Hitachi Energy Patente

There is disclosed herein a computer-implemented method (300) for monitoring an static induction device (3), comprising obtaining time-series data (308) comprising monitoring data (307) for the static induction device (3), and decomposing the time-series data (308) into a plurality of layers (310), each layer (310) comprising a subset of the time-series data (308), wherein at least one layer (310) is configured to contain a respective contribution (420) to the monitoring data (308) from a contributor (418). The method (300) further comprises providing each layer (310), as input, to a respective predictive model (312) of a corresponding plurality of predictive models (312) configured to predict future contributions to the monitoring data (310) from the respective contributor (418), and receiving, as output from the plurality of predictive models (312), respective predictions (314) of future contributions from the contributor (418) of each layer (310). Finally, the method (300) comprises aggregating the predictions (314) from each predictive model (312) to thereby obtain predicted monitoring data (316) for the static induction device (3). In this way, the accuracy of the predictions for how the static induction device monitoring data will change is greatly enhanced, enabling a more pro-active control for a power network.

The invention relates to a system (20) for a gas-insulated switchgear (2), comprisinga housing (30) with a flange (32) to be coupled to an electrical component (3), e.g. a disconnector/earthing switch, of the gas-insulated switchgear (2);a low-power instrument transformer (40) arranged in the housing (30) and configured for conducting current and voltage measurements; anda cable end module (50) arranged in the housing (30) and configured for receiving at least one power line (100) to be electrically connected to the electrical component (3) through the flange (32).

The present disclosure relates to a power supporting system (1) comprising an energy storage system (100), ESS, comprising at least one string (S-1, S-M). Each string (S-1, S-M) comprises at least two energy storage racks electrically connected in series. Each energy storage rack comprises at least one energy storage module (ESM-1, ESM-N). Each energy storage module (ESM-1, ESM-N) comprises at least one energy storage cell (C-1, C-2, C-X), and a cell management system (xMS-1, xMS-N) configured to manage said at least one energy storage cell. The power supporting system (1) comprises at least one control device (10, 10-1, 10-M), and at least one primary communication unit (20-1, 20-M), PCU, each PCU (20-1, 20-M) configured to provide a dedicated communication channel between at least one of said at least one control device (10, 10-1, 10-M) and CMSs (xMS-1, xMS-N) of one of said strings (S-1, S-M). Said at least one control device (10, 10-1, 10-M) and said at least one PCU (20-1, 20-M) are arranged remotely relative to said ESS (100). Said at least one control device (10, 10-1, 10-M) is configured to control, by communication via said dedicated communication channels, a remote firmware loading process to update said CMSs (xMS-1, xMS-N) of said at least one string (S-1, S-M). A method of updating CMSs of a power supporting system (1) is also disclosed.

The invention relates to an operating mechanism for a switching device (601), comprising a motor (1), an axially extending screw rod (701) and a spindle nut (702), whereby the screw rod (701) is connected to the motor (1) and configured for being rotated by the motor (1), the spindle nut (702) is movable engaged with the screw rod (701) and configured to move linearly along the screw rod (701) when the screw rod (701) rotates, and the screw rod (701) comprises a chamfered thread tip (706) insertable into the spindle nut (702), whereby the chamfered thread tip (706) comprises a first chamfered surface (708) extending in a first angle (α) relative to the axial extension of the screw rod (701) and a second chamfered surface (709) extending in a second angle (β) relative to the axial extension of the screw rod (701) and greater than the first angle (α).

There is disclosed herein a protection enhancement device (110) for a power network (200), comprising a control module (112) configured to, in response to a detection of a fault current at a detection location in the power grid, control a power device (120) to inject an additional current to supplement the fault current in the power network to thereby cause a protection response on the power network, and/or control a protection device (120) to initiate a protection response on the power network (200). In the latter case, the detected fault current has a different value to a protection response configuration setting for the protection device at which a protection response of the protection device (120) is caused. Also disclosed herein are associated protection methods performable by devices, and a protection system comprising such devices.

