Device and method for filling a pressurized gas tank

Abstract

Method (100) for filling a pressurized gas tank (1), implemented by a filling device (10) comprising a gas source (2), a filling line (3) connecting the source (2) to the tank (1), a flow and / or pressure regulating valve (4), and an electronic control device (5) configured to stop filling when an estimated temperature (or density) of the gas in the tank (1) reaches a temperature limit (or density limit, respectively), the method comprising, before filling the tank (1), a step (S1) for determining the ambient temperature (Tamb) at the filling device (10), a step (S2) for determining the pressure (Pini) of the gas in the tank (1), and a preliminary step (S3) for estimating the initial temperature (Tini) of the gas in the tank (1).the initial temperature (Tini) of the gas present in the tank being an estimated value based on the ambient temperature (Tamb) and the pressure (P) of the gas present in the tank before filling, the initial temperature (Tini) of the gas present in the tank being greater than or equal to, or less than or equal to, the ambient temperature (Tamb) Figure of the abstract: Fig. 14,

Description

Title of the invention: Device and method for filling a pressurized gas tank

[0001] The invention relates to a device and a method for filling a pressurized gas tank.

[0002] The invention applies in particular to the filling of hydrogen tanks.

[0003] The invention relates more particularly to a device for filling a pressurized gas tank, comprising a gas source, a filling pipe connecting the source to the tank, a flow and / or pressure regulating valve in the filling pipe, a sensor configured to measure the pressure in the tank and / or the ambient temperature at the filling device, and an electronic control member configured to: - stop filling when an estimated temperature (respectively a density) of the gas present in the tank reaches a temperature limit value (respectively a density limit value), - estimate before filling the tank, an initial temperature of the gas present in the tank.

[0004] In a device such as described above, the temperature of the gas present in the tank, and in particular the temperature before filling, remains difficult to access. As is known, the initial temperature used in existing filling protocols is derived from recommendations of the SAE (“Society of Automotive Engineering”) relating to the filling of fuel cell vehicles. The initial temperature is given as a function of the ambient temperature according to a predefined curve for the hot case and a predefined curve for the cold case.

[0005] However, practice shows that the initial temperature of the tank to be filled does not always evolve according to these curves recommended by the SAE. Thus, with an initial temperature of the tank located outside the corridor formed by the curves recommended by the SAE, the tank is exposed in particular to a risk of overheating or overfilling.

[0006] The risk of overheating is increased in the case of successive fillings. Indeed, while filling generally causes the gas present in the tank to increase its temperature, current protocols consider that for a subsequent filling the initial temperature of the tank remains within the corridor defined by the SAE. Thus, due to an underestimated initial gas temperature during the subsequent filling, the temperature limit can be quickly exceeded, in contradiction with the predictions of the models deployed in current filling protocols.

[0007] An aim of the present invention is to overcome all or part of the drawbacks of the prior art noted above.

[0008] To this end, the device according to the invention, moreover in accordance with the generic definition given in the preamble above, is essentially characterized in that the initial temperature is a value estimated as a function of the ambient temperature and as a function of the pressure of the gas present in the tank before filling, the initial temperature of the gas in the tank being greater than or equal to or less than or equal to the ambient temperature.

[0009] Thus, the invention improves the precision with which the initial temperature of the gas present in the tank is estimated, thus limiting the risk of overheating or overfilling of the tank.

[0010] The invention also relates to a method for filling a pressurized gas tank, implemented by a filling device comprising a gas source, a filling line connecting the source to the tank, a flow and / or pressure regulating valve in the filling line, and an electronic control member configured to stop filling when an estimated temperature (or density) of the gas present in the tank reaches a temperature limit value (or respectively a density limit value). The method comprises, before filling the tank, a step of determining the ambient temperature at the filling device, a step of determining the pressure of the gas present in the tank and a preliminary step of estimating the initial temperature of the gas present in the tank.The initial temperature of the gas in the tank is an estimated value based on the ambient temperature and on the pressure of the gas present in the tank before filling, said initial temperature of the gas in the tank being greater than or equal to or less than or equal to the ambient temperature.

