solid insulation.

[0049] Regarding the "De Pablo model", reference should also be made to the CIGRE reference paper "The Condition of Solid Transformer Insulation at End-of-Life", CIGRE ELECTRA No. 321, April 2022, by Christoph Krause et al., in particular to page 4 and the equations (1) to (3) therein along with the accompanying explanations.

[0050] In a preferred embodiment of the method according to the invention, the above De Pablo equation (corresponds to equation (3) on page 4 of the aforementioned CIGRE reference paper) is used, however without the factor 1.2 at the end, i.e. 2 FAL μg / g Papier = 10 6 ∗ DP 0 / DP t − 1 / 162 * DP 0 * 96 * 0 , 3 , to calculate quantities of 2-FAL as at least one aging product in step S3. This is because, due to a lack of transformer-specific information, the factor 1.2 represents a general conversion to the DP value in the hotspot.

[0051] The insulating medium is liquid or gaseous. It can consist of or be oil-based, for example, mineral oil. A gas can also serve as the insulating medium. SF6 is just one example of an insulating gas that can be used, considered, or simulated as an insulating medium. It is also possible for a vacuum to serve as, or be simulated as, the (gaseous) insulating medium. For the sake of completeness, it should be noted that even under vacuum conditions, gas transfer and transport to other locations—in other words, a redistribution of aging products—is possible.

[0052] The rigid insulation arrangement can further comprise or consist of cellulose, in particular paper and / or pressboard. Paper, in particular, can be of different grades, e.g., heat-stabilized and non-heat-stabilized.

[0053] For other materials, it is more practical to determine the material-dependent properties using laboratory techniques. For example, with aramid, a high-temperature plastic, the formation of CO2 or other decomposition products that could serve as aging markers is barely perceptible with current knowledge. However, data does exist to estimate the aging process itself based on the decreasing mechanical strength.

[0054] Cellulose and pressboard may also contain other materials mixed in, especially those that ensure a higher thermal insulation class.

[0055] It may also be provided that different areas of the fixed insulation arrangement, in particular different parts thereof, are designed differently, e.g. by different material qualities.

[0056] Instead of the DP number of cellulose, the generation of substances that can also be used as aging markers, in particular aging product quantities, could alternatively be calculated – without prior formation of a DP number – based on time information, especially a time profile, which can depend on several parameters, such as temperature, humidity, etc. [CO2+CO measurement results versus time: see the aforementioned Japanese paper by Naoki Yamagata et al.].

[0057] Aging products, such as CO2 and CO, can also be generated within the insulating medium itself. This proportion is generally significantly lower than the contribution from the solid insulation.

[0058] In a further development of the method according to the invention, the electrical device can comprise a tank filled with the insulating medium, in which components of the electrical device are arranged, wherein one or more of the components are provided with the solid insulation arrangement, in particular wrapped with the solid insulation arrangement or parts thereof. A transformer, for example, typically comprises several components arranged in a tank filled with, for example, oil, each of which is provided with the solid insulation, e.g., pressboard (cellulose) and / or paper. In the case of paper, the components in question are in particular wrapped with it. Cores, windings, and other adjacent parts are mentioned as purely exemplary examples of such components.

[0059] The overall fixed insulation can therefore comprise several parts or partial insulations for the various components, which is why it is also referred to as a fixed insulation arrangement. If the fixed insulation arrangement consists of multiple parts, two or more of the insulation parts can also be spaced apart from each other.

[0060] If the fixed insulation arrangement is multi-part, the different areas of the fixed insulation arrangement in step S2 can comprise or be defined by different parts of the fixed insulation arrangement. Preferably, different parts of the fixed insulation arrangement are assigned to different components of the electrical device. In particular, it is possible that different components of the electrical device are provided with different parts of the fixed insulation arrangement, e.g., also wrapped.

[0061] For example, in step S2, local temperatures and in particular local aging parameters, such as DP numbers, are calculated for various parts of the fixed insulation arrangement, such as for the part of the fixed insulation arrangement located on the windings, for an insulation part of at least one spacer, the insulation part of at least one pressure ring and the insulation part of at least one mounting plate.

[0062] Aging parameters are those that represent or are associated with an aging process of the insulation, and which change, in particular, due to at least one aging process occurring. In other words, it can be a parameter that represents or indicates the progressive aging of the insulation. In the case of cellulose, this is, for example, the decreasing DP number over time, i.e., with increasing age, which has been described in more detail above.

