WO2014037396A1

WO2014037396A1 - Insulation fluid filling method and filling apparatus

Info

Publication number: WO2014037396A1
Application number: PCT/EP2013/068278
Authority: WO
Inventors: Thomas Alfred Paul; Axel Kramer; Denis Tehlar
Original assignee: Abb Technology Ag
Priority date: 2012-09-04
Filing date: 2013-09-04
Publication date: 2014-03-13
Also published as: EP2893544A1

Abstract

A method and device for providing an insulation fluid (10) and for filling this insulation fluid (10) into medium-voltage or high-voltage switchgear (1) is provided. The method comprises method elements of mixing at least two fluid components (A,B) for yielding a first amount (M1) of insulation fluid (10), monitoring a first mixing ratio (R1) of this first amount (M1), and monitoring a second mixing ratio (R2) of a second amount (M2) of the insulation fluid (10) that is already in the electrical apparatus (1). The first mixing ratio (R1) is controlled such that no condensation of a fluid component (A,B) takes place. Furthermore, the first mixing ratio (R1) and the first amount (M1) of the insulation fluid are controlled using the second mixing ratio (R2), the second amount (M2), a desired target mixing ratio (R), and a target amount (M) of the insulation fluid.

Description

Insulation fluid filling method and filling apparatus

Technical Field

The present invention relates to a method and device for mixing fluid components to yield an insulation fluid mixture and for filling this insulation fluid mix^¬ ture into an electrical apparatus, in particular into gas-insulated medium-voltage or high-voltage switchgear.

Introduction and Background Art

Dielectric insulation media in liquid and/or gaseous states (i.e. fluids) are widely applied to insu^¬ late an electrically active part in a variety of electri^¬ cal apparatuses such as gas-insulated switchgear (GIS) . For example, the electrically active part in medium- voltage or high-voltage metal-encapsulated switchgear is arranged in a gas-tight compartment which encloses an in^¬ sulation gas with a pressure of several bars, which elec^¬ trically separates the compartment of the apparatus from its electrically active part. In other words, the insula^¬ tion gas does not allow the passage of electrical current from the electrically active part to the compartment. A commonly used dielectric insulation gas is sulfur hex- afluoride (SFg) , which exhibits excellent insulation and electric arc extinguishing capabilities. However, SFg is a strong contributor to the green-house effect and thus has a high global warming potential. Therefore, alterna^¬ tive insulation fluids should be found.

Several alternative insulation fluids have been identified. Some of these alternatives comprise mul^¬ ti-component fluid mixtures, i.e. they comprise more than one molecular or atomic species. It is found that certain properties of such insulation fluid mixtures are compul^¬ sory to the safe operation of the electrical apparatus. As an example, the dielectric break-down strength of the insulation fluid is strongly dependent on local concentration ratios of the mixture fluid components and on total fluid pressure. In order to achieve and upkeep the mixture's insulating features and thus the safety and functionality of the electrical apparatus, the fluid com^¬ ponents need to be carefully mixed and delivered to the electrical apparatus.

Niemeyer L., "Cigre Guide for SF6 gas mixtures", l^st international conference on SF6 and the Envi^¬ ronment, November 2000 comments on suitable mixing proce^¬ dures for multi-component insulation gas mixtures.

US 2011/0297149 Al discloses a veterinary an^¬ esthesia monitor which establishes, maintains, and re^¬ ports upon anesthesia gas mixtures.

EP 0 894 506 A2 relates to a medical anesthe^¬ sia delivery system for providing flows of breathing and anesthesia gases to a patient. The anesthesia delivery system comprises a feedback and control circuit to ana^¬ lyze and regulate the mixing ratio of patient inhaled gases .

EP 0 476 502 A2 discloses a gas-insulated electric apparatus in which insulation abnormalities in a compartment are detected and countermeasures can be ini^¬ tiated, e.g. by injecting an insulation gas into the compartment .

The disclosed methods and devices have the disadvantage, however, that they are not suitable for controlled filling of electrical apparatuses, in particu^¬ lar for delivering well-defined mixing ratios at pres^¬ sures above atmospheric pressure.

Disclosure of the Invention

Hence it is a general objective of the pre^¬ sent invention to provide an improved method and device for providing an insulation fluid mixture for an electrical apparatus and for filling this insulation fluid mix^¬ ture into the electrical apparatus.

These objectives are achieved by the method and device of the independent claims. Accordingly, a method for filling a first amount of an insulation fluid which comprises at least a first fluid component and a second fluid component into a compartment of a fluid-insulated electrical apparatus (such as, e.g., gas-insulated medium-voltage or high- voltage switchgear) comprises a method element of

- mixing the at least two fluid components which are to be comprised in the first amount of the in^¬ sulation fluid. This is done at a first mixing ratio, i.e. at or to achieve a desired mixing ratio for the first amount of the insulation fluid. Thus, the first amount of mixed insulation fluid is yielded.

In embodiments, a molecular weight of the first fluid component A differs from a molecular weight of the second fluid component B, in particular at least by a factor of 2, particularly at least by a factor of 5.

The mixing-step as well as the other method elements in the method (see below) are carried out at a filling temperature Tfiii_/ e.g at ambient temperature, e.g. 20°C during commissioning of the electrical appa^¬ ratus .

The method further comprises a method element of

deriving the first mixing ratio of the first amount of the insulation fluid, i.e. an actual mix^¬ ing ratio of the first amount of the insulation fluid that can deviate from a desired first mixing ratio during the mixing step. This is achieved by means of a first sensor, such as a gas chromatography an optical sensor, and/or a p-T-p-sensor (see below for further examples) . Thus, information indicative of the actual first mixing ratio of the first amount of the insulation fluid is ob^¬ tained .

The method comprises a further method element of

- filling the first amount of the insulation fluid into the compartment of the electrical apparatus. There, as an option, this first amount of insulation flu^¬ id can be actively (e.g. via a fan, see below) or pas^¬ sively (e.g. via diffusion) mixed with an optional second amount of the insulation fluid that is already in the compartment (if any) . By filling the first amount into the compartment, the total amount of insulation fluid in the compartment of the electrical apparatus is increased while information indicative of the first mixing ratio prior to and/or after filling is available.

The first mixing ratio of the first amount of the insulation fluid is controlled such, or - in other words - the amount of the first component and the amount of the second component that are mixed to yield the first amount of the insulation fluid are controlled such that condensation temperatures of the first and second fluid components in the first amount (i.e. in the insulation fluid that is to be filled into the compartment) are be^¬ low the filling temperature Tfiii- In other words, the first mixing ratio is controlled in a way that no conden^¬ sation of one or more fluid components takes place during the mixing and filling of the insulation fluid into the electrical apparatus. Thus, it is ensured that during the filling method the insulation fluid is in a gaseous form and that no condensation which could alter the mixing ratio takes place.

In embodiments, a concentration of the first component A in the first amount of the insulation fluid does not exceed a threshold concentration above which a condensation of the first component A would take place at the filling temperature Tf-_j__]__]_.

In embodiments, the same can be achieved for the second component B, i.e. a concentration of the sec^¬ ond component in the first amount of the insulation fluid does not exceed a threshold concentration above which a condensation of the second component B would take place at the filling temperature Tf-_j__]__]_. Thus, the first amount of the insulation flu^¬ id can be kept in a gaseous form.

The first mixing ratio of the first amount and the first amount of the insulation fluid itself are controlled using a target mixing ratio and a target amount of insulation fluid in the compartment of the electrical apparatus. The terms "target mixing ratio" and "target amount" refer to a mixing ratio or an amount of insulation fluid, respectively, that is or are to be reached after completion of the filling process or at least at a later stage during the filling process, e.g. at the end of the commissioning procedure of the electri^¬ cal apparatus .

By controlling the first mixing ratio and the first amount using the target mixing ratio and a target amount, the compartment of the electrical apparatus can be more easily filled with the right or desired amount (i.e. "target amount") of insulation fluid with the right or desired mixing ratio (i.e. "target mixing ratio") for reliable operation of the electrical apparatus.

Alternatively to the method described above, in addition to deriving the first mixing ratio of the first amount of the insulation fluid, the method further comprises a method element of

- deriving a second mixing ratio of a second amount of the insulation fluid which is already present in the compartment of the electrical apparatus. This is achieved by means of a second sensor which can, e.g., be an integral part of the electrical apparatus or which can only be temporarily arranged in the compartment of the electrical apparatus; the second sensor can also be con^¬ nected to the compartment of the electrical apparatus, e.g. via a self-sealing connector. Thus, information indicative of the actual second mixing ratio of the second amount of the insulation fluid in the compartment of the electrical apparatus is obtained. In other words, the mixing ratio of the insulation fluid that is already in the compartment of the electrical apparatus (i.e. the second mixing ratio of the second amount of the insula^¬ tion fluid) is derived by means of the second sensor. Like the first sensor, also the second sensor can for ex^¬ ample comprise at least one of the group of: a gas chro^¬ matography an optical sensor, and a p-T-p-sensor (see below) .

Then, as an alternative to or in addition to controlling the first mixing ratio and the first amount of the insulation fluid which is to be filled into the compartment using the above discussed target mixing ratio and target amount, the first mixing ratio of the first amount and the first amount of the insulation fluid it^¬ self (i.e. the concentrations of the first and second fluid components as well as the total number of particles in the first amount) can also be controlled using the second mixing ratio and the second amount of the insula^¬ tion fluid which is already in the compartment. In other words, the first mixing ratio and the first amount of still-to-be filled insulation fluid can then be con^¬ trolled depending on the second mixing ratio and the sec^¬ ond amount of insulation fluid that is already in the compartment of the electrical apparatus (in addition to or as an alternative to using the target mixing ratio and target amount) .

By controlling the first mixing ratio and the first amount using the second mixing ratio and the second amount and/or a target mixing ratio and a target amount, the compartment of the electrical apparatus can more eas^¬ ily be filled with the right or desired amount (i.e. "target amount") of insulation fluid with the right or desired mixing ratio (i.e. "target mixing ratio") for re^¬ liable operation of the electrical apparatus.

In an embodiment of the method, a filling pressure (which is, in other words, an equivalent to the "target amount" for a fixed volume of the compartment) of the insulation fluid in the compartment of the electrical apparatus after the above described filling step is above 1 bar, preferably above 2 bars, more preferably above 5 bars (measured at an insulation fluid temperature of 20°C, which can also be the filling temperature of the electrical apparatus of the electrical apparatus Tfj__]__]_) . Thus, adequate amounts of insulation fluid and a suffi^¬ cient dielectric breakdown strength, e.g. for medium- voltage or high-voltage operation of the electrical appa^¬ ratus, can be achieved, or can easier be achieved, in the compartment .

In embodiments, the first and second fluid components are brought into (e.g. by a heater) - or, al^¬ ternatively, it is ensured that they already are in - gaseous states, advantageously prior to carrying out the above described mixing step. Thus, the first mixing ratio can more easily be controlled, e.g. by controlling and/or measuring flow rates of at least one fluid component, in particular of both fluid components.

In another embodiment of the method or meth^¬ ods, the first fluid component is in a liquid state at the filling temperature Tf-j__]__]_ and at a pressure of 1 bar and the second fluid component is in a gaseous state at the filling temperature Tf-j__]__]_ and at a pressure between 5 bar and 200 bar. The first fluid component A is then brought into a gaseous state (e.g. by heating) and it is ensured that the second fluid component B is in a gaseous state and the fluid components are subsequently mixed at the filling temperature Tf-j__]__]_ and at a pressure between 3 bar and 10 bar. This has the advantage that the first amount of the insulation fluid is yielded from a first liquid (at the given conditions) component and from a second gaseous (at the given conditions) fluid component.

In another embodiment, the method comprises a further method element of

- homogenizing the first amount (which is to be filled into the compartment) of the insulation fluid for reducing a mixture fluctuation and/or a density fluctuation in the first amount.

Alternatively or in addition, the method com^¬ prises a further method element of

- homogenizing the second amount of the insu^¬ lation fluid for reducing a mixture fluctuation and/or a density fluctuation in the second amount of the insula^¬ tion fluid which is already in the compartment of the electrical apparatus.

This or these homogenization step(s) is or are preferably carried out prior to deriving the first and/or second mixture ratio (s) and lead(s) to a more ho^¬ mogeneous mixture and/or density of the fluid components in the respective amounts of insulation fluid and thus to a more reliable derivation of the first and/or second mixture ratio (s) . Furthermore, due to the reduction of local mixture and/or density fluctuations, a more relia^¬ ble filling and operation of the electrical apparatus is achieved or simplified.

In another embodiment, the method comprises a further method element of

- reducing a or the second amount of the in^¬ sulation fluid in the compartment of the electrical appa^¬ ratus, in particular prior to carrying out the step of filling the first amount of the insulation fluid into the compartment of the electrical apparatus. In other words, the second amount of insulation fluid in the compartment of the electrical apparatus is reduced or removed before the first amount of the insulation fluid is filled into the compartment. This is, e.g., very useful, if during revision of the electrical apparatus the "old" insulation fluid needs to be fully or partially removed from the compartment (e.g. due to contamination) prior to filling the first amount of "freshly mixed" insulation fluid into the compartment .

Alternatively or additionally, the method comprises a further method element of - reducing an amount of a filling gas in the compartment of the electrical apparatus, in particular prior to carrying out the step of filling the first amount of the insulation fluid into the compartment of the electrical apparatus. In other words, a filling gas, e.g. air or nitrogen, can at least partially be removed from the compartment before the first amount of insula^¬ tion fluid is filled into the compartment. This is, e.g., very useful, if during commissioning of the electrical apparatus ambient air that is present in the compartment needs to be fully or at least partially removed from the compartment prior to filling the first amount of "fresh" insulation fluid into the compartment.

This step has or these steps have the ad^¬ vantage that a "target amount" and/or "target mixing ra^¬ tio" of insulation fluid can be as clean as possible and as precisely controlled as possible, which leads to a better insulation performance and thus to a more reliable operation of the electrical apparatus.

In another embodiment of the method, at least the method elements of

- mixing the fluid components at the first mixing ratio, and of

- deriving

a) the first mixing ratio of the first amount of the insulation fluid and/or

b) a or the second mixing ratio of the second amount of the insulation fluid, and of

- filling the first amount of the insulation fluid into the compartment of the electrical apparatus are carried out repeatedly.

In other words, the above mentioned steps are repeated as long as the compartment of the electrical ap^¬ paratus is not sufficiently filled with insulation fluid, or, in other words, as long as the "target mixing ratio" and/or the "target amount" of insulation fluid in the compartment is or are not yet reached. The term "repeat- edly" as used herein refers to both a repeated execution of the above mentioned steps one-after-another with dis^¬ cernible start- and stop-points of all single steps as well as to a (in time) continuous execution, or to an at least temporarily continuous execution, of at least some of the method elements.