[0011] Furthermore, embodiments of the method according to the invention may comprise one or more of the following characteristics: - the initial temperature of the gas in the reservoir is at least one of: a first calculated initial temperature corresponding to a state of the reservoir considered to be recently filled to a first initial density, a second calculated initial temperature corresponding to a state of the reservoir considered to be recently withdrawn to a second initial density; - the first initial temperature of the gas in the tank is within a high temperature range having a determined lower limit greater than or equal to the ambient temperature and an upper limit corresponding to a determined maximum temperature limit, for example equal to 85°C; - the first initial temperature is determined from a first pre-curve determined predictive “TPgas_hot” provided by a predetermined physical model simulating a reference filling of the tank over the high temperature range; - the second initial temperature of the gas in the tank is included in a low temperature range having a determined upper limit less than or equal to the ambient temperature and a lower limit corresponding to a determined minimum temperature limit, for example between zero and -5°C; - the second initial temperature is determined from a second predetermined predictive curve “TPgas_cold” provided by a predetermined physical model simulating a reference emptying of the tank over the low temperature range; - the method comprises a step consisting of modeling, during filling, a first temperature variation curve (or respectively a first density variation curve) of the gas present in the tank, having as starting condition the first initial temperature (respectively the first initial density) of the gas present in the tank, and / or a second temperature variation curve (or respectively a second density variation curve) of the gas present in the tank, having as starting condition the second initial temperature (respectively the second initial density) of the gas present in the tank; - the method comprises a step of estimating, during filling, a temperature variation curve (or respectively a density variation curve) of the gas present in the tank, said curve being between the first temperature variation curve (respectively the first density variation curve) and the second temperature variation curve (respectively the second density variation curve); - the first initial temperature of the tank and / or the second initial temperature is recalculated during filling as a function of the flow rate and as a function of the temperature of the gas present in the filling pipe, said flow rate and said temperature being determined by calculation and / or by sensors at the filling device; - the temperature limit value (respectively the density limit value) is a determined fixed value, for example 85°C (respectively, for example 24.1 kg / m3 or 40.2 kg / m3) or a value given by a reference temperature “TPgas_max” curve (respectively by a reference density “RHO_cold” curve), said curve being provided by a predetermined physical model which simulates the thermodynamic conditions of the gas during a reference filling of the tank; - the physical model is based on a system of equations comprising at least one of: an internal energy balance equation applied to the gas present in the reservoir, a mass balance equation applied to the gas present in the reservoir, an energy conservation equation in a wall of the reservoir, a heat flow continuity equation between the gas present in the reservoir and the reservoir wall, a heat flow continuity equation between the tank wall and the ambient air, and a flow rate equation relating a mass flow rate of the filling device to a pressure difference between the filling device and the reservoir; - the first initial temperature of the reservoir (respectively the second initial temperature of the reservoir) and the initial pressure of the gas present in the reservoir are obtained by solving said system of equations, for example using a tridiagonal matrix type algorithm; - the ambient temperature of the filling device and the initial pressure of the gas present in the tank are determined by calculation and / or measured by sensors at the filling device; - the electronic control member is configured to control the flow and / or pressure control valve to achieve a predetermined pressure curve or ramp during filling; - the electronic control unit is configured to simulate and estimate the density variation curve (and / or the temperature variation curve) of the gas present in the tank dynamically during filling of the tank and / or in advance, i.e. before filling.

[0012] The invention may also relate to any alternative device or method comprising any combination of the above or below features within the scope of the claims.

[0013] Other features and advantages will appear on reading the description below, given with reference to the following figures in which:

[0014] [Fig-1] schematically and partially illustrates a filling device according to the invention;

[0015] [Fig.2] schematically illustrates steps of a filling process using of the filling device;

[0016] [Fig.3] illustrates a reference temperature curve giving, for a tem given ambient temperature, an evolution of the temperature limit of the gas as a function of its pressure;

[0017] [Fig.4] illustrates an evolution with respect to the ambient temperature of a terminal in lower (respectively of an upper limit) associated with a range of high temperatures (respectively a range of low temperatures) in which a reference filling;

[0018] [Fig.5] illustrates curves giving, for a simulated reference filling at a given ambient temperature, the variation in the temperature of the tank considered to be recently filled (respectively considered to be recently drawn off), the variation in the temperature, density and pressure of the gas present in said tank, the variation in the temperature and pressure of the gas at the outlet of the filling device;

[0019] [Fig.6] illustrates a surface curve corresponding to several temperature curves reference temperatures established for different ambient temperatures;

[0020] [Fig.7] illustrates the reference temperature curve of [Fig.l] with a variable “temperature” reduced between 0 (zero) and 1 (one);

[0021] [Fig.8] illustrates reference temperature curves and predictive curves of the first initial temperature of the reservoir or of the gas present in said reservoir;

[0022] [Fig.9] illustrates curves representing an evolution of the temperature of the reservoir and the temperature of the gas present in said reservoir, said evolution being linked to a thermal diffusion subsequent to a reference filling;

[0023] [Fig. 10] illustrates the curves of [Fig.8] after reduction of the “temperature” variable;

[0024] [Fig. 11] illustrates predictive curves of the second initial temperature of the reservoir or of the gas present in said reservoir;

[0025] [Fig. 12] illustrates the curves of [Fig.l 1] after reduction of the variable “temperature” and reduction of the variable “pressure” between zero (0) and one (1);

[0026] [Fig. 13] illustrates an example of implementation of a gas temperature variation curve to control the filling of the tank and prevent it from overheating;

[0027] [Fig. 14] illustrates an example of implementation of a gas temperature variation curve in combination with the reference temperature curve to control the filling of the tank and prevent its overheating;

[0028] [Fig. 15] represents an example of radial discretization of a tank wall;

[0029] [Fig. 16] represents details of a mesh obtained after the discretization of the [Fig.15],

[0030] The device 10 for filling pressurized gas tanks 1 is, for example, a station for filling pressurized hydrogen tanks.

[0031] The device 10 comprises a gas source 2, a filling pipe 3 connecting the source 2 to the tank 1, a valve 4 for regulating the flow rate and / or pressure in the filling pipe 3, a set of sensor(s) 6a, 6b configured to measure the pressure in the tank 1 and / or the ambient temperature Tamb at the filling device 10.

[0032] Furthermore, the device 10 comprises an electronic control member 5 configured to control the filling and in particular to stop the filling when an estimated temperature (respectively a density) of the gas present in the tank 1 reaches a temperature limit value (respectively a density limit value). The control member 5 is also configured to estimate, before filling the tank 1, an initial temperature of the gas present in the tank 1.

[0033] The electronic control member 5 comprises for example a microprocessor, a computer or any suitable electronic controller.

[0034] According to the invention, the initial temperature Tini of the gas present in the tank 1 is a value estimated as a function of the ambient temperature Tamb and as a function of the pressure Pini of the gas present in the tank 1 before filling, said initial temperature Tini being greater than or equal to or less than or equal to the ambient temperature Tamb.

[0035] Thus, for a given ambient temperature Tamb, the initial temperature Tini of the gas present in the tank is no longer a fixed quantity as provided for in the prior art, but rather a quantity which depends on the initial pressure Pini of the gas present in the tank 1.