The present disclosure relates to a control device (20) configured for grid-forming control using a voltage source converter, VSC, (10) electrically connected to a power grid (40) via a point-of-common-coupling, PCC, the control device (20) comprising: a decoupling controller (21) configured to determine a VSC virtual electromotive force (v<EMF>) of said VSC (10) based at least on a first complex signal (K; γ, δ), which first complex signal (K; γ, δ) is based at least on a complex power (S<g>; P<g>, Q<g>) being exchanged between said VSC (10) and the power grid (40), and a complex power reference (S<ref>, P<ref>, Q<ref>) for said complex power (S<g>; P<g>, Q<g>), and a current controller (22) configured to determine a VSC voltage reference (v<c,ref>) for said VSC (10) based at least on a grid voltage (v<g>) for said power grid (40) at said PCC, and a current reference (i<f,ref>) for a current (if) propagating between said VSC (10) and said PCC, a virtual admittance unit (24) configured to determine said current reference (i<f,ref>) based at least on a voltage input (v<in>) multiplied with a virtual admittance (Y<v>), wherein said voltage input (v<in>) is based at least on a voltage difference between said VSC virtual electromotive force (v<EMF>) and said grid voltage (v<g>), and said virtual admittance (Y<v>) is based at least on a virtual impedance between the VSC virtual electromotive force (v<EMF>) and said PCC, wherein the control device (20) is configured to provide said VSC voltage reference (v<c,ref>) to said VSC (10), and wherein said decoupling controller (21) comprises a phase compensation unit (211) configured to apply a phase compensation to said first complex signal (K; γ, δ), which phase compensation is selected to compensate for a phase-shift introduced by said virtual impedance. A method of grid-forming control and a power supporting system is also disclosed.

The disclosure provides a subsea electrical connection comprising: an electrical conductor extending along an axial direction of the electrical conductor; and a first flexible bellow surrounding the electrical conductor in a radial direction perpendicular to the axial direction, wherein an inner surface of the first flexible bellow is spaced apart from the electrical conductor in the radial direction.

There is disclosed herein an energy storage module (10) for an energy storage system (4) connected to a power grid (2). The energy storage module comprises an energy storage unit (20) having an equivalent capacitance (22). The energy storage unit has an internal resistance (24). The energy storage module further comprises a bypass circuit (30) connected in parallel to the energy storage unit. The bypass circuit comprises a bypass switch (32) and a current limiting device (34) connected in series. A resistance of the current limiting device is less than 500 % of the internal resistance of the energy storage unit. There are also disclosed herein an energy storage system and a power grid system.

The present disclosure relates to a method for monitoring a physical asset, the method comprising: receiving data associated with the physical asset from one or more data sources, wherein the data include measurements of one or more parameters of the physical asset; providing a physical model associated with the physical asset, wherein the physical model includes the one or more parameters of the physical asset; providing one or more cost functions, wherein each cost function is associated with one of a plurality of degradation scenarios; determining, based on the data, the physical model and the cost functions, a most probable degradation scenario of the degradation scenarios; and initiating a mitigation action based on the determined most probable degradation scenario indicating a potential failure.

A method of detecting a fault in a DMR line (6a-6e) or in an electrode line (13) in a bipolar HVDC system (1) comprises creating an AC current (i_harm) in the DMR line (6a-6e) or in the electrode line (13) during bipolar operation of the HVDC system (1), measuring currents (i₁, i₂) at two or more locations in the DMR line (6a-6d) or in the electrode line (13), calculating the difference (D_RMS) between the currents (i₁, i₂) and determining if the difference (D_RMS) is larger than a pre-defined threshold level (L) for detecting a fault (12) in the DMR line (6a-6e) or in the electrode line (13).

The present disclosure relates to transformer diagnosis methods for a transformer having a primary side source winding and two secondary side load windings. The method using relations based on a transformer model in which effects of magnetizing currents in the transformer core, leakage inductances and effective resistances of the primary and secondary windings are taken into account, wherein a transformer model of a three-winding transformer is used; transformer currents are measured during variations of the transformer loads wherein the power transferred through the respective windings are proportional; measurements are collected, for at least two different transformer loads, of a selection of primary side currents being indicative of the transformer load; a linear fits is performed by means of relations based on said transformer model, between equivalent load current and the primary current to estimate the value of at least one of the following transformer properties: turn ratio and magnetizing current.