As an example for such a repeated execution of the steps, during commissioning of the electrical ap^¬ paratus, the mixing and filling steps can be carried out continuously while the mixing-ratio-derivation-step ( s ) is or are carried out, e.g., once or twice every second. Then, if the second mixing ratio and/or the second amount of insulation fluid is or are within a preset band around the target mixing ratio and/or target amount of the insu^¬ lation fluid in the compartment (e.g. not differing by more than 5% from the target mixing ratio and/or from the target amount) , the mixing and filling is stopped and the electrical apparatus is (at least as far as its fluid- insulation is concerned) ready for operation.

If the above mentioned steps are carried out repeatedly, the filling steps are preferably carried out against increasing second amounts of the insulation fluid in the compartment (or, equivalently, against increasing insulation fluid pressures in the compartment) . In other words, the total amount of insulation fluid in the com^¬ partment of the electrical apparatus increases over time due to the repeated filling of freshly mixed first amounts of insulation fluid into the compartment. Thus, the target mixing ratio and/or the target amount of insu^¬ lation fluid in the compartment can be reached using re^¬ peated (i.e. discernible and/or continuous) mixing and filling steps.

In another embodiment, the first sensor and/or a or the second sensor (each) comprises at least one of the group of

- a gas chromatograph for carrying out a gas chromatographic measurement on the first and/or second amount of the insulation fluid. The term "gas chromato^¬ graphic measurement" herein relates to an experimental quantification of a physical property (e.g. a retention time) of at least one fluid component of the insulation fluid comprising interactions between atoms or molecules of the insulation fluid and a carrier fluid called the "mobile phase" and a fixed material called the "stationary phase". This stationary phase can e.g. be located in a column of the gas chromatograph .

- An optical sensor for carrying out an optical measurement on the first and/or second amount of in^¬ sulation fluid. The term "optical measurement" herein re^¬ lates to an experimental quantification of a physical property of at least one fluid component of the insula^¬ tion fluid comprising interactions between atoms or mole^¬ cules of the insulation fluid and photons. Examples are

* optical absorption measurements by means of, e.g., a multi-pass spectroscope at e.g. at least one wavelength between 0.2 μιη and 20 μιη or

* fluorescence measurements, e.g., at at least one fluorescence excitation wavelength between 100 nm and 400 nm and/or at at least one fluorescence emission wavelength between 350 nm and 600 nm.

- A photoacoustic sensor for carrying out a photoacoustic measurement on the first and/or second amount of the insulation fluid. Such a photoacoustic measurement comprises optical excitation of at least one fluid component of the insulation fluid, e.g. in the wavelength range between 0.2 μιη and 20 μιη, followed by detection of the acoustic response of the insulation flu^¬ id using, e.g., a microphone. Alternatively or in addi^¬ tion, also acoustic sensors relying on other measurement principles can be used.

- A pressure (p) sensor, a temperature (T) sensor, and

* a density (p) sensor,

* a speed of sound sensor, * a viscosity sensor, and/or

* a thermal conductivity sensor.

More than three-parameter-sensor-systems can be used for error reduction and/or for more complex insulation fluid mixtures (e.g. comprising more than two flu^¬ id components) . The respective mixing ratios of the first and/or second amount of the insulation fluid (or even the concentrations of the first and second fluid components in the first and/or the second amounts) can, e.g., be de^¬ rived using an equation of state (i.e. a "thermodynamic equation describing the state of matter under a given set of physical conditions" (from http://en.wikipedia.org/ wiki /Equation_of_state as accessed on May 03, 2012) which is, e.g., selected from the group consisting of:

* the ideal gas law, i.e. pV = nRT with p being an absolute pressure, V being a volume, n being a number of molecules (usually expressed in moles) , R being the ideal gas constant, and T being an absolute tempera^¬ ture,

* the van-der-Waals equation of state, i.e. (p+a/V_m ²) (V_m-b)=RT with V_m being a molar volume and a, b being substance-specific parameters for the respec^¬ tive insulation fluid component,

* the virial equation of state, i.e. pV_m/(RT) = 1 + B(T)/V_m + C(T)/V_m ² + D(T)/V_m ³ + ... with B(T), C(T), D(T), ... being temperature-dependent terms that correspond to interactions between molecules,

* the Beattie-Bridgeman equation of state, i.e. p = R_UT/ (V_m ²) (1-c/ (V_mT³) ) (V_m+B) -A/ (V_m ²) with A = A_Q(l-a/V_m), B = B_Q(l-b/V_m), R_u being a gas constant in the form R_u = 8.314 kPa m³/ (kmol K) , V_m being a molar volume, and a, b, c, AQ, and B Q being substance-specific parameters for the respective insulation fluid component, and

* the Peng-Robinson equation of state, i.e. p = RT/(V_m-e) - d (T) / (V_m (V_m+e) + e (V_m-e) ) with d(T) and e being empirical parameters. When an equation of state other than the ideal gas law is used, the behavior of a gas can be better predicted than with the ideal gas law alone and the pre^¬ diction can be extended to liquids. This is possible by putting in terms to describe attractions and repulsions between molecules as well as the molecular volume itself which leads to a reduction in the molar volume.

For example, the document US 7 184 895 B2 gives details on how the fluid component concentration ( s ) is or are derived using a pressure, a temperature, and a density measurement.

The term "concentration" herein defines

- a quantity (with units) which is indicative of an amount per volume unit, e.g. a particle number per volume unit, moles per volume unit, or a number density, or

- a number (without units) which is indicative of a ratio such as a mole fraction, a pressure-normalized partial pressure, a volume fraction, a mass fraction, or a density fraction.

The advantage of using a sensor or sensors as described above is that information indicative of the first and/or second mixing ratio (s) is easier to obtain. Specifically, the signals from the sensor (s) are used to derive the actual first and/or second mixing ratio (s), and the desired first mixing ratio is or can be con^¬ trolled using this information. This simplifies the pro^¬ cess of reaching the correct target mixing ratio as well as target amount of insulation fluid in the compartment of the electrical apparatus.

In another embodiment, the method comprises a further method element of

- deriving a mass flow of at least one of the fluid components (or advantageously deriving mass flows of all fluid components in the first amount) prior to or during the step of filling the first amount of the insu^¬ lation fluid into the compartment of the electrical appa- ratus . This has the further advantage that information indicative of the total amount of the respective fluid component that is filled into the compartment of the electrical apparatus is available. Thus, e.g., the mixing and/or filling process (es) can be stopped as soon as a target amount of the respective fluid component in the compartment of the electrical apparatus is reached.

In another embodiment of the method, the first fluid component is selected from the group consist^¬ ing of:

- sulfur hexafluoride,

- partially or fully fluorinated ethers, in particular hydrofluoroethers , hydrofluoro monoethers, hy- drofluoro monoethers containing at least 3 carbon atoms, perfluoro monoethers, perfluoro monoethers containing at least 4 carbon atoms, fluorooxiranes , perfluorooxiranes , hydrofluorooxiranes , perfluorooxiranes comprising from three to fifteen carbon atoms, hydrofluorooxiranes com^¬ prising from three to fifteen carbon atoms, and mixtures thereof,

- partially or fully fluorinated ketones, in particular hydrofluoro monoketones, perfluoro mono^¬ ketones, perfluoro monoketones comprising at least 5 car^¬ bon atoms, perfluoro monoketones comprising exactly 5 or 6 or 7 or 8 carbon atoms, and mixtures thereof,

- fluoroolefins ; in particular: perfluoroole- fines, hydrofluoroolefins (HFO) , hydrofluoroolefins (HFO) comprising at least three carbon atoms, hydrofluoro- olefins (HFO) comprising exactly three carbon atoms, trans-1, 3, 3, 3-tetrafluoro-l-propene (HFO-1234ze) , 2, 3, 3, 3-tetrafluoro-l-propene (HFO-1234yf) , and mixtures thereof; and

- mixtures thereof.

The second fluid component B is selected from the group consisting of:

- nitrogen, - oxygen,

- carbon dioxide,

- nitric oxide,

- nitrogen dioxide,

- nitrous oxide,

- argon,

- methanes, in particular partially or fully halogenated methanes, in particular tetrafluoromethane or trifluoroiodomethane,

- air, in particular technical air or synthetic air, and

- mixtures thereof.

Thus, an improved insulation performance can be achieved for the insulation fluid of the electrical apparatus .

In further embodiments, the first fluid com^¬ ponent is selected from the group consisting of:

cyclic and/or aliphatic fluoropentanones , preferably cyclic and/or aliphatic perfluoropentanones , more preferably 1, 1, 1, 3, 4, 4, 4-heptafluoro-3- (tri-fluoro- methyl) butan-2 -one,

cyclic and/or aliphatic fluorohexanones , preferably cyclic and/or aliphatic perfluorohexanones , more preferably 1, 1, 1, 2, 4, 4, 5, 5, 5-nonafluoro-4- (tri- fluoromethyl ) pentan-3-one,

cyclic and/or aliphatic fluoroheptanones , preferably cyclic and/or aliphatic perfluoroheptanones ,

- sulfur hexafluoride, and

- hydrofluoroethers .

In another embodiment, the second fluid com^¬ ponent consists of:

nitrogen and oxygen with relative partial pressures between p (N₂ ) / (p (O2 ) +p (N₂ ) ) =0.7 , p (0₂ ) / (p (0₂ ) + p(N₂))=0.3 and p (N₂ ) / (p (0₂ ) +P (N₂ ) ) =0.95 , p (0₂ ) / (p (0₂ ) + p (N₂) ) =0.05 or

carbon dioxide and oxygen with relative partial pressures between p (C0₂ ) / (p (0₂ ) +p (C0₂ ) ) =0.6, p(0₂)/ (p(0₂)+p(C0₂) )=0.4 and p (C0₂) / (p (0₂) +p (C0₂) ) =0.99, p(0₂)/ (p(0₂) + p(C0₂) )=0.01, or

carbon dioxide and nitrogen with relative partial pressures between p (C0₂ ) / (p (N₂ ) +p (C0₂ ) ) =0.1 , p (N₂) / (p (N₂) +p (C0₂) ) =0.9 and p (C0₂ ) / (p (N₂ ) +p (C0₂ ) ) =0.9, p(N₂) / (p(N₂)+p(C0₂) )=0.1.

The first fluid component A comprises at least one of the group consisting of:

1,1,1,3,4,4, 4-heptafluoro-3- (tri-fluoro- methyl) butan-2-one with a partial pressure between 0.1 bar and 0.7 bar at a temperature of 20°C,

1, 1, 1, 2, 4, 4, 5, 5, 5-nonafluoro-4- (tri- fluoromethyl ) pentan-3-one with a partial pressure between 0.01 bar and 0.3 bar at a temperature of 20°C,

sulfur hexafluoride with a partial pressure between 0.1 bar and 2 bar at a temperature of 20 °C, and hydrofluoroethers with a partial pressure be^¬ tween 0.2 bar and 1 bar at a temperature of 20 °C.

In another embodiment, the second fluid com^¬ ponent comprises

nitrogen and oxygen with relative partial pressures between p (N₂) / (p (0₂) +p (N₂) ) =0.75, p (0₂ ) / (p (0₂ ) + p(N₂))=0.25 and p (N₂ ) / (p (0₂ ) +p (N₂ ) ) =0.90 , p (0₂ ) / (p (0₂ ) + p (N₂) ) =0.10 and

wherein the first fluid component (A) com^¬ prises 1,1,1,3,4,4, 4-heptafluoro-3- (tri-fluoromethyl) butane-one with a partial pressure between 0.25 bar and 0.5 bar and/or the fluid component 1,1,1,2,4,4,5,5,5- nona-fluoro-4- (tri-fluoromethyl) pentan-3-one with a par- tial pressure between 0.02 bar and 0.3 bar at a tempera^¬ ture of 20°C.

In another embodiment, the method comprises a further method element of

- deriving a first concentration of the first fluid component of a or the second amount of the insula^¬ tion fluid in the compartment of the electrical appa^¬ ratus .

The method comprises a further method element of

- deriving a second concentration of the second fluid component of the second amount of the insula^¬ tion fluid in the compartment of the electrical appa^¬ ratus .

This information can, e.g., be derived using a sensor arrangement as discussed above in the electrical apparatus and/or using integrated mass flow measurements of the filled amounts of the first fluid components. Thus, the fluid component concentrations of the insula^¬ tion fluid in the compartment of the electrical apparatus are derivable.

In embodiments, the method further comprises a method element of deriving a dielectric break-down strength of a or the second amount of the insulation fluid in the compartment of the electrical apparatus. This is achieved using a or the first concentration of the first fluid component of the insulation fluid and us^¬ ing a or the second concentration of the second fluid component of the insulation fluid, and, e.g., using the following equation

^Ebd = S{c_A,c_B) ∑CiE_{cri i}

i=A,B Here, E_cr-j_-|- ^ ^and ^Ecrit,B ^are fluid component specific critical field strengths of the first fluid com^¬ ponent A and the second fluid component B; c^ and eg are the first and second concentrations of the first and sec^¬ ond fluid components A and B; S (c^, eg) is a synergy pa^¬ rameter; and i is an index running over the fluid compo^¬ nents A and B.

Then, the dielectric break-down strength E_j^ of the insulation fluid in the compartment of the elec^¬ trical apparatus is derivable and it can be compared to a target dielectric breakdown strength. As soon as this target dielectric breakdown strength is reached (e.g. within a band of at least +5% above a threshold dielec^¬ tric breakdown strength), the mixing and/or filling can be stopped and the electrical apparatus is (at least as far as its fluid-insulation is concerned) ready for operation.

In other embodiments, the method comprises a further method element of

- deriving an operating state of the electrical apparatus using a or the first concentration of the first fluid component in the second amount of the insula^¬ tion fluid and using a or the second concentration of the second fluid component in the second amount of the insu^¬ lation fluid in the compartment of the electrical appa^¬ ratus .

- deriving an operating state of the electrical apparatus using a or the dielectric break-down strength Ej^ of the second amount of the insulation fluid in the compartment of the electrical apparatus.