[0036] Advantageously, the initial temperature Tini of the gas in the reservoir 1 is at least one of: a first calculated initial temperature Tini_l corresponding to a state of the reservoir 1 considered to be recently filled to a first density pini_l, and a second calculated initial temperature Tini_2 corresponding to a state of the reservoir 1 considered to be recently withdrawn to a second initial density pini_2.

[0037] The ambient temperature Tamb and the initial pressure Pini are determined respectively during a step S1 and a step S2 of a filling method implementing the filling device and illustrated schematically in [Fig.2]. The steps S1 and S2 may be executed simultaneously or not. As for the initial temperature Tini, which is divided into a first initial temperature Tini_l and a second initial temperature Tini_2, it is determined during a step S3 also illustrated in [Fig.2].

[0038] In the following, the tank 1 considered to be recently filled will also be considered a tank recently exposed to direct sunlight or in a relatively hot environment or a tank in the process of being filled while hot. The tank considered to be recently drawn off will also be considered a tank recently exposed to a relatively cold environment or a tank in the process of being filled while cold.

[0039] The reservoir considered to be recently filled and the reservoir considered to be recently drawn are modeled virtual reservoirs which may be identical or No. In the second case, the tank considered to be recently drawn may in particular have a surface area / volume ratio greater than that of the tank considered to be recently filled, which makes it more capable of evacuating heat and cooling down.

[0040] Advantageously, the first initial temperature Tini_l of the tank is included in a high temperature range having a determined lower limit TLgas_min_l greater than or equal to the ambient temperature and an upper limit TLgas_max_l corresponding to a determined maximum temperature limit. The evolution of the lower limit TLgas_min_l as a function of the ambient temperature of the filling device is given by a “Thotsoak” curve (see [Fig.4]) resulting for example from SAE recommendations relating to the filling of a hot tank. The upper limit TLgasmax_l is set for example at 85°C. This value takes into account the thermophysical properties of the tank.

[0041] Advantageously, the second initial temperature Tini_2 of the tank is included in a low temperature range having a determined upper limit TLgas_max_2 less than or equal to the ambient temperature Tamb and a lower limit TLgas_min_2 corresponding to a determined minimum temperature limit, for example between zero and -5°C. In particular, the upper limit TLgas_max_2 is a function of the ambient temperature and follows a “Tcoldsoak” curve (see [Fig.4]) resulting for example from SAE recommendations relating to filling a cold tank.

[0042] Advantageously, the temperature limit value (respectively the density limit value) is a fixed value determined or a value given by a reference temperature “TPgasmax” curve (respectively a reference density curve). In particular, the reference temperature “TPgasmax” curve (respectively the reference density curve) makes it possible, for a given pressure of the gas in the tank, to estimate the temperature limit value (respectively the density limit value).

[0043] An example of a reference temperature “TPgas_max” curve is shown in [Fig.3]. For a maximum pressure of the gas present in tank 1, the temperature limit value is 85°C and coincides with the upper limit TLgasmax_l of the high temperature range. For a minimum pressure of the gas in the tank, the temperature limit value is approximately 20°C and coincides with the lower limit TLgasmin_l of the high temperature range.

[0044] The density limit value depends on the working pressure of the tank. For example, for a working pressure of 350 barg, the density limit value is set, for example, at 24.1 kg / m3. For a working pressure of 700 barg, the density limit value is set, for example, at 40.2 kg / m3. Generally speaking, the density limit value can be defined by a percentage of filling state “SOC” (for “State of Charge”, in English) which defines a maximum quantity of hydrogen recommended in a tank.

[0045] Determination of the reference temperature “TPgasmax” curve and the reference density curve

[0046] Advantageously, the reference temperature curve “TPgasmax” (respectively the reference density curve) is provided by a first physical model which simulates a reference filling of the tank over the high temperature range (respectively over the low temperature range) with pre-cooling at the source. The reference filling is simulated from a minimum pressure which is here set at 5 barg up to a maximum pressure corresponding to the working pressure of the tank.

[0047] In particular, the reference temperature curve “TPgasmax” (respectively the reference density curve) is obtained from a curve “Pgashot” giving the evolution of the pressure of the gas present in the tank as a function of time and from a curve “Tgasmax” (respectively a curve “RHO_cold”) giving the evolution as a function of time of the temperature (respectively the evolution of the density) of the gas contained in the tank. The curves “Pgashot”, “Tgasmax” (respectively “RHO_cold”) are also provided by the first predetermined physical model. They are illustrated in [Fig.5].

[0048] For different ambient temperatures, the first predetermined physical model provides as many reference temperature “TPgasmax” curves (respectively as many reference density curves) which form a reference temperature surface curve (respectively a reference density surface curve). [Fig.6] illustrates an example of a reference temperature surface curve.

[0049] In order to facilitate the exploitation of the reference temperature “TPgasmax” curve (respectively the reference density curve), the “temperature” variable is reduced between the value zero (0) corresponding to the lower limit TLgasmin_l and the value one (1) corresponding to the upper limit TLgasmax_l. [Fig.7] illustrates a “Reduc_TPgasmax” curve obtained after reduction of the “temperature” variable on the reference temperature “TPgasmax” curve.

[0050] Determination of the first initial temperature Tini_l of the gas

[0051] The first initial temperature Tini_l of the gas present in the tank is given by a second predetermined predictive curve “TPgashot” which is a function of the initial pressure Pini of the gas (see [Fig.8]). In other words, for a given initial pressure Pini of the gas present in the tank, the predictive curve “TPgashot” makes it possible to determine the first initial temperature Tini_l.