[0063] In step S2, when optionally calculating the aging parameters for the various areas / parts of the solid insulation, other aging-determining factors can be considered in addition to the temperature distribution, particularly the humidity distribution. For example, local temperatures and local humidity values ​​can be calculated for different areas / parts, especially of the solid insulation assembly of the electrical device. Thus, the influence of humidity on aging is also taken into account.

[0064] In step S2, aging parameters can be calculated, taking into account local temperatures and, in particular, humidity. In a further development, several DP values ​​are determined for the various areas / parts as the aging parameters for a given period. In other words, the decreasing DP values ​​with advancing time and thus increasing age of the solid insulation assembly are preferably determined, especially for different areas / parts of the solid insulation assembly. Particularly preferably, in step S2, aging parameters, especially DP values, are calculated for a period of several hours or days, or several months, or even several years.

[0065] It may also be provided that, within the framework of the simulation or the thermo-hydraulic aging model, parts of the solid insulation are simplified and assumed to be blocks.

[0066] As technology advances, the trend towards mixing different insulation qualities is increasing. For example, the use of heat-stabilized papers in winding insulation, with their thermal hotspots, has been state of the art in US industry for over 30 years. However, current assessment methods fail to consider that this type of cellulose exhibits a completely different behavior in 2-FAL generation compared to the large residual mass of non-heat-stabilized cellulose. 2-FAL formation during aging is extremely reduced. The trend towards mixing different materials, such as aramid, is increasing in order to utilize higher thermal classes than those of cellulose – see also IEC 60076-14:2014. In this context, the use of empirical adjustment factors is becoming increasingly disadvantageous. The use of a network model according to the invention to account for locally individual material properties is particularly advantageous here as well.

[0067] Another embodiment of the method according to the invention is characterized in that a simulation of the operation is carried out for a simulation period of several months, in particular several years. In other words, a long-term simulation of the electrical device is performed within the framework of the method according to the invention. It should be noted that the actual computing time ("real time") required for carrying out such a simulation can, of course, be considerably shorter than the time periods covered by the simulation (simulation time), and usually will be considerably shorter.

[0068] It should also be noted that in the case of extreme overloads or extremely rapidly changing loads, a simulation for short periods, e.g., on the order of hours, can also be of great interest. Here, the risk of bubbling, the formation of gas bubbles in the high-voltage range, is particularly high.

[0069] Another object of the present invention is a computer program comprising program code means which, when the program is executed on at least one computer, cause the at least one computer to carry out the steps of the method according to the invention.

[0070] The computer program can be configured to receive one or more values ​​of aging product quantities measured on the electrical device in order to compare these with the aging product quantities calculated in step S3.

[0071] Furthermore, the invention relates to a computer-readable medium comprising instructions which, when executed on at least one computer, cause that at least one computer to carry out the steps of the method according to the invention.

[0072] The computer-readable medium could be, for example, a CD-ROM or DVD, or a USB or flash memory device. It should be noted that a computer-readable medium is not limited to a physical medium, but can also be, for example, a data stream and / or a signal representing a data stream.

[0073] Further features and advantages of the present invention will become clear from the following description with reference to the accompanying drawing. Therein is Figure 1 is a purely schematic partial representation of a transformer, Figure 2 is a purely schematic representation of a thermo-hydraulic aging model for the transformer. Figure 1Figure 3 shows a graph in which the normalized relative aging rate is plotted against the moisture content in cellulose, Figure 4 shows an ambient temperature profile for one year, Figure 5 shows a graph in which local temperatures for different areas of the transformer are plotted. Figure 1 Figure 6 shows a graph plotted over one day, in which the humidity for different areas of the transformer is plotted. Figure 1 Figure 7 shows a graph plotted over one day, in which local DP values ​​for different areas or parts of the fixed insulation arrangement of the transformer are plotted. Figure 1 Figure 8 shows a graph in which DP numbers are plotted against the amount of CO2 + CO2.

[0074] The Figure 1 shows a highly simplified, purely schematic partial representation of an electrical device given by a transformer 1.