This operating state can for example be se^¬ lected from a group of possible operating states consist^¬ ing of: - normal, i.e. undisturbed, operation of the electrical apparatus,

- underfilling of the compartment, i.e. an insufficient amount (e.g. second amount) of insulation fluid in the compartment of the electrical apparatus,

- overfilling of the compartment, i.e. a su^¬ perfluous amount (e.g. second amount) of insulation fluid in the compartment of the electrical apparatus,

uniform leakage of the insulation fluid (e.g. of the second amount of the insulation fluid) from the compartment of the electrical apparatus, i.e. fluid component-independent loss of insulation fluid from the compartment of the electrical apparatus,

- preferential leakage of one fluid component (e.g. A or B) from the insulation fluid (e.g. from the second amount of the insulation fluid) from the compart^¬ ment of the electrical apparatus, i.e. increased loss of one fluid component compared to the other fluid component from the compartment of the electrical apparatus, thus leading to a change of the mixing ratio (e.g. the second mixing ratio) of the insulation fluid in the compartment of the electrical apparatus,

condensation or preferential condensation of one fluid component from the insulation fluid (e.g. from the second amount of the insulation fluid) in the compartment of the electrical apparatus, e.g. a state transition from gaseous to liquid state or vice versa of only one or at least preferentially one fluid component

(e.g. A or B) of the insulation fluid (e.g. of the second amount of the insulation fluid) in the compartment of the electrical apparatus,

adsorption or preferential adsorption of one fluid component (e.g. A or B) of the insulation fluid

(e.g. of the second amount of the insulation fluid) in the compartment, for example on a component of the elec^¬ trical apparatus and/or on an inner surface of the com^¬ partment of the electrical apparatus, - chemical process or preferential chemical process of one or pertinent to one fluid component (e.g. A or B) of the insulation fluid (e.g. of the second amount of the insulation fluid) , in particular in the compartment of the electrical apparatus,

- appearance of at least one new fluid compo^¬ nent in the insulation fluid (e.g. of the second amount of the insulation fluid) , in particular in the compartment of the electrical apparatus; e.g. due to arcing, partial discharges, evaporation, light, high temperature, or any other chemical process or stemming from any other source, in particular wherein the new fluid component is a contaminant, i.e. an undesired substance in the insula^¬ tion fluid, and

- decomposition or preferential decomposition of at least one fluid component (e.g. A and/or B) of the insulation fluid (e.g. of the second amount of the insu^¬ lation fluid) , in particular in the compartment of the electrical apparatus; e.g. due to arcing, partial dis^¬ charges, light, high temperature, and/or reactions of at least one of the fluid components (e.g. A and/or B) .

Thus, a plurality of different operating states or "scenarios" for the electrical apparatus can be distinguished and, e.g., troubleshooting in the case of malfunctions is enabled or simplified. An identification of the operating state is possible by, e.g., comparing the first and second fluid component concentrations in the insulation fluid (e.g. in the second amount of the insulation fluid), i.e. in the insulation fluid being present in the compartment of the electrical apparatus) to already filled first amounts of the insulation fluid, as are e.g. known from integrated mass flow measurements.

Other optional possible operating states can e.g. be :

- intermolecular reactions between molecules of the at least two fluid components (e.g. A, B) , in par- ticular of the insulation fluid being present in the compartment of the electrical apparatus, and

- removal of at least one of the at least two fluid components (e.g. A, B) , in particular of the insu^¬ lation fluid being present in the compartment of the electrical apparatus, e.g. due to adsorption onto surfac^¬ es .

Thus, even more operating states or "scenarios" for the electrical apparatus can be distinguished and, e.g., troubleshooting in the case of malfunctions is improved or simplified further.

It should be noted that it is also possible to alternatively or additionally derive fluid component concentrations and/or a dielectric breakdown strength of the first amount of the insulation fluid that is to be filled into the compartment of the electrical apparatus and to use these values in the derivation of the operat^¬ ing state of the electrical apparatus.

In other embodiments, the method element of mixing the at least two fluid components for yielding the first amount of the insulation fluid is carried out prior to the method element of filling the first amount into the compartment of the electrical apparatus. Thus, the first mixing ratio and the first amount of the to-be- filled insulation fluid can be adapted to the target mix^¬ ing ratio and the target amount; and/or the first mixing ratio and the first amount relate to the still-to-be filled insulation fluid.

In yet another embodiment, the method element of mixing the at least two fluid components with the first mixing ratio comprises a disturbing of a laminar flow of the at least two fluid components. Thus, a turbu^¬ lence is introduced. In other words, a momentum is trans^¬ ferred to particles of the at least two fluid components, a laminar flow is consequently disturbed and a turbulence is created. As a consequence, the at least two fluid com- ponents are mixed more thoroughly, which results in a more homogeneous mixture of the first amount of the insu^¬ lation fluid. Alternatively, this can also be applied to the second amount of the insulation fluid which is al^¬ ready present in the compartment of the electrical appa^¬ ratus. Thus, the mixture in the compartment is homoge^¬ nized and the insulation strength can be improved.

It should be noted here that - although not shown in figures - it is also possible to fill the insu^¬ lation fluid components into the compartment in a paral^¬ lel but unmixed or even in a sequential manner and mix them therein, instead of or in combination with the herein disclosed premixing them outside the compartment of the electrical apparatus for yielding the first amount of the insulation fluid which is then filled into said com^¬ partment. Any of the herein discussed or known mixing means is arranged outside of the compartment, but it can also be arranged inside the compartment and/or can be connected or can be connectable to the compartment using a fluid communication means, such as e.g. a self-sealing fluid connector.

Advantageously, different means for disturb^¬ ing the laminar flow ("mixing means"), i.e. for transferring of the momentum, can be used. Specifically, the cre^¬ ation of a turbulence is achieved by means of at least one of the group of:

- a tube mixer, in particular comprising a baffle, particularly a plurality of baffles, and/or com^¬ prising a perforated dip tube,

- a rotating mixer, in particular comprising a fan, particularly a foldable fan,

- a convection mixer, in particular comprising a heater,

- a radiation mixer, in particular comprising a UV light source or an IR light source, - a volumetric mixer, in particular compris^¬ ing an expandable volume, particularly a pneumatically expandable volume,

a bypass mixer, and

combinations thereof.

In embodiments, other designs can for example also rely on an impeller or analogous moving parts, a blade, a paddle, or a helix. Also static obstacles can herein be used which cause turbulence, such as grids, meshes, deflectors, porous structures, etc.

All these mixers can also be used to destrat- ify the gas, and to remove concentration and temperature gradients .

As another aspect of the invention, an insulation fluid filler (i.e. a filling device) for filling at least a first amount of an insulation fluid into a compartment of a fluid-insulated electrical apparatus (in particular gas-insulated medium-voltage or high- voltage switchgear) as described above comprises

- a mixer for mixing at least two fluid components at or in a first mixing ratio. This mixer can be an active mixer (e.g. comprising a fan) or mixer with moving parts, or a passive mixer (e.g. relying on diffu^¬ sion and/or convection), or static mixer without moving parts (e.g. comprising static obstacles or guiding means) . Specifically, at least one of the above discussed laminar flow disturbing means can be used.

Furthermore, the insulation fluid filling de^¬ vice comprises

- a first sensor for deriving the first mixing ratio of the first amount of the insulation fluid that is to be filled into the electrical apparatus.

Furthermore, the insulation fluid filling de^¬ vice comprises

- a fluid connector for establishing a connection for insulation fluid between the insulation fluid filling device and the electrical apparatus and for transferring the first (i.e. newly mixed) amount of the insulation fluid from the insulation fluid filling device to the electrical apparatus.

Additionally, the insulation fluid filling device comprises

an analysis and control unit which is adapted and structured to carry out the method elements of a method as described above.

In an embodiment, the insulation fluid fill^¬ ing device comprises

- an interface for receiving a sensor signal indicative of a second mixing ratio of a second amount of the insulation fluid that is already in the compartment of the electrical apparatus. As an example, the electri^¬ cal apparatus comprises a second sensor for deriving this second mixing ratio of the second amount of insulation fluid and transmits a sensor signal indicative of this second mixing ratio via the interface to the insulation fluid filling device.

In another embodiment, the analysis and con^¬ trol unit of the insulation fluid filling device compris^¬ es a computer program element comprising computer program code means for, when executed by a processing unit of the insulation fluid filling device, implementing a method as described above.

The described embodiments and/or features similarly pertain to the apparatuses and the methods. Synergetic effects may arise from different combinations of these embodiments and/or features, although they might not be described in detail.

Brief Description of the Drawings

The invention and its embodiments will be more fully appreciated by reference to the following de^¬ tailed description of presently preferred but nonetheless illustrative embodiments in accordance with the present invention when taken in conjunction with the accompanying drawings .

Fig. 1 shows an insulation fluid filling apparatus or filler 30 according to the invention as well an electrical apparatus 1 ;

Fig. 2 shows a schematic of an optical fluores^¬ cence sensor;

Fig. 3 shows a schematic of an optical absorb^¬ ance sensor 100, 200;

Fig. 4 shows a schematic of a gas chromatograph 100, 200 comprising two columns and a mass spectrometer behind one column;

Fig. 5 shows a chromatogram illustrating the separation of insulation fluid components "N2/O2" and "C5" and "C6";

Fig. 6 shows two chromatograms illustrating loss of a specific fluid component "gas 4";

Fig. 7 shows two chromatograms illustrating de^¬ tection of a contaminant;

Fig. 8 shows an absorption diagram illustrating characteristic optical absorbance signatures of insula^¬ tion fluid components "C5" and "C6" in the infrared re^¬ gion;

Fig. 9 shows absorption diagrams illustrating the detection of a contaminant "HF" of the insulation fluid;

Fig. 10 shows an absorption diagram in the near UV range for "acetone", "C5", and "C6";

Fig. 11 shows an infrared absorption spectrum illustrating characteristic optical absorbance signatures of an insulation fluid component "C5";

Fig. 12 shows infrared absorption spectra illus^¬ trating the characteristic optical absorbance signatures of the insulation fluid component "C5" of Fig. 11 as well as characteristic optical absorbance signatures of a con^¬ taminant "CF₄"; Fig. 13 shows infrared absorption spectra illus^¬ trating the characteristic optical absorbance signatures of the insulation fluid component "C5" of Fig. 11 as well as characteristic optical absorbance signatures of a con^¬ taminant hexafluoropropene "CF3CF=CF2";

Fig. 14 shows infrared absorption spectra illus^¬ trating the characteristic optical absorbance signatures of the insulation fluid component "C5" of Fig. 11 as well as characteristic optical absorbance signatures of a con^¬ taminant heptafluoropropane "CF3CFHCF3";

Fig. 15 shows a schematic of an optical absorb^¬ ance measurement comprising a spectrometer;

Fig. 16 shows an "optical absorption" vs. "C5 pressure" diagram as recorded by an optical setup of Fig. 15;

Fig. 17 shows a schematic of an optical absorb^¬ ance measurement comprising a non-dispersive photodetec- tor and a band-pass filter;

Fig. 18 shows a "transmitted intensity" vs. "C5 pressure" diagram as recorded by an optical setup of Fig. 17;

Fig. 19 illustrates the dependence of UV absorp^¬ tion of the insulation fluid component "C5" on different insulation fluid mixture ratios and/or pressures;

Fig. 20 shows a photoionization detector 100, 200;

Fig. 21 shows a pTp detector 100, 200 in a bypass configuration;

Fig. 22 similarly to Fig. 10 shows reference da^¬ ta for acetone and an insulation fluid component "C6" and absorption diagrams for the insulation fluid components "C5" and "C6" for wavelengths between 200 nm and 400 nm;

Fig. 23 similarly to Fig. 18 shows a "trans- fleeted intensity" vs. "C5 pressure" diagram as well as a relative error dl/I;

Fig. 24 similarly to Fig. 23 shows a "trans- fleeted intensity" vs. "C5 pressure" diagram as well as a relative deviation to a theoretical fit dl/I; Fig. 25 shows absorption spectra of the insula^¬ tion fluid components "C5", "O2", and "CO2", and of the contaminant "hexafluoropropene" for wavelengths between 200 nm and 500 nm;

Fig. 26 shows absorbance spectra for insulation fluid components "C5", "C6", "C7", and "C0₂" for wave^¬ lengths between 1350 nm and 1950 nm;

Fig. 27 shows a zoomed part of the spectra of figure 26 for wavelengths between 1850 nm and 1950 nm;

Fig. 28 shows an absorption spectrum of the insulation fluid component "C5" for wavenumbers between 500 cm^-1 and 2000 cm^"1;

Fig. 29 shows the C5 spectrum of Fig. 28 overlapped with spectra of insulation fluid components "CO2" and contaminants "H2O", "CF4", "hexafluoropropene", and "heptafluoropropane", and a relative transmittance of carbon monoxide "CO" as for wavenumbers between 2900 cm^~l and 500 cm^~l;

Fig. 30 similarly to Fig. 1 shows an insulation fluid filling device 30 and an electrical apparatus 1, wherein the insulation fluid filling device 30 comprises an additional gas blender 339 with a tube mixer 370 and wherein the electrical apparatus 1 comprises a gas blend^¬ er 339 with a foldable fan 305a;

Fig. 31 shows a tube mixer 370 with a plurality of baffles 371;

Fig. 32 shows a tube mixer 380 with a perforated dip tube 381;

Figs. 33a and 33b show a rotating mixer 305 arranged in a compartment 2, wherein the rotating mixer 305 comprises a foldable fan 305a;

Fig. 34 shows a radiation mixer 390 arranged in a compartment 2, wherein the radiation mixer 390 compris^¬ es a light source 391;

Fig. 35 shows a convection mixer 307 comprising a heater 310; Figs. 36a and 36b show a volumetric mixer 400 arranged in a compartment 2, wherein the volumetric mixer 400 comprises a pneumatically expandable volume 401;

Fig. 37 shows a bypass mixer 410 connected to a compartment 2;

Figs. 38a - 38e show different optical and mounting options for a first and/or second optical sensor 100,200; and

Fig 39 shows an optical sensor 100, 200 with an optical measurement channel and an optical reference channel .

Modes for Carrying Out the Invention

Description of the Figures :

Fig. 1 shows a schematic of an insulation fluid filler or filling device 30 according to the invention. The insulation fluid filler 30 operates at a filling temperature Tfj__]__]_ which corresponds to the ambient tempera^¬ ture of, e.g., 20°C. The insulation fluid filler 30 com^¬ prises two fluid component reservoirs 301 and 302 for holding fluid components A and B, respectively.

Here, fluid component A comprises a perfluoro- ketone C5 which is a liquid at room temperature and at a pressure of 1 bar. A pump 303 conveys a stream (upper bold arrow) of liquid fluid component A to a mixer 31. The flux of the liquid fluid component A is monitored and controlled by a mass flow meter and regulator 35. In other words, the mass flow meter and regulator 35 measures and regulates the mass flux of fluid component A that en^¬ ters the mixer 31. Information about the mass flux is transmitted to an analysis and control unit 34 of the in^¬ sulation fluid filler 30 and mass flux regulation commands are received from this analysis and control unit 34. With the mass flux information, also a closed-loop operation of the pump 303, e.g. a variable pump speed controlled by the mass flow meter and regulator 35 that maintains a desired mass flux, is possible (indicated by the upper curved arrow in Fig. 1) .

Fluid component B consists of a pressurized gas mixture consisting of, e.g., 95% carbon dioxide and 5% oxygen at a total overpressure of 15 bars and is gaseous at ambient temperature. Due to the overpressure and a pressure gradient, a flow of gaseous fluid component B (lower bold arrow) automatically reaches the mixer 31 of the insulation fluid filler 30 after passing through a pressure regulator 304 which down-regulates the fluid component pressure to 10 bars and through a mass flow me^¬ ter and regulator 36 which measures and regulates the mass flux of fluid component B that enters the mixer 31. This information is again transmitted to the analysis and control unit 34 and regulation control commands are re^¬ ceived from the analysis and control unit 34 of the insu^¬ lation fluid filler 30. As discussed above, closed loop operation of the pressure regulator 304 and the mass flow meter and regulator 36 is possible (lower curved arrow) .