[0052] The predictive curve “TPgashot” is derived from a second physical model which simulates a reference filling of the tank over the high temperature range with possibly pre-cooling of the gas supplied by the source. Preferably, the second physical model additionally simulates thermal diffusion at the tank and / or at the gas present in the tank. Thus, the second physical model may differ from the first physical model by taking thermal diffusion into account.

[0053] The thermal diffusion is subsequent to the reference filling of the tank and the pre-cooling of the source. Advantageously, the thermal diffusion is simulated over a period of 2 minutes corresponding to the minimum time required between two successive fillings of the tank at the filling station. Also advantageously, during the thermal diffusion period, the pressure of the gas in the tank is assumed to be constant.

[0054] For a given ambient temperature, thermal diffusion induces a temperature drop in the gas present in the tank. Such a drop is represented by a “STgasmax” curve illustrated in [Fig.9].

[0055] It should be noted that the “TPgashot” curve is obtained from the “Pgashot” curve and the “Tgashot” curve, both established for the same ambient temperature. In particular, the “Tgashot” curve results from a subtraction carried out between the “Tgasmax” curve and the “STgasmax” curve.

[0056] Advantageously, in order to facilitate the exploitation of the “TPgashot” curve, the “temperature” variable can be reduced between the value zero (0) corresponding to the lower limit TLgasmin_l and the value one (1) corresponding to the upper limit TLgasmax_l. The “Reduc_TPgashot” curve thus obtained is illustrated in [Fig. 10].

[0057] Determination of the second initial temperature Tini_2 of the gas

[0058] The second initial temperature Tini_2 of the gas is obtained from a third curve “TPgascold” which is a function of the initial pressure Pini of the gas (see [Fig. 11]). In other words, based on the initial pressure Pini of the gas present in the reservoir, the curve “TPgascold” makes it possible to determine the second initial temperature Tini_2 of the gas present in the reservoir.

[0059] The “TPgascold” curve can be provided by a third predetermined physical model (described later) which simulates a reference emptying of the tank over the low temperature range, i.e. between the upper limit TLgasmax_2 and the lower limit TLgasmin_2.

[0060] The upper limit TLgasmax_2 corresponds to the start of the reference emptying where the tank is in a state at least partially filled, and preferably 100%. This upper limit TLgasmax_2 corresponds to a maximum pressure Pgasmax_2, called the service pressure (“Nominal Working Pressure” in English), and a density maximum pgasmax_2. It should be noted that the maximum pressure Pgasmax_2 can be expressed as a function of the upper bound TLgasmax_2 through the following mathematical expression taken from the literature:

[0061] Pgasmax_2 [barg] = A * TLgasmax_2 [°C] + B

[0062] with A between 1 and 2, and B between 300 and 400.

[0063] The lower bound TLgasmin_2 corresponds to the end of reference emptying where the tank has a minimum pressure Pgasmin_2, here set at 5 barg, and a minimum density pgasmin_2, called the second initial density pini_2. In addition, the lower bound TLgasmin_2 can be obtained from the mathematical expression below taken from the third physical model:

[0064] Tgasmin_2 = A * Tamb + B

[0065] where

[0066] Tamb is the ambient temperature at the filling device;

[0067] A is between 0.5 and 1.5;

[0068] B is between -30 and -15.

[0069] For different ambient temperatures, the third physical model provides as many “TPgascold” curves that form a surface curve (not shown). In order to facilitate the exploitation of these “TPgascold” curves, the “temperature” variable on each of them is reduced between the value zero (0) corresponding to the lower bound TLgasmin_2 and the value one (1) corresponding to the upper bound TLgasmax_2. Similarly, the “pressure” variable can be reduced between the value zero (0) corresponding to the minimum pressure Pgasmin_2 and the value one (1) corresponding to the maximum pressure Pgasmax_2 before emptying the tank. The “Reduc_TPgascold” curve illustrated in [Fig. 12] is the result of such a reduction carried out on the “TPgascold” curve.

[0070] Ultimately, for an initial pressure Pini measured in the tank and an ambient temperature Tamb recorded at the filling device, at least one predetermined physical model considers that this initial pressure Pini is the final pressure of a previous filling (“hot case”) or of a previous withdrawal (“cold case”). These two hypotheses respectively define a value for the first initial temperature Tini_l (corresponding to the hot case) and a value for the second initial temperature Tini_2 (corresponding to the cold case). The value of the first initial temperature Tini_l and the value of the second initial temperature Tini_2 thus defined may be different from the initial temperature value determined according to the prior art, for example deviating by 10 to 20°C from the ambient temperature.

[0071] Estimation of the gas temperature variation curve between a first gas temperature variation curve (hot case) and a second gas temperature variation curve (cold case).

[0072] Advantageously, the control member 5 can be configured to estimate, during filling, a variation curve of the temperature of the gas present in the tank as a function of the initial temperature of the gas present in the tank, that is to say with as starting point (or starting condition), the initial temperature of the gas present in the tank.

[0073] More specifically, the control member 5 can be configured to estimate such a curve between a first gas temperature variation curve (hot case) and a second gas temperature variation curve (cold case). Thus, for a given pressure of the gas present in the tank, the corresponding temperature is estimated between a first temperature located on the first gas temperature variation curve (hot case) and a second temperature located on the second gas temperature variation curve (cold case).

[0074] The first curve (respectively the second curve) of variation of the temperature of the gas has as starting point (or starting condition) the first initial temperature Tini_l (respectively the second initial temperature Tini_2) of the gas. Furthermore, the first curve (respectively the second curve) of variation of the temperature of the gas is provided by a predetermined physical model which simulates a reference filling of the reservoir from the first initial temperature Tini_l (respectively from the second initial temperature Tini_2) of the gas, said filling being accompanied by a pre-cooling at the source and followed by a thermal diffusion.