[0075] The transformer 1 comprises a tank 2, shown only partially in the figure, in which several components of the transformer 1 are arranged. Of these, the following are shown in the Figure 1The following components can be identified: a core 3, a winding assembly 4, a compression ring 5, a mounting plate 6, and a winding cylinder 7. The transformer 1 has a fixed insulation assembly I, which primarily comprises or consists of pressboard and paper (cellulose) and is designed in multiple parts. The widely used core design of a power transformer was chosen as an example. The active part of this transformer comprises the winding assembly 4 extending around the iron core 3, the conductors of which are insulated by pressboard elements (cellulose), with the conductors typically wrapped in paper (cellulose). The pressboard elements and paper belonging to the winding assembly 4 are integral parts of the fixed insulation assembly I. The same applies to the compression ring 5, the mounting plate 6, and the winding cylinder 7, each of which consists of one or more pressboard elements.

[0076] The block element marked with reference numeral 8 in Figure 1This is representative of other areas or parts of the insulation assembly I that may be present or are present, in particular further cellulose, also for / on further components of the transformer 1 outside the active part. All paper or pressboard parts located on or attached to the various components together form the solid insulation assembly I.

[0077] The number 9 indicates a "void" in tank 2, in other words, a recess or hole. Also shown schematically are a preservative (oil expansion vessel 10), a dehumidifier 11, and a cooler 12. The void 9 lies between the cooler 12 and the remaining area of ​​tank 2. Expansion vessels are available in open designs, allowing air to enter. However, the trend towards closed systems—e.g., using a rubber membrane—is increasing to prevent the entry of oxygen, which accelerates aging.

[0078] In the illustrated embodiment, tank 2 is further filled with a liquid insulating medium, which is a mineral oil 13. This can also be referred to as an oil-insulated transformer 1. The solid insulation assembly I and the oil 13 together form an insulation system for the transformer 1.

[0079] The solid insulation assembly I is in contact with the insulating oil 13. Substances that arise due to the aging of the solid insulation assembly I, migrate into the insulating medium 13, and can be detected there are also referred to as "aging markers." CO₂ + CO₂ and / or 2-FAL are examples of such "aging markers."

[0080] The oil flow is in Figure 1 indicated by simple arrows, which are exemplified by the reference number 14.

[0081] Double arrows 15 further indicate, purely schematically, the moisture exchange between solid insulation arrangement I and liquid insulating medium 13.

[0082] By carrying out the embodiment of the inventive method described below, an improved assessment of the aging state and thus better condition monitoring of the operation of the transformer 1 is possible.

[0083] In step S1, a thermo-hydraulic aging model (THAM for short) 16 (see Figure 2 The aging model 16 is provided and / or created for transformer 1, and a simulation is performed for transformer 1, in particular for a simulation period of several months or years. Internal processes and procedures within transformer 1 are simulated. The simulation can be carried out in a manner known per se using at least one computer.

[0084] An extended version of the thermo-hydraulic aging model presented in the article "Simulation of long-term transformer operation with a dynamic thermal, moisture and aging model", 5th International Colloquium on Transformer Research and Asset Management, October 2018, Opatija, Croatia, is used here.

[0085] The various components or aspects of the in Figure 1 The depicted transformer 1, including the oil flow, is modeled by various branches with specific properties, connected by nodes, to obtain a thermal network model. Figure 2An exemplary arrangement of branches and nodes for the main components of transformer 1 is shown. Nodes 17-20 represent the core 3, nodes 21-23 the bulk of the oil 13 in tank 2 ("tank bulk oil"), nodes 21, 24, and 23 represent the wall of tank 2, node 25 represents the confluence of the oil 13 from the components below, node 26 the oil 13 in the upper part of tank 2 ("top oil" - the resulting hot oil) and the membrane of the conservator 10, node 23 the oil 13 in the upper part of tank 2 including the seals, nodes 27-38 the transformer windings, and nodes 39-41 the cooling element. Nodes 42 to 44 show an exemplary design of the oil supply to the windings of the winding assembly 4. The branches are arranged in Figure 2 indicated by arrows connecting nodes 17-44.

[0086] Within the framework of the simulation with the aging model 16, local temperatures are calculated in step S2 for various areas of the electrical device 1, here for various parts of the solid insulation arrangement I and in particular areas of the oil 13, which adjust themselves dynamically with a time delay due to the electrical losses generated from current and voltage. Figure 1 It is indicated that the press ring 5 has the temperature T TO of the oil 13 in the upper area of ​​the tank 2 ("Topoil"), the winding arrangement 4 with the windings and pressboard spacers has the hotspot temperature TH and the mounting plate 6 has the temperature T BO of the oil 13 in the lower area ("Bottom Oil").