In the mixer 31 the liquid fluid component A is vaporized by a heater 310 and mixed with the gaseous flu^¬ id component B in a turbulent jet mixing zone (curved ar^¬ rows in 31) . Thus, the first amount Ml of insulation flu^¬ id 10 is yielded. Due to the controlled amounts of fluid components A and B that are mixed in the mixer 31, the first amount of the insulation fluid 10 has a first mix^¬ ing ratio Rl, or - equivalently - the first fluid compo^¬ nent A has a first concentration c^ and the second fluid component B has a second concentration eg .

It should be noted in this respect that the mix^¬ ing and thus yielding of the first amount Ml, measuring of mixing ratios, and filling of the first amount Ml of the insulation fluid 10 can be continuous processes (see above) .

Then, the first amount Ml of the insulation flu^¬ id 10 is transferred to the compartment 2 of the electri- cal apparatus 1 via a fluid connector 33 of the insula^¬ tion fluid filler 30 and via suitable tubing (bold ar^¬ rows) . Before the first amount Ml of the insulation fluid 10 leaves the insulation fluid filler 30, however, the first mixing ratio Rl and/or the first amount Ml is de^¬ rived using a first sensor 100, which is an optical fluo^¬ rescence sensor as described in Fig. 2 below and/or a mass flow sensor. Alternatively, an optical absorbance sensor from Fig. 3, a photoionization sensor from Fig. 20, a pTp sensor from Fig. 21, or any other suitable sensor or sensor combination can be used. The information indicative of the first mixing ratio Rl from the first sensor 100 is transmitted to the analysis and control unit 34 and compared to the mixing ratio as determined by the mass fluxes of the single fluid components A and B (see above) . Thus, the detection of failure states such as condensation in simplified.

Furthermore, before filling of the first amount Ml into the compartment 2, the second mixing ratio R2 and the second amount M2 of the insulation fluid 10 which is already in the compartment 2 of the electrical apparatus are also measured by a second sensor 200 (e.g. comprising any of the sensors mentioned above, or combinations thereof) and transmitted to the analysis and control unit 34 via an interface 32.

As soon as the first amount Ml reaches the com^¬ partment 2 of the electrical apparatus 1, the transferred first amount Ml of the insulation fluid 10 with the first mixing ratio Rl is then mixed with the already present second amount M2 of the insulation fluid 10 which has (before mixing) a second mixing ratio R2 (see above) . This mixing is accelerated by a circulator 305. The cir^¬ culator 305 homogenizes the density and/or the mixture of the insulation fluid components A and B.

By filling the first amount Ml into the compart^¬ ment 2, the total amount of insulation fluid 10 in the compartment 2 of the electrical apparatus is increased. During all these steps, the first mixing ra^¬ tio Rl of the first amount Ml of the insulation fluid 10 (that has been transferred into the compartment 2) is controlled such that condensation temperatures of the fluid components A and B of the insulation fluid 10 are below the filling temperature Tf-_j__]__]_. In other words, it is ensured that no condensation takes place.

The first mixing ratio Rl and the first amount Ml of the insulation fluid 10 that is transferred into the compartment 2 are further controlled using

* the second mixing ratio R2 and the second amount M2 of insulation fluid 10 in the compartment 2 (before filling) and using

* a target mixing ratio R (e.g. 5.5 % per- fluoroketone C5, 94.5 % CC^-C^-mixture) and a target amount M (e.g. equivalent to 7.7 bar at 20°C) of insula^¬ tion fluid 10 which are to be reached after completion of the filling process.

In other words, knowing the second mixing ratio R2 and the second amount M2 of insulation fluid 10 in the compartment 2, the analysis and control unit 34 de^¬ rives how much (i.e. the first amount Ml) insulation flu^¬ id 10 at which mixture (i.e. the first mixing ratio Rl) needs to be filled into the compartment 2 so that the target amount M at target mixing ratio R are reached.

Furthermore, after filling another measure^¬ ment of R2 and M2 are taken and this information from the second sensor 200 enables the analysis and control unit 34 to derive concentrations c^ and eg of the first and second fluid components A and B in the compartment 2. Using these data, the analysis and control unit 34 can then optionally in addition derive an operating state 0 of the electrical apparatus 1. Specifically, dielectric break-down strength of the insulation fluid 10 is derived according to

^Ebd = S{c_A,c_B) ∑CiE_{cri i}

i=A,B with Ecrit,A ^anc^ ^crit,B being known and preset fluid-component-specific critical field strengths of the fluid components A and B. S (c^, eg) is a known and preset synergy parameter and i is an index running over the fluid components A and B. From this dielectric breakdown strength E_j^ of the insulation fluid 10, the operating state 0 is derived which is indicative of the availabil^¬ ity of the electrical apparatus for normal operation, e.g. current conduction or high-voltage switching.

With this information available, the insulation fluid filler 30 can stop the filling procedure as soon as a threshold of the dielectric breakdown strength Ej^ is reached or exceeded.

Fig. 2 shows a schematic of an optical fluores^¬ cence sensor 100, 200 as it can be used in the insulation fluid filler 30 of Fig. 1 or in the electrical apparatus 1. Here, fluorescence excitation light from a light source 23 (e.g. a laser, LED, VCSEL) is collimated by a lens 24 and passes a dichroic beam splitter 28. It then passes through the wall of a glass tube whose inner vol^¬ ume is filled with to-be-measured insulation fluid 10, either from a by-pass-arrangement at an insulation fluid tubing or from an extraction of a small amount (e.g. 1 ml at 1 bar at room temperature) of insulation fluid 10. This glass tube forms the measurement cell 21 of the op^¬ tical sensor 100. In another application, the measurement cell 21 of the optical sensor 200 can be formed by the compartment 2 of the electrical apparatus 1 itself, i.e. fluorescence is then measured inside the compartment 2. The fluorescence excitation light then excites molecules of the insulation fluid 10 and resulting fluorescence emission light is in part travelling back towards the beam splitter 28. To increase light collection efficiency, a mirror 27 is arranged on a side of the glass tube opposing the beam splitter 28. Fluorescence emission photons are then deflected by the dichroic beam splitter 28, pass an emission filter 26 (which blocks leftover excita- tion light), and are focused onto a detector 25 (e.g. an avalanche photodiode or a photomultiplier tube) by a col^¬ lection lens 24. The electrical fluorescence signal (in^¬ dicative of c^ and eg) from the detector 25 is then read out, optionally preprocessed (not shown) , and transmitted to the analysis and control unit 34 of the insulation fluid filler 30 for further processing. It should be noted that different light sources and optical setups are possible, e.g. monochromatic light at one or more wave^¬ lengths (e.g. at 305 nm) from one or more laser (s) 23, narrow spectrum light from a narrow band LED light source 23 (e.g. in the near UV range), or polychromatic or white light (optionally with a monochromator such as a grating) from a conventional light source 23. It is also possible to use different optical sensors 100, 200 for the differ^¬ ent fluid components A and B.

Fig. 3 shows a schematic of an optical absorb- ance sensor 100, 200 as it can be used in the insulation fluid filler 30 from Fig. 1 or in the electrical appa^¬ ratus 1. Here, light is monochromized inside a light source 23 by a grating (only schematically shown) and split into two beams by a e.g. 50:50 beam splitter 28. One light beam is guided through a reference cell 21a with a known absorbance per wavelength. The other light beam is guided through a measurement cell 21 of the opti^¬ cal sensor which comprises the insulation fluid 10 which is to be measured optically. Again, bypass configurations (arrows) or extractions of small amounts of insulation fluid 10 (not shown) are possible. After traveling through the measurement cell 21 or the reference cell 21a, respectively, both light beams are propagated through band-pass filters 29 and focused onto photodetec- tors 25 (e.g. avalanche photodiodes) by lenses 24. By tuning the wavelength from the light source, an absorb^¬ ance spectrum of the insulation fluid 10 over wavelength which is indicative of the fluid component concentrations c^ and eg (or the mixing ratio) is measured. As discussed above with regard to the fluorescence optical sensor 100, 200, it should again be noted that different light sources and optical setups are possible as it is obvious to the person skilled in the art.

Advantages of using an optical sensor 100, 200 for deriving the fluid component concentrations c^, C-Q, (or the mixing ratios of the insulation fluid 10) are:

(i) high specificity to individual fluid components,

(ii) high sensitivity, (iii) broad applicability to any insulation fluid mixture comprising optical absorption or fluorescence, and (iv) nonextractive measurement princi^¬ ple, i.e. no insulation fluid needs to be removed.

Fig. 4 shows a schematic of a gas chromatograph 100, 200 as it can be used in the insulation fluid filler 30 from Fig. 1 or in the electrical apparatus 1. The gas chromatograph 100, 200 comprises a carrier gas reservoir 331 and a sample injector 332 which accepts a small amount (e.g. 0.5 ml at 1 bar at 20°C) of the insulation fluid 10 from the first or the second amount of the insu^¬ lation fluid 10. This insulation fluid 10 is then inject^¬ ed into the flowing carrier gas and propagated through two columns 333 onto detectors 334 (e.g. thermal conduc^¬ tivity detector, flame ionization detector, or electron capture detector, electron impact ionization detector) . Due to different retention times of the fluid components A and B with a stationary phase in the columns, the con^¬ centrations values c^ and CQ (or the mixing ratio) can be measured. For example, the column Fluorocol from the com^¬ pany Supelco ( Sigma-Aldrich) can separate a mixture of 2/O2 and C5/C6. A mass spectrometer 335 is arranged behind one column 333 for carrying out an additional mass spectrometric measurement for detecting and/or discriminating contaminants (i.e. undesired substances) in the insulation fluid 10.

Advantages of using a gas chromatographic meas^¬ urement for deriving the fluid component concentrations c^ and CQ (or the mixing ratios of the insulation fluid 10) are: (i) good separation and quantification capabil^¬ ity to individual fluid components, (ii) very good sensi^¬ tivity, and (iii) the ability to diagnose unknown contam^¬ inants, e.g. by optionally using an additional mass spec^¬ trometer (see below) .

Fig. 5 shows a chromatogram (i.e. a chromato- graph detector signal as a function of retention time in the column) illustrating the separation of insulation fluid components "N2/O2" (technical air) and "C5" and "C6", which all appear as single separate peaks in the chromatogram. A Fluorocol column from the manufacturer Supelco can, e.g., be used for separation.

Fig. 6 shows two chromatograms illustrating electrical stress-induced loss of a specific fluid compo^¬ nent (gas 4, peak drops, see arrow and dotted lines) . An FC column can, e.g., be used for such a measurement.

Fig. 7 shows two chromatograms illustrating de^¬ tection of a contaminant of the insulation fluid 10. An additional peak (arrow, dotted circle) appears after a fresh insulation gas mixture (fresh gas mixture) has un^¬ dergone electrical stress (aged gas mixture) . An FC col^¬ umn can, e.g., be used for such a measurement.

Fig. 8 shows an infrared absorption spectrum illustrating characteristic signatures of insulation fluid components C5 and C6 in the infrared region. The use of infrared spectroscopy offers an easy, specific, and accu^¬ rate method for the determination (type and concentra^¬ tion) and monitoring of the fluid components that make up the insulation fluid 10. Many molecules, such as e.g C5 (i.e. C5-fluoroketone) and C6 (i.e. C6-fluoroketone) show characteristic spectral signatures (spectral fingerprint) in the infrared region which are, e.g., due to vibration^¬ al excitation. Specifically, measurements of the bands labeled C5-signature and C6-signature in the spectrum of the insulation fluid unambiguously indicate the presence and allow the concentration determination of C5 and/or C6, respectively. Note that the spectrum in the region 1200cm^~l to 1350 cm^~l is partially saturated.

Fig. 9 shows absorption diagrams illustrating the detection of a contaminant "HF" in the insulation fluid 10. Specifically, after electrical stress induced aging of an insulation fluid component "C5", clear ab^¬ sorption signatures of the contaminant "HF" can be seen in the spectrum. An HF reference spectrum is given for comparison. Using calibrated HF spectra, the concentra^¬ tion of the contaminant HF can also be derived. Analogous procedures exist for other contaminants.

Fig. 10 shows an absorption diagram in the UV range for acetone, C5, and C6 together with reference da^¬ ta for acetone.

Fig. 11 shows an infrared absorption spectrum illustrating characteristic optical absorption signatures of an insulation fluid component "C5". It is found that the insulation fluid component "C5" shows absorption peaks that do not overlap with spectral signatures of contaminants (indicated by, e.g., the arrow, see below) . Therefore, by selecting such an appropriate spectral sig^¬ nature, the insulation fluid component "C5" can be unam^¬ biguously monitored without cross-sensitivity to contami^¬ nants .

Fig. 12 shows infrared absorption spectra illus^¬ trating the characteristic optical absorbance signatures of the insulation fluid component "C5" of Fig. 11 as well as characteristic optical absorbance signatures of a con^¬ taminant "CF4" (as indicated by the arrow) . Because these peaks do not overlap, they allow an unambiguous detection of the contaminant "CF4", even in the presence of "C5".

Fig. 13 shows infrared absorption spectra illus^¬ trating the characteristic optical absorbance signatures of the insulation fluid component "C5" of Fig. 11 as well as characteristic optical absorbance signatures of a con^¬ taminant hexafluoropropene "CF3CF=CF2" (as indicated by the arrows) . Similar to the situation in Fig. 12, these signatures allow an unambiguous detection of the contami^¬ nant "CF₃CF=CF2", even in the presence of "C5".

Fig. 14 shows infrared absorption spectra illus^¬ trating the characteristic optical absorbance signatures of the insulation fluid component "C5" of Fig. 11 as well as characteristic optical absorbance signatures of a con^¬ taminant heptafluoropropane "CF3CFHCF3". Similar to the situation in Figs. 12 and 13, these signatures allow an unambiguous detection of the contaminant heptafluoropro^¬ pane "CF3CFHCF3", even in the presence of "C5".

Fig. 15 shows a schematic of an optical absorb^¬ ance measurement comprising a spectrometer, i.e. a wave^¬ length-discriminating or dispersive photodetector 25. Light from a light source 23 (e.g. a deuterium light source) is propagated through a measurement cell 21 com^¬ prising the insulation fluid 10. Then, a part of the light that has not been absorbed in the measurement cell 21 is detected by photodetector 25.