In one embodiment, the electric system (1) comprises a static electric induction device (2) and an electrochemical impedance spectroscopy, EIS, sensor (3), wherein- the electric system (1) comprises a protective coating (4) to protect a component (21, 22, 23, 24, 25) of the electric system (1) from an exterior (A),- the EIS sensor (3) comprises a working electrode (31), a reference electrode (32) and a counter electrode (33), a test portion (40) of the protective coating (4) directly extends across the working electrode (31), the reference electrode (32) and the counter electrode (33),- the working electrode (31) is of areal fashion and is surrounded by the reference electrode (32), the latter being surrounded by the counter electrode (33), seen in top view of the test portion (40), and- the EIS sensor (3) is configured to provide information about a corrosion state of the protective coating (4).

An electric appliance (1) comprising a tank (2) and a shielding plate (2) mounted to an inner wall (21) of the tank (2) is specified, wherein the electric appliance (1) comprises a vibration damping system (4) configured to dampen a transfer of vibrations from the shielding plate (2) to the tank (2).

The invention relates to a tulip contact (10) for a gas-insulated circuit-breaker (1), comprising at least one contact section (12) having an inner contour (18) extending along a switching axis (X) between a tip (14) and a root (16), whereinthe inner contour (18) forms a contact throat at the tip (14) and exhibits a first radius (R1) in respect to the switching axis (X) that defines a narrowest cross section of the contact throat;the inner contour (18) comprises a first curved segment (C1) and, optionally, adjacent to the first curved segment (C1) a first straight segment (S1), at least one of the first curved segment (C1) and the first straight segment (S1) form the contact throat;the first curved segment (C1) exhibits a curving radius (RU) defining a convexity of first curved segment (C1) towards the switching axis (X) in a longitudinal sectional view; and1 ≤ RU/R1 and, optionally, RU/R1 ≤ 50.

Transformer (100) comprising a core (10) extending along a winding axis (A). The transformer (100) further comprises a winding (20) wound around the core (10) and extending along the winding axis (A), wherein the winding (20) terminates in a first axial end surface (21) and in a second axial end surface (22). The first axial end surface (21) is arranged opposite the second axial end surface (22) along the winding axis (A). The first axial end surface (21) and the second axial end surface (22) extend along a radial direction (R), which is perpendicular to the winding axis (A). The transformer (100) further comprises a ring element (30) arranged at the first axial end surface (21). The ring element (30) comprises ferromagnetic sheets (31), aligned along the winding axis (A). The ring element (30) further comprises an insulating shell (32) covering the ferromagnetic sheets (31).

The invention relates to a method for determining a faulted phase on a power transmission line, the method comprising obtaining a voltage signal and a current signal of the transmission line, selecting, based on a magnitude of the voltage signal and a magnitude of the current signal, a process for determining a faulted phase from a group of processes, and determining a faulted phase using the selected process.

There is provided mechanisms for differential protection of a transmission line of a transmission system. A method comprises obtaining a restraining current and a differential current from the transmission line. The method comprises determining a compensation current for the differential current. The method comprises providing the differential current as compensated for by the compensation current and the restraining current to a differential protection arrangement for making a trip decision. The method comprises detecting an internal fault for the transmission system. The method comprises, as a result thereof, providing the differential current without being compensated for by the compensation current and the restraining current to the differential protection arrangement for making the trip decision.

The invention relates to a method for determining a faulted phase on a power transmission line, the method comprising obtaining a voltage signal and a current signal of the transmission line, determining a change in the magnitude of the apparent power, ΔS, for each phase based on the voltage and current signal, determining a fault voltage angle, and determining a faulted phase based on ΔS and the fault voltage angle.

The present disclosure relates to a method for optimizing power generation in a power network using Benders decomposition, the method comprising performing one or more iterations of the following steps: a) determining a solution for each master problem, MPi, of a plurality of master problems, wherein each MPi is based on a different slack weight, Wi, of a plurality of slack weights; b) selecting a slack weight wi based on a first predetermined condition; c) selecting the solution for the MPi that corresponds to the selected wi as an input for one or more subproblems; d) determining one or more solutions for the one or more subproblems; and e) when a second predetermined condition is met, inputting the determined one or more solutions of the one or more subproblems of step d) into each MPi of step a) of the subsequent iteration.

In one embodiment, the electric system (1) comprises a static electric induction device (2) and a corrosion-sensitive coating (31), wherein- the electric system (1) further comprises a protective coating (4) to protect a component (21, 22, 23, 24, 25) of the electric system (1) from an exterior (A) of the electric system (1),- the corrosion-sensitive coating (31) comprises a luminescent material (30) and is applied on a substrate (34) of the component so that the corrosion-sensitive coating (31) is located between the substrate (34) and the protective coating (4), and- the luminescent material (30) is configured to emit luminescence (L) upon activation by at least one of a change of a pH value or occurrence of metal ions caused by corrosion of the substrate (34) so that the corrosion-sensitive coating (31) is configured to provide information about a corrosion state of the component.