[0075] It should be noted that the physical model at the origin of the “TPgashot” curve predicting the first initial temperature Tini_l of the gas is preferably the same as that at the origin of the first curve of variation of the temperature of the gas. The two curves are distinguished by their starting points which are respectively the lower bound TLgasmin_l (for the “TPgashot” curve predicting the first initial temperature Tini_l) and the first initial temperature Tini_l (for the first curve of variation of the temperature of the gas).

[0076] Estimation of the density of the gas present in the reservoir between a first density variation curve (hot case) and a second density variation curve (cold case)

[0077] Advantageously, the control member 5 can be configured to estimate (for example calculate), during filling, a curve of variation of the density of the gas present in the tank as a function of the initial density of the gas present in the tank. This means that the curve of variation of the density has as its starting point the initial density of the gas present in the tank. It should be noted that the initial density can be calculated by the ideal or real gas equation knowing the volume of the tank (known or estimated, for example via the known technique based on a pressure “pull” pulse before filling), the pressure and the temperature of the gas.

[0078] More specifically, the control member 5 can be configured to estimate the curve of variation of the density of the gas between a first curve of variation of the density of the gas (hot case) and a second curve of variation of the density of the gas (cold case). Thus, for a given pressure of the gas present in the tank, the corresponding density is estimated between a first density located on the first curve of variation of the density (hot case) and a second density located on the second curve of variation of the density (cold case).

[0079] The first curve (respectively the second curve) of variation of the density has as its starting point the first initial density (respectively the second initial density) of the gas present in the reservoir. Furthermore, the first curve (respectively the second curve) of variation of the density of the gas is provided by a physical model which simulates a reference filling of the reservoir from the first initial density (respectively from the second initial density) of the gas with pre-cooling at the source.

[0080] Determination of the reservoir reference temperature curve and the reservoir temperature variation curves

[0081] Advantageously, the condition for stopping filling of the tank relating to a limit value of the temperature of the gas present in the tank can be replaced by a stopping condition relating to a limit value of the temperature of the tank.

[0082] The limit value of the tank temperature can be given by a reference temperature curve of the tank, said curve being provided by the first physical model. In particular, for a tank formed of several layers, for example an internal layer called “liner” in contact with the gas and an external layer called “composite” in contact with the ambient air, the first physical model provides three reference temperature curves illustrated in [Fig.8], namely a first curve “TPgasliner_max” for the internal wall, a second curve “TPlinercompo_max” for the internal layer / external layer interface, and a third curve “TPexternal_max” for the external wall.

[0083] The curves “TPgasliner_max”, “TPexternal_max”, “TPlinercompo_max” can be obtained from the curve “Pgashot” in combination respectively with curves “Tgasliner_hot”, “Textemal_hot”, “Tlinercompo_hot” giving the evolution of the temperature as a function of time for the internal wall, the internal layer / external layer interface, and the external wall. The curves “Tgasliner_hot”, “Texternal_hot” are illustrated in [Fig.5].

[0084] Furthermore, to be easily exploitable, the curves “TPgasliner_max”, “TPexternal_max”, “TPlinercompo_max” can be transformed respectively into a curve “Reduc_TPgasliner_max”, a curve “Reduc_TPextemal_max”, and a curve “Reduc_TPlinercompo_max” each defined between a value zero (1) corresponding to the lower limit TLgasmin_l of temperature and a value one (1) corresponding to the upper limit TLgasmax_l of temperature.

[0085] Estimation of the initial temperature of the reservoir between a first initial temperature (hot case) and a second initial temperature of the reservoir (cold case)

[0086] The control member 5 can be configured to estimate the initial temperature of the reservoir between a first initial temperature of the reservoir corresponding to a state of the reservoir considered to be recently filled to the first initial density (hot case), and a second initial temperature of the reservoir corresponding to a state of the reservoir considered to be recently drawn down to a second initial density (cold case).

[0087] The first initial temperature Tini_l of the tank can be given by a predictive curve “TPgasliner_hot”, “TPlinercompo_hot”, “TPexternal_hot” (see [Fig.8]) as a function of the initial pressure Pini of the gas. In other words, for a given initial pressure Pini of the gas present in the tank, the predictive curve “TPgasliner_hot”, “TPlinercompo_hot”, “TPexternal_hot” makes it possible to determine the first initial temperature Tini_l of the tank.

[0088] The curves “TPgasliner_hot”, “TPexternal_hot”, “TPlinercompo_hot” can be provided by the second physical model. More specifically, these curves are obtained from the curve “Pgashot” in combination respectively with curves “Tgasliner_hot_bis”, “Texternal_hot_bis” “Tlinercompo_hot_bis ” (not shown) giving the evolution of the temperature as a function of time for respectively the internal wall, the external wall and the internal layer / external layer interface of the tank.

[0089] The “Tgasliner_hot_bis” curve is obtained from the “Tgasliner_hot” curve modulated by an “STgasliner” curve. The “Texternal_hot_bis” curve is obtained from the “Texternal_hot” curve modulated by an “STextemal” curve. The “Tlinercompo_hot_bis” curve is obtained from the “Tlinercompo_hot” curve modulated by an “STlinercompo” curve. The “STgasliner”, “STexternal”, “STlinercompo” curves are illustrated in [Fig.9]. They represent a thermal diffusion subsequent to the reference filling and which occurs at the level of the internal wall, the external wall and the internal layer / external layer interface of the tank.

[0090] In the example illustrated in [Fig.9], the thermal diffusion corresponds to a drop in temperature over time for the inner wall of the tank. On the other hand, for the outer wall and the inner layer / outer layer interface of the tank, thermal diffusion corresponds to an increase in temperature.