[0087] In step S2, local aging parameters, specifically local DP values, are calculated for various areas of the fixed insulation assembly I, specifically for different insulation components. This calculation takes into account the determined local temperatures and other aging-influencing factors, namely oxygen and local humidity. The calculation of local aging parameters included in this embodiment is optional.

[0088] In step S3, quantities of at least one aging product, which arises due to the aging of the solid insulation arrangement I and passes into the insulating medium 13, are calculated for the various areas / parts of the insulation arrangement I, taking into account the local DP numbers obtained in step S2, and in particular the masses of the insulation arrangement I.

[0089] The thermo-hydraulic aging model 16 is designed to calculate the following parameters with respect to moisture and aging behavior: The moisture exchange between the solid insulation assembly I and the insulating medium 13, including any potential moisture exchange with the atmosphere. Calculation of local aging parameters, in particular the degree of polymerization (DP value), of different areas / parts of the cellulose insulation I, especially the hotspot, taking into account: ∘ the influence of moisture and oxygen ∘ the cellulose quality (e.g., non-thermostabilized or thermostabilized paper) ∘ the moisture formation due to the aging process itself. Influence of aging (DP value) on the moisture absorption of the cellulose, and the risk of blistering when the pressure of dissolved gases and moisture in the insulating medium 13 exceeds the ambient pressure acting upon it. This pressure or temperature should also be slightly increased to overcome the interfacial tension of the liquid and thus enable the formation of a gas bubble in the insulating oil 13.

[0090] Aging leads to a deterioration in the mechanical strength of the cellulose or the tensile strength of the paper. A decrease in tensile strength to below 40% of the initial value is often considered a risky condition with regard to short-circuit strength. Instead of the measured tensile strength, the average length of the cellulose fibers, the degree of polymerization (DP value or DP number), is a common parameter for describing the condition of the cellulose. Tensile strengths of approximately 40% correlate with DP values ​​of about 200. New transformers typically have DP numbers above 900. The aging model 16 calculates the decrease in the DP value in each modeled solid insulating part 4-8 of the transformer 1 according to Figure 1 .

[0091] The following equation is preferably used to calculate the local DP numbers: D P t = D P t − 1 1 − p T + O t ⋅ M t ⋅ Δ t ⋅ p T − 1 1 1 − p T

[0092] In this equation, Mt is a humidity factor, p(T) is a temperature function, and Ot is an oxygen factor, which strongly influence the DP drop according to the equation.

[0093] The following applies to the temperature function: Non-thermostabilized paper and cellulose thermally stabilized paper p ( T ) = 1,182455 · e 0,005791· T < p ( T ) = 1, 024508 · e 0,006021· T <

[0094] The following equation is used for the oxygen factor: O T ppm = 1 + O T − 1 ⋅ O ppm / O sat , with O sat = 37000 ppm (< 1,000 m sea level, in mineral oil) and O ppm equal to the oxygen concentration present in the transformer.

[0095] The graph according to Figure 3Furthermore, the graph shows the influence of humidity on the relative aging rate. The graph plots the normalized relative aging rate (NRAR - Mt factor normalized to 1 at 0.5% humidity) against the humidity in cellulose WC in %. The solid line corresponds to values ​​calculated using aging model 16, while the crosses and dots represent values ​​according to IEC 60076-7:2018, with the crosses for non-thermostabilized paper and the dots for thermostabilized paper. As can be seen, the normalized relative aging rate increases with increasing humidity of the solid insulation assembly I, in this case cellulose.

[0096] The graph from Figure 4 This figure shows the ambient temperature profile used or assumed in the simulation for one year. The same ambient temperature profile, shown in this figure, was used for each simulation year.

[0097] Additionally, the load, voltage, and switch position are usefully taken into account. If multiple cooling stages are present, these can be defined (e.g., oil pumps, fans). The graph according to Figure 5 Figure 16 shows temperatures calculated using the aging model for various parts of the insulation arrangement I and areas of the oil 13. The temperatures are plotted over time, with the x-axis representing an exemplary day, specifically the hottest day of the year, which occurs in Figure 4 is marked.