Embodiments of present disclosure provide power conversion system (100) and method (200) for converting power. Power conversion system (100) comprising energy storage elements (102), first converter (103) connected to grid (104) to provide bi-directional flow of power between first converter (103) and grid (104). Second converter (107) connected to energy storage elements (102) at one end and other end connected to electric machine (106). Output of second converter (107) is supplied as input to electric machine (106). Controller (108) controls at least one of: current flowing from grid (104) to energy storage elements (102) and corresponding voltage between grid (104) and energy storage elements (102), current flowing from energy storage elements (102) to grid (104) via first converter (103) and corresponding voltage between energy storage elements (102) and grid (104), and current flowing from energy storage elements (102) to control operation of electric machine (106) via second converter (107).

The invention relates to a gas-insulated electrical application (1), particularly gas-insulated switchgear or gas-insulated line, comprisingtwo flanges (10, 12) connected to each other in a connection region (14),an insulator (30) with an annular base (32) arranged in the connection region (14) to be held by the two flanges (10, 12), anda sealing ring (50) arranged with a first side (51) in a groove (34) of the annular base (32) and configured for sealing with a second side (52) against one of the two flanges (10, 12), whereinthe sealing ring (50) comprises a recess (54) on at least one of the first side (51) and the second side (52), and/orthe groove (34) comprises stepped edges (36, 38) of which a radially outwards stepped edge (38) is wider in a radial direction than a radially inwards stepped edge (36).

The present disclosure relates to a charging management device for electric vehicles comprising at least one processor configured to obtain a predetermined state of charge, SoC, of a battery of the electric vehicle; determine a power consumption for reaching the predetermined SoC; determine a power consumption for preconditioning of the electric vehicle; and assign a power reserve of at least one electric charger to the electric vehicle based on the determined total power consumption. The disclosure further relates to a corresponding method, system, storage device and computer program.

The present disclosure relates to a charging management device for electric vehicles comprising at least one processor configured to obtain a predetermined state of charge, SoC, of a battery of the electric vehicle, obtain a target time for reaching the predetermined SoC, determine a power consumption for reaching the predetermined SoC within the target time and assign a sequence of chargers of a plurality of chargers based on the predetermined SoC, the target time, and the determined power consumption. The present disclosure further encompasses a corresponding method and system.

A transformer core (1) comprising a plurality of legs (2) is specified, wherein- the legs (2) comprise a plurality of core sheets (21),- a flitch plate (3) is attached to the core sheets (21) of one of the legs (2), the flitch plate (3) being configured to mechanically stabilize the leg (2),- the flitch plate (3) comprises a plurality of sheets (31), and- adjacent sheets (31) of the flitch plate (3) are bonded to one another via an adhesive layer (32).Further, a transformer (10) and a method of producing a transformer core (1) are specified.

In one embodiment, the method is for laminating a core part (20) of a static electric induction device (1) and comprises:A) providing a first stack (41) of sheets (25) for the core part (20),B) assembling a second stack (42) of the sheets (25), the sheets (25) are taken one by one from the first stack (41), the sheets (25) are oriented between 30° and 70° relative to a vertical direction (V),C) forming a third stack (43) by taking the whole second stack (42), the sheets (25) are oriented at most 25° relative to the vertical direction (V), andD) taking the third stack (43) and incorporating it into the core part (20), the sheets (25) are oriented at most 5° relative to the vertical direction (V) in the core part (20), so that the core part (20) is laminated of the sheets (25).

The present disclosure relates to a method of controlling a power supporting system (1) for a power transmission or distribution system, said power supporting system (1) being controllable to: a first power supporting mode (S8) in which said power supporting system (1) is capable of providing active power support and reactive power support, said active power support being provided by means of an energy storage system (20), ESS, having been charged to an energized state, and said reactive power support being provided by means of a voltage source converter (10), VSC, and a second power supporting mode (S5) in which said power supporting system (1) is capable of providing reactive power support by means of said VSC (10), and in which said ESS (10) is in a de-energized state, the method comprising: determining if said power supporting system (1) is operating with normal switching behavior or with emergent switching behavior; during normal switching behavior, controlling the power supporting system (1) so that, when switching mode from said first power supporting mode (S8) to said second power supporting mode (S5), said ESS (20) is discharged by means of operating at least one switch of the VSC (10), and in response to a detection of emergent switching behavior, controlling the power supporting system (1) so that, when switching mode from said first power supporting mode (S8) to said second power supporting mode (S5), said ESS (20) is discharged by means of a protection circuit (40) configured to consume energy of the ESS (20). A control device (30) and a power supporting system (1) is also disclosed.