[0091] It should be noted that the curves “TPgasliner_hot”, “TPextemal_hot”, “TPlinercompo_hot” can be reduced respectively into a curve “Reduc_TPgasliner_hot”, a curve “Reduc_TPextemal_hot”, and a curve “Reduc_TPlinercompo_hot”, each defined between a value zero (1) corresponding to the lower limit TLgas_min_l of temperature and a value one (1) corresponding to the upper limit TLgas_max_l of temperature (see [Fig. 10]).

[0092] The second initial temperature Tini_2 of the tank is given by a predictive curve “TPgasliner_cold”, “TPextemal_cold”, “TPlinercompo_cold” as a function of the initial pressure Pini of the gas (see [Fig. 11]). In other words, for a given initial pressure Pini of the gas, the predictive curve “TPgasliner_cold”, “TPexternal_cold”, “TPlinercompo_cold” makes it possible to determine the second initial temperature Tini_2 of the tank.

[0093] It should be noted that the predictive curve “TPgasliner_cold”, “TPexternal_cold”, “TPlinercompo_cold” can be provided by the third physical model which simulates a reference emptying of the tank. At the end of this reference emptying, the temperature of the tank reaches a minimum value which varies across the thickness of the tank. For a tank formed of two layers, the minimum temperature Tmin at the inner wall, the outer wall or the inner layer / outer layer interface can be given for example by a linear law of type Tmin = A*Tamb +B,

[0094] where

[0095] Tamb is the ambient temperature;

[0096] A is between 0.5 and 1.5;

[0097] B is between -20 and -10.

[0098] The predictive curves “TPgasliner_cold”, “TPextemal_cold”, “TPlinercompo_cold” can be obtained from the curve “Pgascold” (giving the evolution of the gas pressure as a function of time) in combination respectively with the curves “Tgasliner_cold”, “Texternal_cold”, “Tlinercompo_cold” giving the evolution of the tank temperature as a function of time (cold case). The curves “Tgasliner_cold”, “Textemal_cold” and “Pgas_cold” are illustrated in [Fig.5].

[0099] Estimation of the reservoir temperature variation curve between a first reservoir temperature variation curve (hot case) and a second reservoir temperature variation curve (cold case).

[0100] The control member 5 can be configured to estimate, during filling, a temperature variation curve of the tank (see step S5 of the method illustrated in [Fig.2]). This estimation is made between a first curve of variation of the temperature of the reservoir (hot case) and a second curve of variation of the temperature of the reservoir (cold case). Said first curve and said second curve are determined during a step S4 of the method illustrated in [Fig.5].

[0101] The first variation curve has as its starting point the first initial temperature Tini_l of the reservoir. The second variation curve has as its starting point the second initial temperature Tini_2 of the reservoir.

[0102] The first curve (respectively the second curve) of variation of the temperature of the reservoir can be obtained from a model which simulates a reference filling of the reservoir from the first initial temperature Tini_l (respectively from the second initial temperature Tini_2) of the reservoir, with a pre-cooling of the gas at the source, and a thermal diffusion subsequent to the reference filling.

[0103] Effect of thermal diffusion on the predictive curves of the first initial temperature Tini_l of the reservoir or of the gas present in the reservoir

[0104] An analysis of [Fig.8] giving the reference temperature curves of the reservoir - or of the gas present in the reservoir - (i.e. the curves “TPgasliner_max”, “TPlinercompo_max”, “TPexternal_max”, “TPgas_max”), and the predictive curves of the first initial temperature Tini_l of the reservoir - or of the gas in the reservoir - (i.e. the curves “TPgasliner_hot”, “TPlinercompo_hot”, “TPexternal_hot”, “TPgas_hot”) shows that the gas and the internal wall of the reservoir see their temperatures decrease as a result of thermal diffusion, while the external wall and the internal layer / external layer interface of the reservoir see their temperatures increase as a result of thermal diffusion.

[0105] Taking into account this singular evolution of the temperature of the tank following thermal diffusion makes it possible to improve the accuracy of the reference temperature curve of the tank and / or the reference temperature curve of the gas present in the tank. Thus, the invention makes it possible to improve safety during filling by limiting the risk of exceeding the temperature limit value.

[0106] Examples of implementation of the reference temperature curve “Tgas_max” and the first gas temperature variation curve “Tgas_hot_bis”

[0107] When filling the tank, in order to prevent any overheating, the first curve “Tgas_hot_bis” of variation of the temperature of the gas present in the tank can be used alone (see example in [Fig. 13]) or in combination with the curve “Tgasmax” of reference temperature of the gas present in the tank (see example in [Fig. 14]).

[0108] As each of these examples shows, after connecting the vehicle to the end of the filling line (step 1) and initialization of the filling by the filling device (step 2), a first overheating is detected at the level of the gas present in the tank (step 3). In the example of [Fig. 13], the overheating is detected at the moment when the “Tgas_hot_bis” curve reaches the temperature limit value represented here by a horizontal line at 85°C. In the example of [Fig. 14], the first overheating is detected below the temperature limit value (85°C). This overheating corresponds to the moment when the “Tgas_hot_bis” curve “catch up” with the “Tgas_max” curve.

[0109] Following this first overheating, the control member stops filling while the tank is not filled to its maximum. The stoppage continues (step 4) for a certain required duration before a new filling cycle. Steps 5, 6, 7 correspond to this new filling cycle for which the initial temperature of the gas present in the tank is determined as a function of the temperature reached by the same gas at the end of the previous cycle, but also as a function of the pressure of this same gas at the start of the new cycle.