[0098] At the local temperatures from Figure 5 It concerns: T1 the temperature of the paper in the hot spot, T2 the temperature of the spacers at the hot spot, T3 the temperature of the oil 13 at the top of the winding pressure ring, T4 the temperature of the hot oil in the cooler 12 (top), T5 the temperature of the cooled oil 13 (bottom) at the inlet under the winding assembly 4, TU the ambient temperature

[0099] Figure 6shows the calculated local humidity level for different areas / parts for the exemplary hottest day, namely: M1 the moisture content of the paper wrapping in the hotspot M2 the moisture content of the spacers at the hotspot M3 the moisture content of the winding pressure ring 5 (above the windings) M4 the moisture content of the winding mounting plate 6 (below the windings) M5 the water activity at cold oil temperature T_bottom M6 the moisture content in the mineral oil 13 in ppm

[0100] It should be noted that the scale of the Y-axis on the left shows the moisture content of cellulose in % as well as the moisture content of oil expressed as water activity (aw) in % - that is the relative moisture vapor pressure, which refers to the vapor pressure of water at the same temperature, while that of the right Y-axis shows the moisture content of oil in ppm.

[0101] Within the simulation, the aging model 16 is used to calculate corresponding values ​​for each day, as described in the Figures 5 and 6 The applied values ​​were calculated.

[0102] The Figure 7 Step S2 shows the local DP numbers DP1-DP5 calculated in the aging model 16 (optional) for the various fixed insulation parts 4-8 of the transformer 1. Figure 1 Plotted over time in years. As you can see, a simulation period of 50 years has been chosen here, although this is purely for illustrative purposes. In other words, the decrease in the local DP numbers DP1-DP5 is simulated over a period of 50 years. A notable feature of this example is that the use of thermostabilized paper in the windings can result in less aging of these windings than other insulating components.

[0103] The local DP figures from Figure 7 Specifically, it concerns: DP1DP values ​​of the paper wrapping in the hotspot; DP2DP values ​​of the spacer at the hotspot; DP3DP values ​​of the paper wrapping, averaged over a winding; DP4DP value in the winding pressure ring 5 (above the windings); DP5DP value in the winding mounting plate 6 (below the windings); DPav Average value of all local DP values ​​in the network model

[0104] It should be noted that instead of the DP numbers of cellulose, the generation of substances that can also be used as aging markers can alternatively be calculated – without prior formation or calculation of the DP numbers – based on a time profile that depends on several parameters, such as temperature, humidity, etc. [CO2+CO measurement results versus time: see the aforementioned Japanese paper by Naoki Yamagata et al.].

[0105] As noted above, in step S3 – in the embodiment described here, taking into account the local aging parameters, here local DP numbers DP1-DP5 – quantities of at least one aging product that arises due to the aging of the solid insulation arrangement I and transfers into the insulating medium are calculated. Specifically, quantities of CO₂ + CO₂ and / or quantities of 2-FAL are calculated here, which is to be understood as an example.

[0106] The calculation of the quantities of CO₂ + CO₂ is carried out using at least one formula that was created on the basis of metrologically recorded data, in this case on the basis of metrologically recorded data that link a decreasing DP number with an increasing quantity of CO₂ + CO₂. Figure 8 The figure shows, purely as an example, a corresponding graph with relevant pairs of values ​​and a corresponding fit function F. It should be emphasized that the Figure 8This is purely schematic and serves only for illustration. The x-axis is unscaled and unitless and can – as in the present case – also be logarithmic, for example.

[0107] For example, one can refer to specific measurement results such as those published in the article "Diagnosis of Thermal Degradation for Thermally Upgraded Paper in Mineral Oil" by Naoki Yamagata et al., 2008 International Conference on Condition Monitoring and Diagnosis, Beijing, China, April 21-24, 2008, for instance. Figure 8 from this article. Alternatively or additionally, measurements on experimental setups and / or measurements on real electrical devices can be carried out to obtain data for a correlation of DP numbers and amounts of CO₂ + CO₂.

[0108] To alternatively or additionally calculate quantities of 2-FAL, the equation is preferably used. 2 FAL μg / g Papier = 10 6 ∗ DP 0 / DP t − 1 / 162 * DP 0 * 96 * 0 , 3 , used.

[0109] When calculating the amounts of aging products, their equilibrium state between solid insulation arrangement I and insulating medium 13 is taken into account, in particular to determine the absorption of the respective aging product(s), here CO 2 +CO and / or 2-FAL, into the insulating medium 13.