Fig. 16 shows an "optical absorption" vs. "C5 pressure" diagram as recorded by an optical setup of Fig. 15. In other words, absorption A in arbitrary units is plotted versus pressure p in mbar of pure insulation flu^¬ id component C5. The graph shows measured data (diamonds) together with a linear fit line (slope a = 0.009872, offset A_offset = -0.019615) as well as relative errors of measured values compared to the fit. Each data point cor^¬ responds to the area of the absorption peak for the re^¬ spective "C5" concentration. The inset shows a typical absorption spectrum (i.e. wavelength in nm on x-axis, wavelength-dependent absorption in arbitrary units a.u. on y-axis, here for a partial pressure of C5- perfluoroketone of e.g. p(C5) = 91.5 mbar) of insulation fluid component "C5" where the hatched region (240 nm < λ < 350 nm) represents the integration area which is used to measure the (integral) absorption A. The relative er^¬ ror dA/A ≤ ±2% demonstrates that concentration determina^¬ tion of insulation fluid component "C5" is possible with high sensitivity and high precision, using simple cali^¬ bration methods. The used equipment comprises: Light source 22: Deuterium light source, DT-MINI-2-GS, Ocean Optics. Gas cell 21: Stainless steel cell (with optical path length 47cm) "Expando" from Solvias. Fiber-optic ca^¬ bles 23: length 0.5 m, core diameter 600 μιτι, UV- 0.5m/1204191 from Solvias. Detector 25: High-resolution spectrometer, HR4000 from Ocean Optics.

Fig. 17 shows a schematic of an optical absorb- ance measurement comprising a non-dispersive photodetec^¬ tor 25 and a band-pass filter 29. Light from a light source 23 (e.g. a deuterium light source) is propagated through a measurement cell 21 comprising the insulation fluid 10. Then, a part of the light that has not been ab^¬ sorbed in the measurement cell 21 propagates through a band-pass filter 29 and is detected in a non-wavelength- discriminating-fashion by photodetector 25.

Fig. 18 shows a "transmitted intensity" vs. "C5 pressure" diagram as recorded by an optical setup of fig. 17. In other words, transmitted intensity I in p.W is plotted versus pressure p in mbar of pure insulation flu^¬ id component C5. The graph shows measured data (diamonds) together with exponential fit (line, Lambert-Beer law; with offset intensity I_offset = 21.9 μΐ/ί , intensity coef^¬ ficient I_0 = Io = 35.8 μΐ/ί , and exponential coefficient ε = 0.0032 1/mbar) and relative error dl/I of measurement I compared to fit. Each data point corresponds to the to^¬ tal integrated intensity I measured by the silicon photo^¬ detector 25. Most of the light of the Deuterium light source which has a larger wavelength than the insulation fluid components' absorption peaks is blocked by a filter (the light which is not blocked contributes to the offset -'-offset) · ^T^^e relative error dl/I ≤ ±1% demonstrates that concentration of insulation fluid component "C5" can be determined with high sensitivity and high precision.

The used equipment comprises: Light source 22: Deuterium light source, DT-MINI-2-GS, Ocean Optics. Gas cell 21: Stainless steel cell (with optical path length 47cm) «Expando» from Solvias. Fiber-optic cables 23: length 0.5 m, core diameter 600 μιτι, UV-0.5m/1204191 from Solvias. Filter 29: UG-11 from Schott. Detector 25: Si Photodiode, UV-818 from Newport.

Fig. 19 illustrates the dependence of UV absorp^¬ tion of the insulation fluid component "C5" on insulation fluid mixtures and/or on insulation fluid pressure. In other words, the absorption of pure insulation fluid component "C5" in gaseous form and of two different insula^¬ tion fluid mixtures consisting of insulation fluid compo^¬ nent "C5" and insulation fluid component of, e.g., "syn^¬ thetic air" with different mixture ratios and total pres^¬ sures is acquired with the optical setup of Fig. 19. The absolute amount of insulation fluid component "C5" for all 3 samples (i.e. 3 graphs, only two are distinguish^¬ able here) is constant: p(C5)=91.5 mbar. The amount of synthetic air differs between 0 and 5775 mbar. The graph illustrates that absorption values for all 3 insulation fluid mixtures do not or not significantly deviate for high or low pressures when synthetic air is added to the insulation fluid component "C5". This shows that a con^¬ centration of an insulation fluid component "C5" using a C5 UV-absorption peak can be determined independently from the admixture of synthetic air. No effect is ob^¬ served even up to a total insulation fluid pressure of p_tot = 8.9 bar (data not shown) .

The used equipment comprises: Light source 22: Deuterium light source, DT-MINI-2-GS, Ocean Optics. Gas cell 21: Stainless steel cell (with optical path length 47cm) "Expando" from Solvias. Fiber-optic cables 23: length 0.5 m, core diameter 600 μιτι, UV-0.5m/1204191 from Solvias. Detector 25: High-resolution spectrometer, HR4000 from Ocean Optics.

Fig. 20 shows a photoioni zation detector 100, 200 as it can be used in the insulation fluid filling de^¬ vice 30 from fig. 1 or in the electrical apparatus 1. Here, UV light from a light source 23 is propagated through the measurement cell 21 to ionize molecules of the insulation fluid 10. The degree of ionization can then be measured as an ion current by electrodes 22 and thus the mixing ratio can be derived.

Fig. 21 shows a pTp detector 100, 200 in bypass configuration, as it can be used in the insulation fluid filling device 30 from Fig. 1 or - in a different config^¬ uration - also inside the electrical apparatus 1. Here, the pressure p, the temperature T, and the density p of a static, isothermal (cf. the heaters, diagonal lines) gas sample of the insulation fluid 10 are measured and the fluid component concentrations c^ and eg (or the mixing ratio) can thus be determined. The valves 337 are period^¬ ically opened and closed to measure a fresh insulation fluid sample. Alternatively, such a pTp detector 100, 200 can also be used in an in-line configuration. With regard to this Fig. 21, this means that one valve 337 could be closed permanently or alternately to the other valve 337, or that only a single flange connection would be re^¬ quired .

Fig. 22 similarly to Fig. 10 shows reference ab^¬ sorption data for acetone and the insulation fluid compo^¬ nent "C6" on the left hand side and measured absorption diagrams for the insulation fluid components "C5" and "C6" on the right hand side. An optical path length of 20 mm at temperatures in the range between 23.5°C and 24.2°C was used. Specifically, it can be concluded from the spectral measurements on the right hand side of figure 22 that the absorption peak of C5 is shifted by about 5 nm to the blue and that it is reduced by a factor of around 0.6. No absorption features were apparent in the wave^¬ length regime between 400 nm and 1100 nm (data not shown for clarity) .

Fig. 23 similarly to Fig. 18 shows a "trans- fleeted intensity" vs. "C5 pressure" diagram. In other words, transmitted intensity I in p.W is plotted versus pressure p in mbar of an insulation fluid component C5. The graph shows measured data (diamonds) together with an exponential fit (line, Lambert-Beer law) and relative er^¬ rors dl/I of measurements I compared to the fit (right y- axis) . The results demonstrate a very good agreement of the measured data with theory (Lambert-Beer law) with relative errors of smaller than 0.2% over a large C5 con^¬ centration range. A sensitivity of the measurement at a typical nominal C5 concentration (around 400 mbar) was evaluated as dI/dp=-0.5 μνί/mbar (cf. the black triangle ΔΙ/Δρ) . Assuming a signal fluctuation (due to the stabil^¬ ity of the light source, detector and interfaces in the optical system) of dl/l= ±1%, this value results in an overall pressure sensitivity of ±6 mbar.

The used equipment comprises: Light source: Sandhouse 293 nm LED from Ocean Optics. Gas cell: Optical insertion probe 661.686-UVS from Hellma with an optical path length of 20 mm inserted into a pressure vessel from Swagelok. Detector: Si-Photodiode, 818-UV from Newport in combination with power meter 2931-C.

Fig. 24 similarly to Fig. 23 shows a "trans- fleeted intensity" vs. "C5 partial pressure" diagram. In other words, transmitted intensity I in V [here measured in volts as detector output, 1 V corresponds to an inten^¬ sity of 17.72 nW] is plotted versus pressure p in bar of an insulation fluid component C5. The graph shows meas^¬ ured data (open circles) together with an exponential fit (line, Lambert-Beer law) and relative deviations dl/I be^¬ tween the measured data and the theoretical fit (trian^¬ gles) . The results demonstrate a very good agreement of the measured data with theory (Lambert-Beer law) with a relative deviation of smaller than ±0.5% over a large C5 concentration range. The sensitivity of this measurement at a typical nominal C5 concentration (e.g., around 400 mbar) was evaluated as dl/dp= -0.4 V/bar = -7.1 nW/bar . The signal-to-noise ratio is approximately SNR = I/dl = 1000 at an integration time of ~ls. These values result in a pressure sensitivity of Δρ = ΔΙ/ (dl/dp) = [±I (400mbar) / (SNR) ] / ( 0.4V/bar) = ±0.4 mbar.

The used equipment comprises: Light source: LED LLS290, ρ_θ3]^ = 293 nm, Ocean Optics. Gas cell: Optical insertion probe TI-300-UV-VIS from Ocean Optics with an optical path length of 20 mm inserted into a pressure vessel with a volume of 3.6 liters. Detector: Si- photodiode 818-UV from Newport in combination with power meter 2931-C. Measurements were carried out at tempera^¬ tures between 25.3°C and 25.6°C.

Fig. 25 shows absorption spectra of the insula^¬ tion fluid components "C5", "O2", and "CO2", and of the contaminant "hexafluoropropene" for wavelengths between 200 nm and 500 nm. The spectra of O2 and CO2 are from reference datasets while those of C5 and hexafluoropro- pene were measured. The spectral signatures of the con^¬ taminants "CF4" and "heptafluoropropane" were also meas^¬ ured and did not show any spectral overlap with the UV absorption of "C5" (data not shown for clarity) . As a conclusion, an optical absorbance measurement in the UV range is not hampered by these contaminants to a large degree, in particular if a narrow-band light source (e.g., an LED at around 300 nm with a full width at half maximum FWHM = 12 nm) is used.

Fig. 26 shows absorbance spectra for insulation fluid components "C5", "C6", "C7", and "CO2" in the near infrared (NIR) region, specifically between 1350 nm and 1970 nm. A spectrometer from Axsun (Analyzer XLP910) was used for the spectral characterizations. The spectrometer is capable of recording the absorbance in the above- referenced wavelength region with a resolution of 3 cm^~l which is equivalent to a wavelength resolution of about 1 nm.

Fig. 27 shows a zoomed part of the spectra of Figure 26 for wavelengths between 1850 nm and 1950 nm. As evident from figures 26 and 27, the insulation fluid component "C6" shows an absorption peak at 1891 nm. The insulation fluid components "C5" and "C7" show two overlapping absorption peaks at 1873 nm / 1882 nm and 1894 nm/1903 nm, respectively. These absorption bands are presumably second overtones of a C=0 stretch vibration. CO2 data on the right hand side of figures 26 and 27 is reference data for comparison. Please note that the data in figure 27 is normalized to equal number density of molecules (i.e. P_Q5=974 mbar) .

Since the insulation fluid can be a gas mixture of insulation fluid components including "C5", "CO2", and "O2" and since a specific analysis method for "C5" is advantageous, potential cross interference of the C5 absorption bands with those of CO2 in the NIR region were checked for. O2 does not have a permanent dipole moment and therefore does not exhibit a vibration spectrum in the NIR region. CO2 shows a significant band at 1960 nm and around 2000 nm. These features, however, do not overlap with the C5 carbonyl band discussed above.

Thus, in summary, the identified absorption bands of C5 at 1873 nm / 1882 nm represent a possible basis for a C5-specific NIR optical absorption measurement.

Fig. 28 shows an absorption spectrum of the insulation fluid component "C5" for wavenumbers between 500 cm^~l (wavelength 20 μιη) and 2000 cm^~l (wavelength 5 μιη) , i.e. in the mid infrared (MIR) region.

For acquisition of the MIR spectra, a Fourier Transform Infrared Spectroscope (FTIR spectrometer, Digi- lab FTS-40 Pro, 0.5 cm^-1 resolution, 400-4000 cm^-1, 11.25 m path length, cell volume 0.005 m^ = 5 liters) was employed. To obtain a good spectral separation of individual peaks and to avoid saturation, the pressures were reduced to several thousand Pa.

The infrared absorption of the insulation fluid component C5 in the spectral region between 4000 cm^~l (2.5 μιη) and 600 cm^~l (16.6 μιη) shows absorption bands between 600 cm^~l and 1900 cm--'-. These absorption band presumably stem from vibrational transitions of the car- bon framework (below about 1100 cm ^l) , of the C-F bonds (1100 cm^-1 to 1400 cm ^λ) , and of the C=0 carbonyl group (1800 cm^~l) . In the spectral region from 2000 cm^~l to 4000 cm^~l, C5 does not display any absorption features (data not shown) .

Some of the above discussed bands can be used for quantification of the insulation fluid component C5 by IR absorption, for example by using a broad-band, in^¬ candescent light source and a notch filter permitting the transmission of only a selected wavelength that interrogates a narrow spectral region in which only C5 absorbs. For a measurement like this to work, cross-sensitivity to other insulation fluid components and contaminants that may be present should be excluded. As an example, insula^¬ tion fluid components O2 and CO2 can be present. In addi^¬ tion, water vapor "H2O" and the contaminants "HF", "CF4", "hexafluoropropene", and "heptafluoropropane" may appear (note that there may be further contaminants that are not shown here) . Further note that molecular oxygen does not have an infrared spectrum due to its lack of a permanent dipole moment (data not shown) .

Fig. 29 shows the C5 spectrum of Fig. 28 overlapped with spectra of insulation fluid components "C02" (Fig. 29a), of "H2O" (Fig. 29b), and of contaminants "CF4" (Fig. 29c), "hexafluoropropene" (Fig. 29d), and "heptafluoropropane" (Fig. 29e) as a function of wave- numbers between 2000 cm^~l and 500 cm--'-. Furthermore, Fig. 29f shows a relative transmittance of "CO" as a function of wavenumbers between 2900 cm^~l and 500 cm^~l as a reference spectrum.

It is noted that infrared absorption measure^¬ ments are well suited to track e.g. water vapor content without interference of the C5 content. This is because here are infrared absorption bands of water available that do not overlap with those of C5 and these can be ad^¬ dressed for monitoring (see Fig. 29b) . Not shown in Fig. 29 is the data of the contami^¬ nant "HF": The lowest vibrational transition of HF lies around 4000 cm^~l and pure rotational transitions above 500 cm^~l carry extremely small intensities at ambient temperatures .

The use of infrared spectroscopy for C5 detec^¬ tion is therefore herewith proven. However, great care shall be taken to avoid cross-sensitivities to other gas species by choice of appropriate spectral signatures. One suitable band for cross-interference-free C5 characteri^¬ zation is around 990 cm--'-. In addition, the absorption intensities and band shapes may vary slightly with tem^¬ perature because of the temperature-dependence of the population of the rovibrational ground states given by the Boltzmann distribution. These effects should be taken into account.