In one embodiment, the electric system (1) comprises a static electric induction device (2) and an electric resistance, ER, sensor (3), wherein - the electric system (1) further comprises a protective coating (4) to protect a component (21, 22, 23, 24, 25) of the electric system (1) from an exterior (A) of the electric system (1), - the ER sensor (3) comprises a measurement conductor track (31) and a reference conductor track (32), - the protective coating (4) and the ER sensor (3) are configured to be in direct contact with the exterior (A), - the reference conductor track (32) is permanently shielded from the exterior (A), - the measurement conductor track (31) is less shielded from the exterior (A) than the reference conductor track (32), and - the ER sensor (3) is configured to provide information about a corrosion state of the protective coating (4).

The invention relates to a switchgear (1) for breaking an electrical connection, comprising a first contact structure (10) with a first base (12) and a first contact member (14); a second contact structure (20) with a second base (22) and a second contact member (24) and corresponding to the first contact structure (10); a spring accumulator (30) configured for resiliently mounting the first (14) or the second (24) contact member on the respective first (12) or second (22) base; a drive system (40) configured for at least partially moving the first (14) and the second (24) contact members relative to one another along a switching axis (2) in order to break the electrical connection to ground (GND); and a locking means (50) configured for locking the first (14) and the second (24) contact members along the switching axis (2) in a releasable manner.

The invention relates to a circuit-breaker (1), comprisinga gas compression arrangement (4) configured for compressing a quenching gas to be contained therein and which quenching gas is free of sulphur hexafluoride;a first contact (20) and a second contact (30) movable relative to one another along a switching axis (2);a nozzle (40) providing an inflow channel (IN) connecting the gas compression arrangement (4) to an arc zone (3), the inflow channel (IN) constituting a minimal inflow area to be passed by the quenching gas, the nozzle (40) radially restricting the arc zone (3) between the first contact (20) and the second contact (30);wherein, for the quenching gas to exit the arc zone (3), the circuit-breaker (1) provides at least one outlet (OUT1, OUT2) constituting a minimal outflow area, andthe minimal inflow area divided by the minimal outflow area defines an inflow-outflow-ratio being at least 0.5.

A contact structure (10) for an on-load tap changer (100) is provided, wherein the contact structure (10) is formed in one-piece. The contact structure (10) comprises a dense part (4) and at least one hollow part (2, 2A, 2B) arranged next to the dense part (4). The contact structure (10) comprises at least one shielding area (3, 3A, 3B), wherein at least one shielding area (3, 3A, 3B) is arranged next to the at least one hollow part (2, 2A, 2B). Moreover, an on-load tap changer (100) comprising at least one contact structure (10) and a method for producing a contact structure (10) are provided.

Embodiments of present disclosure provide converter system and method connecting first AC system to second AC system. First converter comprises first phase legs arranged by cascading first sub-converters between two DC lines connected to terminals of DC links. Connections of first sub-converters form AC connection terminals and first phase legs are arranged as first phase-arms in sub-converters. Second converter comprises second phase legs arranged by cascading sub-converters, between two DC lines connected to terminals of DC links. Connections of sub-converters form AC connection terminals and second phase legs are arranged second phase-arms in the cascaded sub-converters. First converter or second converter operates rectifier or converter. First phase legs or second phase legs comprises bipolar cell in each phase to control potential difference between DC links. Potential difference between DC links and output voltage of the inverter is changed by value of voltage blocked by bipolar cell.

The present disclosure relates to a method for determining a status of a power grid, wherein the power grid comprises a plurality of branches each connected between two buses. The method comprises obtaining current measurements of at least one of the plurality of branches, obtaining a grid model, the model comprising the plurality of branches and buses, estimating, using the grid model and the current measurements, a voltage value of at least one of the plurality of buses, determining, using the voltage values and current measurements, current estimates of remaining branches of the plurality of branches having an unknown status and determining the status of the remaining branches of the plurality of branches based on the estimated currents.