[0110] Furthermore, a redistribution of the respective aging product(s), here CO2 + CO2 and / or 2-FAL, from locations with a higher generation rate to locations with a lower generation rate via transport through the insulating medium 13 is taken into account.

[0111] For this purpose, a gas pressure pA,oil is assigned to the aging product "A" dissolved in the oil, gA / g oil (g...grams) or µlA / l oil (l...liters), which is proportional to the concentration of the aging product gA / g oil or µlA / l oil. In cellulose, a concentration of the aging product in the cellulose is assigned to the gas pressure, gA / g cellulose = Fcellulose · pA,oil. This factor Fcellulose should be adjusted to the measurement results obtained from the network model if laboratory results on the absorption capacity of the aging product in the cellulose are lacking. In the first step, the temperature dependence of the absorption properties is neglected, as it is significantly lower than with humidity.In network model 16, each iterative calculation step first calculates the additional amount of aging product generated by aging and adds it to the previous value of the aging product dissolved in cellulose. Then, in a second step, due to the transfer into the oil, the amount of aging product gA is redistributed between each individual insulation node and the surrounding oil until the pressures in solid pA,cellulose and liquid / gaseous insulation pA,oil are equal in each insulation element of the network model. The oil flow, which is taken into account in the iterative calculation, leads to a gradual mixing of the aging product dissolved in the oil.Given the low temperature dependence of the cellulose factor F and the slow increase in the amount of aging products compared to mixing in the flowing oil, a simplification is permissible: the individual amounts of aging products dissolved in the oil are immediately and homogeneously distributed throughout the entire oil. Following this, as described above in the second step, equilibrium is reached by shifting the amount of aging products with the individual insulation components (nodes). This is a dynamic process, as the individual local parameters, such as temperature, are usually in flux. Should a temporary, balanced steady state occur, the values ​​in all individual local oil volumes are identical, and the exchange between the solid insulation arrangement and, in particular, the liquid insulating medium, comes to a standstill, at least briefly, before the increase in the aging product in the cellulose becomes noticeable in the oil as described above.

[0112] As noted above, the thermo-hydraulic aging model 16 used in the embodiment described here is an extended version of the model from the article "Simulation of long-term transformer operation with a dynamic thermal, moisture and aging model", 5th International Colloquium on Transformer Research and Asset Management, October 2018, Opatija, Croatia. The extension concerns the calculation of the quantities of CO₂ + CO₂ and / or 2-FAL in the manner described above. The additional generation of the aging marker is performed for each individual point in the network model; in a second step, a redistribution via the equalization processes in the oil 13 takes place. In principle, the focus is on evaluating the parts subjected to the highest thermal stress, the so-called hotspots in the insulation arrangement I. However, the aging markers at the hotspots interact through equalization processes and redistributions.The Model 16 is trained accordingly.

[0113] The quantities of CO₂ + CO and / or 2-FAL calculated via the simulation are then preferably used to distinguish between a normal and a faulty state of transformer 1. For this purpose, the calculated quantities of CO₂ + CO and / or 2-FAL are summed to obtain a total simulation value. Furthermore, corresponding measurement results from a real transformer, for which transformer 1 is used, are obtained. Figure 1 , on which the simulation is based, preferably represents a digital twin, provided or received by a computer program used to carry out the simulation.

[0114] In particular, a comparison between the simulation result(s) and the measured value(s) allows a decision to be made as to whether transformer 1 is in a normal or faulty condition. In the latter case, a warning can be issued. Even if no measurement of an "aging marker" such as 2-FAL is performed, i.e., no corresponding measurement results are available for such a comparison, an assessment of the aging state is still carried out, since model 16 provides information about critical aging in the hotspot. This supports the lifetime analysis and thus the timely planning of new investments.

[0115] Although the invention has been illustrated and described in detail by the preferred embodiment, the invention is not limited by the disclosed examples and other variations can be derived by the person skilled in the art without leaving the scope of protection of the invention.

[0116] Regardless of the grammatical gender of a particular term, persons with male, female or other gender identities are included.