[0110] By taking into account the temperature reached during a previous filling cycle, the invention limits the risk of overheating during a subsequent filling cycle, and opens up the possibility of reaching a maximum charge in the tank.

[0111] Of course, the “Tgas_max” and “Tgas_hot_bis” curves of the examples above can be replaced by the “TPgasmax” and “TPgas_hot” curves to control the filling of the tank. When the stop condition relates to a limit value of the tank temperature, the filling can be controlled using one of the curves (hot case) giving the variation of the tank temperature (for example, the “TPgasliner_hot” curve). This variation curve can be used alone or in combination with the associated reference temperature curve (for example, the “TPgasliner_max” curve). Presentation of physical models

[0112] The physical models mentioned above at the origin of the different predictive curves presented in this description are based on a system of energy balance equations applied to the reservoir and to the gas present in the reservoir. These equations are presented below.

[0113] a. The internal energy balance equation of the gas in the reservoirs is written as follows:

[0114] = kj Sj + Q (1)

[0115] This balance involves the mass of the gas in the reservoirs ms, the internal mass energy of the gas in this reservoir us, the heat exchange between the gas in the reservoirs and the internal wall of the latter with the exchange coefficient and the surface internal of these tanks Sf, the total mass flow of hydrogen p) _ as well as ~ dt the incoming mass enthalpy of the gas hg iniet in the reservoirs. Ts and Twi are respectively the spatial average temperatures of the gas in the volume of these reservoirs and of the internal wall of the latter.

[0116] One of the two safety criteria not to be exceeded involves density in the model. As the total volume of the vehicle tanks Vchss (Compressed Hydrogen Storage System) can be known by the automaton, equation (1) can be written, using mg = VCHSS ps:

[0117] 4¾¾), - (T TS.nh <2) ' CHSS dt “ V ~ 'S> + & 'lg inlet

[0118] To solve this equation (2), it is necessary to discretize it by replacing the time derivatives with finite differences. We thus obtain the equation, by dividing by VCHSS:

[0119] {T^A)_T / t + A,))+_^_hsi^

[0120] where is the value of a time step (approximately one second).

[0121] The material balance equation in the reservoirs allows us to write:

[0122] VCHSS p^t + zlf) = V CHSS p^t) + Q (4)

[0123] or, by dividing by VCHSS; [°1241 ph + AtA ph}+^- <5)

[0125] Equation (5) allows us to obtain the value of + At)- Equation (3) therefore allows us to obtain the value of the internal energy up + Al):

[0126] / \ u J t + An = —~?—-------—_____ \ /

[0127] a. The energy conservation equation in the wall is written:

[0128] pc Ç = 2A(,Æ)(7)

[0129] This balance involves the density of the wall p, its heat capacity cp and its thermal conductivity X, as well as the radius r from the center of the tank.

[0130] At the interface between the gas and the liner, the flow continuity equation is written:

[0131] TL [ À —r / Ç (8) ' ' S 4 wij J Sf liner

[0132] At the interface between the composite and the ambient medium, the flow continuity equation is written:

[0133] _f —AS = k S (t -T

[0134] Where ke is the heat exchange coefficient between the ambient environment and the external wall of the tank, Se is the external surface of these tanks and Tamb is the tem-

[0135]

[0136]

[0137]

[0138]

[0139]

[0140]

[0141]

[0142]

[0143]

[0144]

[0145]

[0146]

[0147] room temperature. Finally, at the interface between the two liner and composite materials inside the wall, the flow continuity equation is written: (ân - / 373 v(10) a. Discretization of the equations in the wall To solve equations (7), (8), (9) and (10), it is necessary to discretize the wall radially into different meshes. The number of meshes in the liner n_liner and in the composite n_comp can vary. The wall temperatures are calculated at each node, and there are n_mesh = n_liner + n_comp + 1 nodes. The mesh size is therefore (5r)}jne[ = eUner / nliner in the liner and (dr) ecmp = ecamp / 11^ in the composite, as shown in Figure 13. The meshes at the interfaces have a thickness of ) j 2- In the illustrated example, 3 meshes are provided in the liner and another 3 meshes are provided in the composite. Equation (7), to be solved, is discretized and applied for each of the different cells constituting the wall. Figure 14 shows the radial discretization used to solve the equation. In this figure, a cell, centered on point P, is surrounded by a point W to the west and a point E to the east, the interior of the tank being to the west. Between points W and P is defined a fictitious face w, and between points P and E is defined a second fictitious face e. The distances between points W and P on the one hand, P and E on the other are noted respectively (5r) and (Ôr)e. The faces w and e are distant from ArP. The points W, P and E are located respectively at rP and rE from the center of the tank. Equation (7) is integrated over a time step and spatially from west to east, to obtain: A = fvJf (pc^rdrdt^ L[è(^dï)] drdt = The term on the left is called A, while the term on the right is called B. Assuming that the density p and the heat capacity cp change little between times t and t+At, the left-hand term is written: = (pc}^)r dr dt^npc^ ^dt^r dr - t + dt)-T(r, The assumption is made that the temperature difference T(r, t + dt) - T(r, t) is constant between points w and e and is equal to T(rP, t + dt) - T(rp, t)- The term A is written: f, (13) AP^p jf 3^^

[0148]

[0149]

[0150]

[0151]

[0152]

[0153]

[0154]

[0155]

[0156]

[0157]

[0158]

[0159]

[0160]

[0161]

[0162]

[0163]

[0164]

[0165]