There is disclosed herein a method for operating a converter unit (10) of a STATCOM (4), configured to be connected to a power grid (2) connected to at least one voltage flicker source (6). The method comprises determining (110) a grid-forming reference based on an active power (110a) and/or an energy level (110b) of the STATCOM, determining (120) a grid-following reference based on a reactive power (120a) and/or a voltage (120b) of the power grid. The method further comprises obtaining (130) a first current reference value by subjecting (125) the grid-forming reference and the grid-following reference to a mathematical operation and subjecting the result thereof to a virtual admittance, determining (140) a second current reference value based on a current at the voltage flicker source, and obtaining (150) a current reference, for controlling the converter unit, based on the first current reference value and the second current reference value. There are also disclosed herein a control unit and a power grid system.

An electric circuit (1) comprising a power thyristor (2) device (2) is specified, wherein - a gate (21) of the power thyristor device (2) is connected to an internal reference (10) of the electric circuit (1) via a first switch (31); and - a cathode (22) of the power thyristor device (2) is connected to the internal reference (10) of the electric circuit (1) via a second switch (32). Further, a method of operating an electric circuit (1) is specified.

The present disclosure relates to a method for determining a status of a power grid, wherein the power grid comprises a plurality of branches each connected between two buses. The method comprises obtaining current measurements of at least one of the plurality of branches, obtaining a grid model, the model comprising the plurality of branches and buses, estimating, using the grid model and the current measurements, a voltage value of at least one of the plurality of buses, obtaining, from a controller, at least one parameter of at least one of remaining branches of the plurality of branches, determining, based on the at least one parameter and current measurements of the at least one of the plurality of branches, currents of the remaining branches of the plurality of branches and determining or correcting the status of the remaining branches of the plurality of branches based on the estimated currents.

The disclosure provides a method for determining a faulted section in a collector grid feeder comprising a plurality of sections, the method comprising obtaining a voltage at a collector bus at a local end of the feeder (8), obtaining a first current signal through the feeder and a second current signal through the feeder, determining a first fault direction value, based on the first current signal and a derivative of the first current signal, and determining a first fault direction based on the first fault direction value, and determining a second fault direction value based on the second current signal and a derivative of the second current signal and determining a second fault direction based on the second fault direction value, and determining a faulted section based on a comparison between the first fault direction and the second fault direction.

The disclosure provides a method for determining a fault location in a faulted section in a collector grid feeder, the method comprising obtaining a first voltage and a first current signal at a local end of the section, obtaining a second current signal at a remote node of the section; and determining a fault location based on the first voltage signal, the first current signal, and the second current signal.

Green-hydrogen plants, connected to weak power grids in geographically remote areas, are prone to frequency fluctuations and instability due to the variability in renewable-energy generation. Disclosed embodiments provide balancing control within the green-hydrogen plant to provide both fast frequency response during contingency conditions and automatic frequency regulation during normal conditions. The fast frequency response may be centralized or distributed, and may combine frequency droop response with fast frequency response or synthetic inertial response. The automatic frequency regulation may determine optimal allocations of a total net power adjustment, which compensates a balance control error in the system, to a plurality of different power resources of the green-hydrogen plant. These power resources may comprise electrolyzer plant(s), renewable energy source(s), and/or energy storage system(s).

There is provided a conservator tank (100) for a transformer tank (200) comprising a housing (10) with a magnetic inner surface (11), which houses an insulation oil (20) inside the housing (10). The conservator tank (100) further comprises a rubber bag (30) with an upper part (31) facing away from the insulation oil (20), and a lower part (32) opposite the upper part (31) facing towards the insulation oil (20). The rubber bag (30) further has a first peripheral side (33) arranged between the upper part (31) and the lower part (32) and a second peripheral side (34) arranged between the upper part (31) and the lower part (32) opposite the first peripheral side (33). The rubber bag (30) is configured to inflate or deflate inside the housing (10), such that the lower part (32) contacts and immerses into an oil surface (21) of the insulation oil (20). The rubber bag (30) has a rubber layer (35) with a first magnetic fastening element (36) arranged on the upper part (31), wherein the first magnetic fastening element (36) is configured to be attached to the magnetic inner surface (11) in a mounted state of the rubber bag (30).