Fig. 30 similarly to Fig. 1 shows an insulation fluid filling device 30 and an electrical apparatus 1. The devices and their mode of operation are very similar to the ones described above with regard to Fig. 1, with the exception that the fluid filling device 30 additionally to the mixer 31 comprises a gas blender 339 with a tube mixer 370 (see Fig. 31 below for details of the tube mixer 370) . In other words, after being mixed in the mixer 31 (wherein the heater 310 here introduces a convec^¬ tion to the insulation fluid components A and B and thus additionally acts as a convection mixer 307), the first amount Ml of the insulation fluid 10 passes through the gas blender 339 of the fluid filling device 30 which fur^¬ ther improves the mixing of the first component A and the second component B in the first amount Ml. Thus, a better insulation performance can be achieved in the electrical apparatus 1.

As an alternative or additional option, the gas blender 339 can also be arranged in the fluid connection between the fluid filling device 30 and the electrical apparatus 1 (dotted rectangle 339) or even in the elec^¬ trical apparatus 1 itself (not shown) .

In Fig. 30, as another difference to Fig. 1, the electrical apparatus 1 now comprises, instead of the

(fixed and relatively small) circulator 305 that is inter alia used for homogenizing the insulation fluid 10 before sensing its mixing ration R2 (see above) , an additional gas blender 339 with a (e.g. large) foldable fan 305a

(see Fig. 33a and 33b below for details of the arrange^¬ ment) . This additional gas blender 339 is used to thor^¬ oughly mix the first amount Ml of insulation fluid (after being filled into the compartment 2 of the electrical ap^¬ paratus 1) with the second amount M2 of the insulation fluid 10 which was already present in the compartment 2 before filling. Thus, a better insulation performance can be achieved in the electrical apparatus 1.

Reasons for providing the gas blenders 339 in the fluid filling device 30 and/or in the electrical ap^¬ paratus 1 are that the fluid components A and B (e.g. C5 and C02^_02-mixture, see above) shall thoroughly be mixed before filling the first amount Ml into the compartment 2 and also after filling. Otherwise, the fluid components A and B could suffer from incomplete mixing which could re^¬ duce the reliability of dielectric performance of the electrical apparatus 1. For example:

(1) The more easily condensable fluid component A (e.g. C5) and the less easily condensable fluid compo^¬ nent B (e.g CO2 and O2) shall be mixed with care to avoid or reduce diffusion times, that otherwise would be needed to mix the first amount Ml diffusively without additional mixing means being present. Similarly, a diffusive mixing of the first amount Ml with the second amount M2 would last longer without additional mixing means being pre^¬ sent. Thus, the additional mixing means, in particular the gas blender 339 or gas blenders 339, allow to reduce or eliminate a waiting time after filling the first amount Ml until it is safe to start operation of the electrical apparatus 1.

(2) The homogeneity of the insulation fluid 10 could change even after the insulation fluid 10 has been filled, e.g. due to a decrease in partial pressure or density of one dielectric insulation fluid component. Such cases may occur for various reasons, for example:

- In case of a condensation event (e.g. of C5) , subsequent reevaporation (e.g. of C5) can safely be achieved by additional mixing means, in particular by gas blender 339 or gas blenders 339, inside the electrical apparatus 1, and a waiting time until it is safe to switch on or to restart the electrical apparatus 1 can again be reduced or eliminated.

- After a switching or circuit breaking event of the electrical apparatus 1, local concentration gradients of e.g. C5 could develop: In an arc, C5 might be consumed so that the C5 number density of the insulation fluid 10 in the arcing zone and in the vicinity of the arcing zone could be lowered. A waiting time until the electrical ap^¬ paratus 1 is operable again, e.g. for opening or closing or reclosing or circuit breaking, can again be reduced or eliminated by providing additional mixing means, in par^¬ ticular gas blender 339 or gas blenders 339, inside the electrical apparatus 1.

Providing the gas blender 339 in the electrical apparatus 1 also allows to fill the first and second com^¬ ponents A and B of the first amount Ml of the insulation fluid 10 in parallel or sequentially into the compartment 2 of the electrical apparatus 1 (i.e. even without out^¬ side premixing, even if not shown in the figures) and then mix them inside the compartment 2. Thus, the fluid filling device 30 could be simplified. Specifically, it would be possible to inject the first insulation fluid component A (e.g. C5) into the evacuated compartment 2 in liquid form before addition of the carrier gas (i.e. the second insulation fluid component B) . As well, liquid first insulation fluid component A (e.g. C5) can be in^¬ jected after filling the carrier gas. This requires ap^¬ plication of the necessary filling pressure, e.g. by a piston or syringe type of injector, to overcome the coun^¬ ter-pressure of the carrier gas. In embodiments, liquid first insulation fluid component A (e.g. C5) is preferen^¬ tially injected so that the liquid evaporates and does not collect in liquid form in the compartment 2 of the electrical apparatus 1. This can for example be accom^¬ plished by spraying the liquid into the compartment 2 through nozzles. It may also be nebulized by an ultrason- ically driven piezo transducer located at the injection point to aid vaporization. To ensure that the right amount is injected, the injected mass or the mass flow can be tracked.

Fig. 31 shows a tube mixer 370 with a plurality of baffles 371. Such a tube mixer 370 can e.g. be com^¬ prised in a gas blender 339 as used in the fluid filling device 30 and/or in the electrical apparatus 1 of Fig. 30. When combining a stream of a gaseous fluid component A (e.g. C5) with a stream of a gaseous fluid com^¬ ponent B (e.g. a carrier gas), adequate turbulence should be provided to thoroughly mix the fluid components A and B. Otherwise, the fluid components A and B could at least in part remain segregated, e.g. owing to their rather different molecular masses. For mixing two streams of different fluid components A and B, a tube mixer 370 can be used: The tube mixer 370 consists of a large bore tube with a series of internal baffles 371 (fixed or static alternating right and left hand helical mixing elements) which create turbulence in the streams by transferring momentum to the particles of the fluid components A and B. Thus, a more rapid and more thorough mixing to a more homogenous stream and a more thoroughly mixed first amount Ml of insulation fluid 10 is achieved.

Fig. 32 shows a tube mixer 380 with a perforated dip tube 381. Such a tube mixer 380 can e.g. be used in a gas blender 339 as used in the fluid filling device 30 and/or in the electrical apparatus 1 of Fig. 30. Here, the first gaseous fluid component A (e.g. C5) enters a large bore tube which encloses a perforated dip tube 381 through which the second gaseous fluid component B (e.g. a carrier gas) enters. The fluid components A and B can then circulate turbulently as momentum is transferred to their particles before exiting the tube mixer 380 as the first amount Ml of the insulation fluid 10. In embodi^¬ ments, the gas blender 339 can additionally comprise a heater 310 to avoid condensation that may occur in regions of high dynamic pressure (dotted) . Thus, a more rapid and more thorough mixing to a homogenous stream and a more thoroughly mixed insulation fluid 10 achieved.

Alternatively, when sequentially filling one gas after the other into a compartment 2, a dip tube 381 can be used (e.g. at the entrance of the compartment 2) which provides more circulation inside the compartment 2, par^¬ ticularly when the light gas is injected into the heavy gas (not shown) .

Figs. 33a and 33b show a rotating mixer 305 in the compartment 2 of an electrical apparatus, the rotat^¬ ing mixer 305 comprising a foldable fan 305a. While Fig. 33a shows the foldable fan 305a in a parking posi^¬ tion (e.g. like a retracted umbrella) in which it does not or only negligibly interfere with the operation of the electrical apparatus 1, in Fig. 33b the foldable fan 305a is in an active position (e.g. like a stretched um^¬ brella) . A rotating mixer 305 can e.g. be used in a gas blender 339 (see Fig. 30) as it provides forced convec^¬ tion by mechanically transferring momentum to the parti^¬ cles of the insulation fluid components A and B. Thus, a more rapid and more thorough mixing to a homogenous stream and a more thoroughly mixed insulation fluid 10 is achieved .

When the or a gas blender 339 is arranged inside the compartment 2 of the electrical apparatus 1, the fan 305a must be arranged in a region that offers the best compromise between high insulation fluid agitation and minimum interference with the operation of the electrical apparatus 1, i.e. it should not negatively affect the di^¬ electric insulation performance (i.e. the fan 305a should be placed in an electric field shadow, and fan materials should be compatible with insulation fluid mixture, etc.) . A foldable fan 305a need not be installed within each compartment permanently, but can be removable. This may require a "foldable" fan 305a that can be introduced through an opening and be retracted after use without causing gas loss from the compartment 2.

In embodiments, a small recess or bulge (as in^¬ dicated e.g. in Fig. 33a, 33b) for housing the folded fan 305a (e.g. permanently or temporarily) is provided at the compartment encapsulation 2.

In embodiments, the fan housing can also be mounted to the compartment 2 by using a self-sealing con^¬ nector (not shown) through which the foldable fan 305a protrudes into and retracts out of the compartment 2 of the electrical apparatus 1. When removing the fan housing from the compartment 2 after having mixed the insulation fluid 10, only a small amount of insulation fluid 10 (i.e. that which is contained in the housing volume) is lost .

Fig. 34 shows a radiation mixer 390 arranged in a compartment 2, wherein the radiation mixer 390 compris^¬ es a light source 391. Here, momentum is transferred to the particles of the insulation fluid 10 radiatively, e.g. by using an IR or UV light source 391 arranged out^¬ side of the compartment 2 and shining into the compart^¬ ment 2 of the electrical apparatus 1 through a window (not shown) of the compartment 2.

Fig. 35 shows a convection mixer 307 arranged in or at the compartment 2 of the electrical apparatus 1 and comprising a heater 310. Forced convection of the insulation fluid 10 is provoked through generation of thermal gradients, i.e. by heating the compartment 2 from one side, for example by using a resistive heating pad or a chemical heater pad for one-time use.

Figs. 36a and 36b show a volumetric mixer 400 arranged in a compartment 2 of the electrical apparatus 1, the volumetric mixer 400 comprising a pneumatically expandable volume 401. By expanding the expandable volume 401 (Fig. 36b), mechanical energy can be transferred to the insulation fluid particles pneumatically, i.e. by having an expandable volume 401 that can be inflated and deflated rapidly from outside the compartment (e.g. by bellows) .

Fig. 37 shows a bypass mixer 410 connected to a compartment 2 of the electrical apparatus 1. The bypass mixer 410 comprises e.g. two valves 337 and a pump 303. Bypass circulation mixing of the insulation fluid 10 can be used to force the insulation fluid components A and B and/or the first amount Ml and the second amount M2 to mix more thoroughly. The bypass mixer 410 comprises a pump 303 which is connected to the compartment 2 at two ports and circulates the insulation fluid 10 there^¬ between. In embodiments, the bypass mixer 410 can be re^¬ movable (e.g. by using self-sealing connections) such that a single bypass mixer device 410 can be used to mix many compartments 2 of the electrical apparatus 1 one af^¬ ter the other.

Thus, an aspect of the invention is an insula^¬ tion fluid filling device 30 adapted to implement the method for filling at least a first amount Ml of an insu^¬ lation fluid 10 into a compartment 2 of a fluid-insulated electrical apparatus 1, in particular of a gas-insulated medium-voltage or high-voltage switchgear 1, according to the filling method and embodiments thereof, as disclosed herein. For that purpose the fluid filling device 30 com^¬ prises : - a mixer 31 for mixing at least two fluid components A, B at a first mixing ratio Rl,

- a first sensor 100 for deriving the first mixing ratio Rl of the first amount Ml of the insulation fluid 10,

- a fluid connector 33 for connecting the insulation fluid filling device 30 to the electrical appa^¬ ratus 1 and for transferring the first amount Ml of the insulation fluid 10 from the insulation fluid filling device 30 to the electrical apparatus 1, and

- an analysis and control unit 34 adapted and structured to carry out the method element or method ele^¬ ments of the method or methods disclosed herein.

In embodiments, the insulation fluid filling de^¬ vice 30 can further comprise an interface 32 for receiving a sensor signal indicative of a second mixing ratio R2 of a second amount M2 of the insulation fluid 10 in the compartment 2 of the electrical apparatus 1.

In embodiments, the analysis and control unit 34 comprises a computer program element comprising computer program code means for, when executed by a processing unit, implementing a filling method or filling methods as disclosed herein.

In embodiments, the insulation fluid filling de^¬ vice 30 comprises in addition:

- vaporization means for bringing the fluid components A, B into gaseous state; and/or

homogenization means for homogenizing the first amount Ml of the insulation fluid 10 inside the in^¬ sulation fluid filling device 30; and/or

- evacuation means for evacuating a or the second amount M2 of the insulation fluid 10 in the compart^¬ ment (2 of the electrical apparatus 1; and/or

- the first sensor 100, 200 being selected from the group of: gas chromatography optical sensor, acoustic sensor, photoacoustic sensor, pressure sensor and temperature sensor in combination with one of speed of sound sensor and viscosity sensor and thermal conductivity sen^¬ sor, and combinations thereof; and/or

- mass flow measurement means for determining a mass flow of at least one of the fluid components A, B; and/or

first measurement and calculation means for determining a dielectric breakdown strength Ebd of the insulation fluid 10 in the compartment 2 of the electri^¬ cal apparatus 1; and/or

- second measurement and calculation means for determining an operating state 0 of the electrical appa^¬ ratus 1; and/or

- the mixer 31 being arranged for mixing the at least two fluid components A, B inside the insulation fluid filling device 30 and/or inside the compartment 2 of the electrical apparatus 1, in particular the mixer 31 being selected from the group consisting of: a tube mixer 370, 380, a rotating mixer 305, a convection mixer 307, a radiation mixer 390, a volumetric mixer 400, a bypass mixer 410, embodiments from claim 23 thereof, and combi^¬ nations thereof.

A further aspect of the invention is an electrical apparatus 1 comprising a compartment 2 housing an electrically active part immersed in an insulation fluid 10, the compartment 2 being adapted to be filled with an insulation fluid filling device 30 according to any one of the disclosure and embodiments herein.

A further aspect of the invention is an electrical apparatus 1 comprising a compartment 2 housing an electrically active part immersed in an insulation fluid 10, the compartment 2 being adapted to be filled by a method for filling or any embodiments thereof, as dis^¬ closed herein.

In embodiments, the electrical apparatus 1 com^¬ prises : - an adapter for receiving a or the fluid connector 33 for connecting the insulation fluid filling device 30 to the electrical apparatus 1 and for transfer^¬ ring the first amount Ml of the insulation fluid 10 from the insulation fluid filling device 30 to the electrical apparatus 1; and/or

- means for homogenizing, for example a foldable fan 305, the first amount Ml of the insulation fluid 10 inside the compartment 2 of the electrical apparatus 1 ; and/or

- a mixer for mixing the at least two fluid com^¬ ponents A, B inside the compartment 2 of the electrical apparatus 1, in particular the mixer being selected from the group consisting of: a tube mixer 370, 380, a rotat^¬ ing mixer 305, a convection mixer 307, a radiation mixer 390, a volumetric mixer 400, a bypass mixer 410, embodi^¬ ments thereof as disclosed herein, and combinations thereof .