Claims

1. A computer-implemented method for assessing ageing and for monitoring a condition of an electrical device (1) comprising a solid insulation arrangement (I) and a liquid and / or gaseous insulation medium (13) that is contact with the solid insulation arrangement (I), wherein the electrical device is a transformer or a choke, in which S1) a thermo-hydraulic ageing model (16) of the electrical device (1) is provided and / or created and a simulation for the electrical device (1) is performed with the aging model (16), S2) local temperatures (T1-T5) are calculated for different regions of the electrical device for different regions of the solid insulation arrangement (I) and / or for different regions of the volume occupied by the insulation medium (13) as part of the simulation, wherein local ageing variables (DP1-DP5) are calculated for different regions of the solid insulation arrangement (I), taking into account the local temperatures and further influencing variables that determine ageing, in particular local humidity and / or oxygen content of the insulation medium (13), S3) quantities of at least one ageing product that originates due to the ageing of the solid insulation arrangement (I) and passes into the insulation medium (13) are calculated for the different regions of the solid insulation arrangement (I), taking into account the calculated local temperatures and / or, if calculated in step S2, the local ageing variables (DP1-DP5) and additionally taking into account masses of the solid insulation arrangement (I), wherein the calculated ageing product quantities are used to differentiate between a normal and a faulty condition of the electrical device (1).

2. The method according to claim 1, characterised in that a comparison of the ageing product quantities calculated in step S3 with one or more ageing product quantities metrologically acquired at the electrical device (1) takes place and a normal and a faulty condition of the electrical device (1) are differentiated based on the comparison result.

3. The method according to any one of claims 1 or 2, characterized in that the thermo-hydraulic ageing model (16) is configured as a network model.

4. The method according to any one of the preceding claims, characterised in that in step S3, quantities of 2-FAL and / or CO2+CO are calculated as ageing product quantities.

5. The method according to any one of the preceding claims, characterised in that in step S3, when calculating the ageing product quantities for the different regions of the solid insulation arrangement (I), the equilibrium state of the or of the respective ageing product between the solid insulation arrangement (I) and the insulating medium (13) is taken into account in each case, taking into account the increase of the or of the respective ageing product due to ageing and the mixing of the or the respective aging product in the insulating medium (13) in order to determine the absorption of the at least one ageing product into the insulating medium (13).

6. The method according to any one of the preceding claims, characterised in that in step S3, when calculating the ageing product quantities, a redistribution of the at least one ageing product from locations having a higher generation rate to locations having a lower generation rate via transport through the insulating medium (13) is taken into account.

7. The method according to any one of the preceding claims, characterised in that in step S3, the calculation of the ageing product quantities is carried out using at least one formula which is or has been created based on metrologically acquired data.

8. The method according to any one of the preceding claims, characterised in that in step S3, quantities of 2-FAL are calculated as ageing product quantities and the formula 2 FAL μg / g paper = 10 6 ∗ DP 0 / DP t − 1 / 162 * DP 0 * 96 * 0.3 , is used for the calculation, wherein DP0 is the initial DP number, DPt is the DP number at any time during ageing, and "g paper" is the mass of the solid insulation arrangement (I) or of the respective part of the solid insulation arrangement (I), wherein the fixed insulation arrangement (I) comprises paper and / or pressboard or is provided by paper and / or pressboard.

9. The method according to any one of the preceding claims, characterised in that the insulation medium (13) comprises an oil or is provided by an oil, and / or that the solid insulation arrangement (I) comprises cellulose or is provided by cellulose.

10. The method according to any one of the preceding claims, characterised in that the electrical device (1) comprises a tank (2) filled with the insulating medium (13), in which components of the electrical device (1) are arranged, wherein one or more of the components are provided with the solid insulation arrangement (I).

11. The method according to any one of the preceding claims, characterised in that wherein the solid insulation arrangement (I) is configured in multiple parts and, in step S2, the different regions of the solid insulation arrangement (I) comprise different parts of the solid insulation arrangement (I) or are provided thereby, wherein different components of the electrical device (1) are assigned different parts of the solid insulation arrangement (I), wherein different components of the electrical device (1) are provided with different parts of the fixed insulation arrangement (I).

12. The method according to any one of the preceding claims, characterised in that a simulation of the operation is carried out for a simulation period of several hours, months or years.

13. A computer program comprising program code means which, upon execution of the program on at least one computer, cause the at least one computer to perform the steps of the method according to any one of claims 1 to 12.

14. A computer-readable medium comprising instructions which, when being executed on at least one computer, cause the at least one computer to perform the steps of the method according to any one of claims 1 to 12.

Images