Figs. 38a - 38e show various embodiments of optical and mounting configurations for optical sensors, e.g. a first optical sensor 100 and/or second optical sensor 200. In particular, various optical and mounting layouts are shown which can be operated by injecting and retrieving light through optical fibers or by directly connecting a light source and detector to the (first and/or second) optical sensors 100, 200.

Fig. 38 (a) and Fig. 38 (b) are based on inser^¬ tion probes, which can be of the transflexion type or the transmission type. The cavity forming the gas-open measurement cell 21 is arranged inside the measurement space (e.g. the compartment 2 of the electrical apparatus 1, or the insulation fluid filling device 30) . In embodiments, it can be protected from particle contamination by an insulation-fluid-permeable protective cover 500, e.g. made from a sintered material, a gaze, a mesh, a porous mate^¬ rial, a porous metal, or thin membranes of PTFE. The pro- tective cover 500 shall be impermeable to particles that would contaminate optical components or the optical meas^¬ urement path. In addition, protective means such as gas adsorbers (e.g. a zeolite) can be provided, if contami^¬ nants, e.g. decomposition gases, are created which may damage the optical sensor (not shown) .

The design in Fig. 38 (a) is advantageously sim^¬ ple, but requires installation in a position located in the electric field shadow so that no arcing in the vicin^¬ ity of the first and/or second optical sensor 100, 200 occurs. Also, the location must be chosen so that the protruding optical probe does not interfere with the me^¬ chanical motion of other parts in the electrical appa^¬ ratus 1.

Fig. 38 (b) shows an alternative embodiment in which the insertion probe is positioned in a bulge of the compartment 2 that is separated from its main volume by a protective cover 500.

Fig. 38 (c) and Fig. 38 (d) show designs which rely on optical collimators that act as optical feed- throughs into a bulge of the main volume of the compart^¬ ment 2 in a transmission configuration (Fig. 38 (c) ) and a transflexion configuration (Fig. 38 (d) ) . In embodiments, a self-sealing connector can be used (e.g. a DILO connector, not shown) to plug-in and pull-out the meas^¬ urement cell 21 to and from the compartment 2 of the electrical apparatus 1. This allows e.g. cleaning optical surfaces, exchanging particle protection tubes or windows 500, calibration, removal or replacement of the first and/or second optical sensors 100, 200 without taking out of service and depressurizing the electrical apparatus 1.

Fig. 38 (e) shows another alternative configura^¬ tion, in which light source and detector are arranged in^¬ side the compartment 2 of the electrical apparatus 1, and thus within the insulation fluid filled volume. For this concept, electrical feedthroughs which are mechanically much less demanding and of lower cost are used and opti- cal feedthroughs are eliminated. This design can also be applied to configurations using a self-sealing type con^¬ nector, as described above.

As a further alternative, the optical sensor 100, 200 can be arranged (also e.g. via optical fiber) at a viewing port of the electrical apparatus 1. Thus, the viewing port forms not only part of the compartment 11 of the electrical apparatus 1, but also part of the optical sensor 20. The viewing port window shall be transmissive at the measurement wavelength, and possibly also at the reference wavelength, e.g. by using glass with transmis^¬ sion in the near infrared or UV-transmitting material, such as quartz or sapphire, or mid infrared transmitting material, such as potassium bromide (KBr) . In embodi^¬ ments, a mirror or a reflective coating or another viewing port for receiving transmitted light can be present in the compartment 11.

Fig 39 shows an embodiment of an optical sensor 100, 200 with an optical measurement channel or beam at a first wavelength (e.g. that is modified, in particular absorbed, by the first fluid component (A) ) , and an opti^¬ cal reference channel or beam at a second wavelength that is not modified, in particular not absorbed, by the first fluid component (A), such as a fluoroketone . This helps to secure the long-term stability of an optical sensor 100, 200. Several factors could adversely affect the measurements :

• light source emission drift (wavelength and intensity e.g. due to aging, internal temperature and surrounding temperature)

• optical fiber transmission changes (e.g. due to bending losses, physical movement of fibers, stress, temperature)

• transmission changes at optical connectors (e.g. due to physical movement, mechanical shock, vibra^¬ tion, stress, temperature) • transmission changes at optical interfaces (e.g. due to contamination)

• optical detector stability (e.g. due to aging)

• analyzer electronics (e.g. affected by elec^¬ tromagnetic interference or e.g. due to ageing) .

The first five factors can be mitigated by using an optical reference channel integrated into an optical sensor 100, 200.

A blue light source BS emitting at around 293 nm (open rectangles), i.e. within the absorption band of e.g. C5 as first insulation fluid component A, is coupled to an optical fiber (not shown) which delivers the light to an optical probe (sensor head) . Alternatively, the blue light source BS is directly mounted to an optical feedthrough directing the light through the measurement cell 21 in fluid communication with the compartment 2 of the electrical apparatus 1 (see above) .

Prior to being coupled into the optical fiber or the measurement cell 21, part of the light from the blue light source BS is split off using an optical beam split^¬ ter BSP1 and is send to a blue reference detector BRD which measures the stability of the emitted light inten^¬ sity of the blue light source BS .

To eliminate artifacts in the transmitted signal (such as changes in the fiber bending losses or the pres^¬ ence of particle contamination on the optical interfaces within the compartment) , a reference channel is used.

Light at a slightly red-shifted wavelength (black beams) which is not absorbed by e.g. C5 as first insulation fluid component A, i.e. at wavelengths λ > 360 nm, emitted by a red light source RS is used to in^¬ terrogate the optical path for optical transmission changes. The emission stability of RS is recorded by a red reference detector RRD using a second beam splitter BSP2.

The red and blue light is combined (e.g. by a first dichroic mirror DM1) . The red light traverses the same optical path (hatched beams) as the blue light, but is not absorbed by the insulation fluid component A (e.g. C5) . When the red light returns from the sensor head, it is split off using a second dichroic mirror DM2 to the red light detector RD . To ensure that none of the red light falls onto the blue light detector BD and vice ver^¬ sa, a short pass filter FSP is arranged in front of the blue light detector BD and a long pass filter FLP is arranged in front of the red light detector RD.

The number density of the insulation fluid com^¬ ponent A (e.g. C5) can be obtained from the transmitted intensities, and at the same time losses introduced in the optical paths and variations in the intensity of the light sources can be corrected for. In particular the following formula can be used:

with

I-_|-_r (b) = transmitted blue light intensity (falling onto blue light detector BD)

tgp = transmissivity of short pass filter FSP

= transmissivity of dichroic mirror DM1, DM2 for blue light

t-BS = transmissivity of beam splitter BSP1 for refer^¬ ence blue light (to blue reference detector BRD)

I_ref (k) = reference blue light intensity (falling onto blue reference detector BRD)

k = conversion factor for wavelength dependence of optical losses, defined by k = Δΐ_]_₀₃₃ (b) / Δΐ_]_₀₃₃ (^r) with

ΔΙ (b)

loss blue light intensity losses on forward opti^¬ cal path to the gas (i.e. reduction of blue light inten^¬ sity after BSP1 and DM1 by losses up to gas) and on the backward optical path from the gas to the detector BD (i.e. reduction of sensor return blue light intensity by losses) and Δΐ_]_₀₃₃ (^r) = red light intensity losses on forward optical path to the gas (i.e. reduction of red light intensity after BSP1 and DM1 by losses up to gas) and on backward optical path from the gas to the detector RD (i.e. reduc^¬ tion of sensor return red light intensity by losses) I-_|-_r(^r) = transmitted red light intensity (falling onto red light detector RD)

= transmissivity of long pass filter FLP

rDM^^r^ ⁼ reflectivity of dichroic mirror DM1, DM2 for red light

t-BS ^ ⁼ transmissivity of beam splitter BSP2 for refer^¬ ence red light (to red reference detector RRD)

I_ref (^r) = reference red light intensity (falling onto red reference detector RRD)

0 = absorption cross section of dielectric insulation fluid component A (e.g. C5)

1 = absorption path length (in gas; in absorbing gas at at least one wavelength)

N = number density of dielectric insulation fluid compo^¬ nent A (e.g. C5 ) .

A periodic measurement, e.g. a pulsed measure^¬ ment, is preferable to minimize temperature-induced drift effects on the light sources. In this context, it is practical to alternate between the red and the blue chan^¬ nel. Then, by time-gated detection (e.g. via a lock-in amplifier) , the red light detector RD, one dichroic mirror and the filters FLP and FSP can be omitted using just one common detector for both beams, given that detector sensitivity at the different wavelengths is sufficient or similar and the ratio of those sensitivities is known.

Electronics, i.e. light source and detector, can be arranged at the optical components. In this case, proper shielding from electromagnetic radiation is necessary. If electronics cannot be shielded from electromag^¬ netic radiation from the electrical apparatus 1, a fiber optic link can be used. In that case the reference chan^¬ nel setup is particularly useful, if the fibers cannot be held rigidly in place. Alternatively or in addition, they can be immobilized in a duct. In any case, whether the system requires fiber optic links depends on whether electromagnetic interference is critical or not or can be shielded or not.

Definitions :

The term "air" herein shall include "technical air", i.e. pressurized and dried ambient air, or "syn^¬ thetic air", i.e. mixtures of nitrogen (¾) and oxygen (O2) with various mixing ratios, or ambient air.

The term "aliphatic" herein shall relate to both "linear aliphatic" and "branched aliphatic".

The term "fluid" herein shall relate to "a sub^¬ stance, such as a liquid [and/] or gas, that can flow, has no fixed shape, and offers little resistance to an external stress" (from http://www.thefreedictionary.com/ fluid, accessed on 9/11/2011).

The term "high-voltage" herein shall relate to voltages larger than 50 kV.

The term "medium-voltage" herein shall relate to voltages larger than 1 kV.

The compound class "hydrofluoroethers" herein relates to specific partially or fully fluorinated ethers as, e.g., available from 3M.

The compound "C5" herein particularly relates to a fluoroketone selected from the group consisting of the compounds defined by the following structural formulae in which at least one hydrogen atom, preferably all hydrogen atoms, is/are substituted with a fluorine atom/ fluorine atoms :

(lb),

( Ic) , and

The compound "C6" herein particularly relates to a fluoroketone selected from the group consisting of the compounds defined by the following structural formulae in which at least one hydrogen atom, preferably all hydrogen atoms, is/are substituted with a fluorine atom/ fluorine atoms

O (Ilf), and

d i g ) ; The compound "C7" herein particularly relates to a fluoroketone selected from the group consisting of the compounds defined by the following structural formulae in which at least one hydrogen atom, preferably all hydrogen atoms, is/are substituted with a fluorine atom/ fluorine atoms :

( 11 Im) , and

Note :

C5 or "C5", C6 or "C6", C7 or "C7" etc. denote partially or fully fluorinated fluoroketones that have 5, 6, 7 etc. interconnected carbon atoms. Insulation fluid in particular means dielectric insulation fluid.

While there are shown and described presently preferred embodiments of the invention, it is to be dis^¬ tinctly understood that the invention is not limited thereto but may be otherwise variously embodied and prac^¬ ticed within the scope of the following claims. In par^¬ ticular, disclosed device features also disclose the cor^¬ responding method features, and disclosed method features also disclose the corresponding device features. Reference numbers

1 : electrical apparatus

2 : compartment of electrical apparatus 1

10: insulation fluid, dielectric insulation fluid

A: first fluid component of insulation fluid 10

B: second fluid component of insulation fluid 10

Ml: first amount of insulation fluid 10

M2 : second amount of insulation fluid 10

Rl : first mixing ratio of the first amount Ml of the insulation fluid 10

R2 : second mixing ratio of the second amount M2 of the insulation fluid 10

Tf_j__]__]_: filling temperature

c^: first concentration of first fluid component A c-_Q : second concentration of second fluid

component B

Eb^: dielectric breakdown strength of insulation

fluid

21: measurement cell

21a: reference cell

22: electrodes

23: light source

24: lens

25: detector

26: emission filter

27 : mirror

28, BSP: beam splitter

29: band-pass filters

30: insulation fluid filler, insulation fluid filling device

31 : mixer

32 : interface

33: fluid connector

34: analysis and control unit

35, 36: mass flow meter and regulator 100, 200: first and second sensor

301, 302: fluid component reservoirs

303: pump

304: pressure regulator

305: circulator; rotating mixer

305a: foldable fan

307: convection mixer

310: heater

331: carrier gas reservoir

332: sample injector

333: column

334: detector

335: mass spectrometer

337 : valve

339: gas blender

370, 380: tube mixer

371: baffle

381: perforated dip tube

390: radiation mixer

391: light source

400: volumetric mixer

401: expandable volume

410: bypass mixer

500: protective cover

BS : blue light source

RS : red light source

BRD: blue reference detector

RRD: red reference detector

DM2, DM: dichroic mirror

RD: red light detector

BD: blue light detector

FLP: long pass filter

FSP: short pass filter

Claims

1. A method for filling a first amount (Ml) of an insulation fluid (10) into a compartment (2) of a fluid-insulated electrical apparatus (1), in particular of a gas-insulated medium-voltage or high-voltage switch- gear ( 1 ) ,

wherein the method is carried out at a fill^¬ ing temperature (Tf n) ,

wherein the insulation fluid (10) comprises at least a first fluid component (A) and a second fluid component (B) ,

wherein the method comprises method elements of

- mixing the at least two fluid components (A, B) with a first mixing ratio (Rl) for yielding the first amount (Ml) of the insulation fluid (10),

- deriving the first mixing ratio (Rl) of the first amount (Ml) of the insulation fluid (10) by means of a first sensor (100), and

- filling the first amount (Ml) of the insu^¬ lation fluid (10) into the compartment (2) of the elec^¬ trical apparatus (10),

wherein the first mixing ratio (Rl) of the first amount (Ml) of the insulation fluid (10) is con^¬ trolled such that condensation temperatures of the fluid components (A, B) in the first amount (Ml) of the insula^¬ tion fluid (10) are below the filling temperature

(^Tfill)' ^and

wherein the first mixing ratio (Rl) and the first amount (Ml) of the insulation fluid (10) are con^¬ trolled using a target mixing ratio (R) and a target amount (M) of the insulation fluid (10) in the compart^¬ ment of the electrical apparatus (1) .

2. A method for filling a first amount (Ml) of an insulation fluid (10) into a compartment (2) of a fluid-insulated electrical apparatus (1), in particular of a gas-insulated medium-voltage or high-voltage switch- gear ( 1 ) ,

wherein the method is carried out at a fill^¬ ing temperature (Tf n) ,

wherein the method comprises method elements of

- deriving

a) the first mixing ratio (Rl) of the first amount (Ml) of the insulation fluid (10) by means of a first sensor (100) and

b) a second mixing ratio (R2) of a second amount (M2) of the insulation fluid (10) in the compart^¬ ment (2) of the electrical apparatus (1) by means of a second sensor (200), and wherein the method comprises a method element of

(^Tfill)' ^and

wherein the first mixing ratio (Rl) and the first amount (Ml) of the insulation fluid (10) are con^¬ trolled using * the second mixing ratio (R2) and the second amount (M2) of the insulation fluid (10) and/or us^¬ ing

* a target mixing ratio (R) and a target amount (M) of the insulation fluid (10) in the com^¬ partment of the electrical apparatus (1) .

3. The method of any one of the preceding claims, wherein a filling pressure (Pfi]_]_) of the insula^¬ tion fluid (10) in the compartment (2) after the method element of filling the first amount (Ml) of the insula^¬ tion fluid (10) into the compartment (2) is above 1 bar, preferably above 2 bars, more preferably above 5 bars, at an insulation fluid temperature of 20°C.

4. The method of any one of the preceding claims further comprising a method element of

- bringing the fluid components (A, B) into gaseous states.

5. The method of claim 4, wherein the method element of bringing the fluid components (A, B) into gas^¬ eous states is carried out prior to carrying out the method element of mixing the fluid components (A, B) .

6. The method of any one of the claims 4 to 5, wherein the first fluid component (A) is in a liquid state at the filling temperature (Tfiii) and at a pres^¬ sure of 1 bar, and wherein the second fluid component (B) is in a gaseous state at the filling temperature (Tfiii) and at a pressure between 5 bar and 200 bar, and wherein the first and the second fluid components (A, B) are brought into gaseous states and then mixed at the filling temperature (Tfiii) and at a pressure between 3 bar and 10 bar.

7. The method of any one of the preceding claims, further comprising a method element of

- homogenizing the first amount (Ml) of the insulation fluid (10) for reducing a mixture fluctuation and/or a density fluctuation in the first amount (Ml), in particular wherein the method element of homogenizing the first amount (Ml) is carried out before deriving the first mixing ratio (Rl) .

8. The method of any one of the preceding claims, further comprising a method element of

- homogenizing a or the second amount (M2) of the insulation fluid (10) in the compartment (2) of the electrical apparatus (1) for reducing a mixture fluctua^¬ tion and/or a density fluctuation in the second amount (M2), in particular wherein the method element of homoge^¬ nizing the second amount (M2) is carried out before de^¬ riving a or the second mixing ratio (R2) .

9. The method of any one of the preceding claims, further comprising a method element of

- reducing a or the second amount (M2) of the insulation fluid (10) in the compartment (2) of the elec^¬ trical apparatus (1), and/or a method element of

- reducing an amount of a filling gas in the compartment (2) of the electrical apparatus (1),

in particular prior to carrying out the method element of filling the first amount (Ml) of the insu^¬ lation fluid (10) into the compartment (2) of the elec^¬ trical apparatus (1).

10. The method of any one of the preceding claims, wherein at least the method elements of

- mixing the fluid components (A, B) at the first mixing ratio (Rl), and

- deriving a) the first mixing ratio (Rl) of the first amount (Ml) of the insulation fluid (10) and/or

b) a or the second mixing ratio (R2) of the second amount (M2) of the insulation fluid (10), and

- filling the first amount (Ml) of the insu^¬ lation fluid (10) into the compartment (2) of the elec^¬ trical apparatus (10)

are carried out repeatedly.

11. The method of any one of the preceding claims, wherein the method element of filling the first amounts (Ml) of the insulation fluid (10) into the com^¬ partment (2) are carried out against increasing second amounts (M2) of the insulation fluid (10) in the compart^¬ ment (2 ) .

12. The method of any one of the preceding claims, wherein the first sensor (100) and/or a or the second sensor (200) comprises at least one of the group of:

- a gas chromatography

- an optical sensor,

- an acoustic and/or a photoacoustic sensor, a pressure sensor, a temperature sensor, and a density sensor,

a pressure sensor, a temperature sensor, and a speed of sound sensor,

a pressure sensor, a temperature sensor, and a viscosity sensor, and

a pressure sensor, a temperature sensor, and a thermal conductivity sensor.

13. The method of any one of the preceding claims, further comprising a method element of

- deriving a mass flow of at least one of the fluid components (A, B) , and in particular the method comprising a method element of

- deriving a mass flow of the first fluid component (A) in the first amount (Ml) and deriving a mass flow of the second fluid component (B) in the first amount (Ml) prior to or during the method element of filling the first amount (Ml) of the insulation fluid

(10) into the compartment (2) of the electrical apparatus

(10) .

14. The method of any one of the preceding claims, wherein the first fluid component (A) is selected from the group consisting of:

- sulfur hexafluoride,

- partially or fully fluorinated ethers; in particular: hydrofluoroethers , hydrofluoro monoethers, hydrofluoro monoethers containing at least 3 carbon at^¬ oms, perfluoro monoethers, perfluoro monoethers contain^¬ ing at least 4 carbon atoms, fluorooxiranes , perfluo- rooxiranes, hydrofluorooxiranes , perfluorooxiranes com^¬ prising from three to fifteen carbon atoms, hydrofluo- rooxiranes comprising from three to fifteen carbon atoms, and mixtures thereof;

- partially or fully fluorinated ketones; in particular: hydrofluoro monoketones, perfluoro mono^¬ ketones, perfluoro monoketones comprising at least 5 car^¬ bon atoms, or perfluoro monoketones comprising exactly 5 or 6 or 7 or 8 carbon atoms, and mixtures thereof;

- mixtures thereof, and wherein the second fluid component (B) is se^¬ lected from the group consisting of:

- nitrogen,

- oxygen,

- carbon dioxide,

- nitric oxide,

- nitrogen dioxide,

- nitrous oxide,

- argon,

- air, in particular technical air or synthetic air, and

- mixtures thereof.

15. The method of claim 14, wherein the first fluid component (A) is selected from the group consisting of:

- sulfur hexafluoride, and

- hydrofluoroethers .

16. The method of any one of the claims 14 to 15, wherein the second fluid component (B) consists of nitrogen and oxygen with relative partial pressures between p (N₂ ) / (p (O2 ) +p (N₂ ) ) =0.7 , p (0₂ ) / (p (0₂ ) + p(N₂))=0.3 and p (N₂ ) / (p (0₂ ) +P (N₂ ) ) =0.95 , p (0₂ ) / (p (0₂ ) + p (N₂) ) =0.05 or

carbon dioxide and nitrogen with relative partial pressures between p (C0₂ ) / (p (N₂ ) +p (C0₂ ) ) =0.1 , p (N₂) / (p (N₂) +p (C0₂) ) =0.9 and p (C0₂ ) / (p (N₂ ) +p (C0₂ ) ) =0.9, p(N₂) / (p(N₂)+p(C0₂) )=0.1, and

wherein the first fluid component (A) com^¬ prises at least one of the group consisting of:

17. The method of any one of the claims 14 to 16, wherein the second fluid component (B) comprises

wherein the first fluid component (A) com^¬ prises 1,1,1,3,4,4, 4-heptafluoro-3- (tri-fluoromethyl) butane-one with a partial pressure between 0.25 bar and 0.5 bar and/or the fluid component 1,1,1,2,4,4,5,5,5- nona-fluoro-4- (tri-fluoromethyl) pentan-3-one with a partial pressure between 0.02 bar and 0.3 bar at a tempera^¬ ture of 20°C.

18. The method of any one of the preceding claims, further comprising method elements of

- deriving a first concentration (cp of the first fluid component (A) in a or the second amount (M2) of the insulation fluid (10) in the compartment (2) of the electrical apparatus (1) and

- deriving a second concentration (CQ) of the second fluid component (B) in the second amount (M2) of the insulation fluid (10) in the compartment (2) of the electrical apparatus (1) .

19. The method of any one of the preceding claims, further comprising a method element of

deriving a dielectric breakdown strength EbcL of a or the second amount (M2) of the insulation flu^¬ id (10) in the compartment (2) of the electrical appa^¬ ratus (1) using a or the first concentration (cp of the first fluid component (A) and using a or the second con^¬ centration (CQ) of the second fluid component (B) , in particular according to

^Ebd = S{c_A,c_B) ∑CiE_{cri i}

i=A,B

with ^cT ,A Si d ^cr , being fluid- component-specific critical field strengths of the fluid components A and B; with c^ and C being the first and second concentrations of the first and second fluid com^¬ ponents A and B; with S (c^, CQ) being a synergy parame^¬ ter; and with i being an index for the fluid components A and B .

20. The method of any one of the preceding claims, comprising a further method element of

deriving an operating state (0) of the electrical apparatus (1) using a first concentration (cp of the first fluid component (A) and using a second con^¬ centration (C-Q) of the second fluid component (B) in a or the second amount (M2) of the insulation fluid (10) in the compartment (2) of the electrical apparatus (1), and/or comprising a further method element of

- deriving a or the operating state (0) of the electrical apparatus (1) using a or the dielectric breakdown strength (Ε^) of a or the second amount (M2) of the insulation fluid (10) in the compartment (2) of the electrical apparatus (1).

21. The method of any one of the preceding claims, wherein the method element of mixing the at least two fluid components (A, B) for yielding the first amount (Ml) of the insulation fluid (10) is carried out prior to the method element of filling the first amount (Ml) of the insulation fluid (10) into the compartment (2) of the electrical apparatus (10); and/or

wherein the first mixing ratio and the first amount is of the still-to-be filled insulation fluid (10) .

22. The method of any one of the preceding claims, wherein the method element of mixing the at least two fluid components (A, B) with the first mixing ratio (Rl) comprises a disturbing of a laminar flow of the at least two fluid components (A, B) for introducing a tur^¬ bulence .

23. The method of claim 22, wherein the dis^¬ turbing of the laminar flow is carried out by a gas blender (339) comprising at least one of the group of:

- a tube mixer (370, 380), in particular comprising a baffle (371), particularly a plurality of baffles (371), and/or comprising a perforated dip tube (381) ,

- a rotating mixer (305) , in particular comprising a fan (305a) , particularly a foldable fan (305a) , a convection mixer (307), in particular comprising a heater (310), - a radiation mixer (390), in particular comprising a UV light source (391) or an IR light source (391) ,

a volumetric mixer (400), in particular comprising an expandable volume (401), particularly a pneumatically expandable volume (401),

a bypass mixer (410), and

combinations thereof.

24. The method of any one of the preceding claims, wherein a concentration of the first component

(A) in the first amount (Ml) of the insulation fluid (10) is below a threshold concentration, above which a condensation of the first component (A) takes place at the filling temperature (Tfj__]__]_) .

25. The method of any one of the preceding claims, wherein a concentration of the second component

(B) in the first amount (Ml) of the insulation fluid (10) is below a threshold concentration above which a condensation of the second component (B) takes place at the filling temperature (Tfj__]__]_) .

26. An insulation fluid filling device (30) adapted to implement the method for filling at least a first amount (Ml) of an insulation fluid (10) into a com^¬ partment (2) of a fluid-insulated electrical apparatus

(1), in particular of a gas-insulated medium-voltage or high-voltage switchgear (1), according to any one of the preceding claims, the insulation fluid filling device

(30) comprising:

- a mixer (31) for mixing at least two fluid components (A, B) at a first mixing ratio (Rl),

- a first sensor (100) for deriving the first mixing ratio (Rl) of the first amount (Ml) of the insula^¬ tion fluid (10) , - a fluid connector (33) for connecting the insulation fluid filling device (30) to the electrical apparatus (1) and for transferring the first amount (Ml) of the insulation fluid (10) from the insulation fluid filling device (30) to the electrical apparatus (1), and

- an analysis and control unit (34) adapted and structured to carry out the method element or method elements of the method of any one of the preceding claims .

27. The insulation fluid filling device (30) of claim 26 further comprising

- an interface (32) for receiving a sensor signal indicative of a second mixing ratio (R2) of a sec^¬ ond amount (M2) of the insulation fluid (10) in the com^¬ partment (2) of the electrical apparatus (1) .

28. The insulation fluid filling device (30) of any one of the claims 26 to 27, wherein the analysis and control unit (34) comprises a computer program ele^¬ ment comprising computer program code means for, when executed by a processing unit, implementing a method of any of the claims 1 to 25.

29. The insulation fluid filling device (30) of any one of the claims 26 to 28, comprising:

- vaporization means for bringing the fluid components (A, B) into gaseous state; and/or

- homogenization means for homogenizing the first amount (Ml) of the insulation fluid (10) inside the insulation fluid filling device (30); and/or

- evacuation means for evacuating a or the second amount (M2) of the insulation fluid (10) in the compartment (2) of the electrical apparatus (1); and/or

- the first sensor (100, 200) being selected from the group of: gas chromatograph, optical sensor, acoustic sensor, photoacoustic sensor, pressure sensor and temperature sensor in combination with one of speed of sound sensor and viscosity sensor and thermal conduc^¬ tivity sensor, and combinations thereof; and/or

- mass flow measurement means for determining a mass flow of at least one of the fluid components (A, B) ; and/or

- first measurement and calculation means for determining a dielectric breakdown strength of the insulation fluid (10) in the compartment (2) of the elec^¬ trical apparatus (1); and/or

second measurement and calculation means for determining an operating state (0) of the electrical apparatus (1); and/or

- the mixer (31) being arranged for mixing the at least two fluid components (A, B) inside the insu^¬ lation fluid filling device (30) and/or inside the com^¬ partment (2) of the electrical apparatus (1), in particu^¬ lar the mixer (31) being selected from the group consist^¬ ing of: a tube mixer (370, 380), a rotating mixer (305), a convection mixer (307), a radiation mixer (390), a volumetric mixer (400), a bypass mixer (410), embodiments from claim 23 thereof, and combinations thereof.

30. An electrical apparatus (1) comprising a compartment (2) housing an electrically active part im^¬ mersed in an insulation fluid (10), the compartment (2) being adapted to be filled with an insulation fluid filling device (30) according to any one of the claims 26 to 29.

31. An electrical apparatus (1) comprising a compartment (2) housing an electrically active part im^¬ mersed in an insulation fluid (10), the compartment (2) being adapted to be filled by a method for filling ac^¬ cording to any one of the claims 1 to 25.

32. The electrical apparatus (1) of any one of the claims 30 to 31, comprising:

an adapter for receiving a or the fluid connector (33) for connecting the insulation fluid filling device (30) to the electrical apparatus (1) and for transferring the first amount (Ml) of the insulation fluid (10) from the insulation fluid filling device (30) to the electrical apparatus (1); and/or

means for homogenizing (305a) the first amount (Ml) of the insulation fluid (10) inside the com^¬ partment (2) of the electrical apparatus (1); and/or

- a mixer for mixing the at least two fluid components (A, B) inside the compartment (2) of the elec^¬ trical apparatus (1), in particular the mixer being se^¬ lected from the group consisting of: a tube mixer (370, 380), a rotating mixer (305), a convection mixer (307), a radiation mixer (390), a volumetric mixer (400), a bypass mixer (410) , embodiments from claim 23 thereof, and combinations thereof.