[0166] By writing: T; , + - Av j - rw — ) - Arp ) tarnished A is finally written: 4 ~ nr T(xAt+At}-T(rP.t) . / rE + \ (14) A _ pCp---------Arp ___ j The term B, the right-hand term of equation (11), can be written: B = L [Xr& t)]dt(15) Assuming that ri _ jV+XMTw) and âL(r / 1 — , and £' / 6H dr\ w' / (ÔÀW assuming that the thermal conductivity X changes little between times t and t+At, the term B is written: T(rEA~T(^ (5r)e dt - -yw> t lz wJ t ? (16) We denote with an exponent 0 the values ​​taken at time t, and with an exponent 1 the values ​​taken at time t+At. We denote with an index w, P and E the values ​​taken at points W, P and E respectively. For example, the value T(r^ 14- Zfr) is denoted Tp- An implicit scheme is used, so equation (11) becomes: T' 1 P CP At . t^Tw (17) E (dr) ~AWrW(ffr) At the interface between the gas and the liner, the flow continuity equation (8) is dis cretized in the following manner: 7.0ç / T° Tl\_ ,0 c (18) A * 0 ~ ^liner^ l\ / v -liner At the interface between the composite and the ambient medium, the flow continuity equation (9) is discretized as follows: ]0 c Tnms / T„imA _ ? Op «yl \ (19) - Deposit (fr) ~ 1 nmesh “ 1 amb l Finally, at the interface between the two liner and composite materials inside the wall, the flow continuity equation (10) is discretized as follows: ,0 T^T1^ ,0 (20) ~~^p (5r)mmp The system of equations (17) for each point P that is not at an interface and (18), (19) and (20) can be solved by the Thomas algorithm, also called TDMA (for Tri-Diagonal Matrix Algorithm). Physical properties and coefficients a. Tg and Pg from p and u At each time step, the model calculates the temperature and pressure of the gas from the density p and the internal energy u of the gas: T„ = end u\ (21)

[0167] Pg = f^ y) (22)

[0168] These functions fi and f2 are adjusted to best match the thermophysical properties of the gas.

[0169] a. Heat transfer coefficients ki and ke

[0170] The model also needs values ​​for the heat transfer coefficients k; and ke. For the internal heat transfer coefficient k,, a function is used, calibrated from the correlation presented by Bourgeois, T. et al. (see “Bourgeois T, Ammouri F, Weber M, Knapik C. Evaluating the temperature inside a tank during a filling with highly-pressurized gas. International Journal of Hydrogen Energy 2015;40:11748-55”) [01711 Nu,Xn, = ^= aRa^,, + cRe^j <*>

[0172] For the external transfer coefficient ke, a function is used, calibrated from the correlation presented by Massard, F. (see Aide-mémoire du thermicien. ELSEVIER; 1997): 101731 [°1741 *.-1.02 4^. 10- <fe / )„,<102<24-2’

[0175] ke = 0.85, 102 < RaDexf < 104 (243)

[0176] ke = 0.48 104 < RaDext < 108 (24'4) [°177] ^0.1254« io8< W24*

[0178] Advantageously, the physical models described above can be associated with a heat transfer model between the filling device and the reservoir(s). This transfer model is described below.

[0179] The incoming gas mass enthalpy term h„ jnie! of equation (1) is obtained from the gas mass enthalpy at the dispenser using a heat transfer model. This aims to estimate the heat exchanges of the gas between the dispenser of the filling device and the entry into the vehicle's tanks, along the hose and other elements connecting the dispenser to the tanks.

[0180] The energy conservation equation applied to the gas inside the piping, called piping, is written in equation (25), for each elementary length dx along the piping. Here we assume that the piping has a homogeneous temperature along its entire length:

[0181] v ^'^gasjnfje _ / / Aj fr MT 4 f^(25) —KLpipAt jji ainipjpAi pipA'! 1 gas.pip^^ i1]

[0182] where nù is the mass flow rate of gas, cp^ is the specific heat capacity of the gas, x is the abscissa along the piping, Tgaspipe is the temperature of the gas in the piping

[0183]

[0184]

[0185]

[0186]

[0187]

[0188]

[0189]

[0190]

[0191]

[0192]

[0193]

[0194]

[0195]

[0196] dependent on x etf, is the internal exchange coefficient in the piping, dint.pipe is the average internal diameter of the piping and Tpipe is the temperature of the piping depending only on ?. Equation (25) can be integrated between 0 (the dispenser) and the position x along the piping to obtain the expression for T„aspipJ <x, i). Pour plus de clarté, nous notons = mcpœ / une longueur cara^ dTgaspipe / (TpipJ^-Tga^^ = 1 / / 4^(26.1) [- ln(TPip^- Tgaspip((x, = 1 / ^26-2) pipent) - Tga&:pip^x, t^KTpipj^-Tdi^ = ex}^-xI(26.3) The energy conservation equation applied to piping is written in equation (27): (wA - mAh ■ (t\-k ■ Mç ■ (t ■ (t\-T \ru^p)pipe dt ~ >^e,pipt.y pipt\l / 1 andA^J where(mcp) is the heat capacity of the piping, dhpioe is the energy transferred from the gas to the piping, kepjpe is the external exchange coefficient of the piping and SepijX is the external surface of the piping. The term rhAhpipJJ) can be written, using equation (26.3): d^Ahplp(^tj — mcp^Tpil^t) Tdisp^xp^ Lpipel1] (28.3) where Lptpe is the piping length. Equation (27) is discretized: (29)

Claims

Claims

1. Method (100) for filling a tank (1) with pressurized gas, implemented by a filling device (10) comprising a gas source (2), a filling line (3) connecting the source (2) to the tank (1), a valve (4) for regulating the flow rate and / or pressure in the filling line, and an electronic control member (5) configured to stop the filling when an estimated temperature (or density) of the gas present in the tank (1) reaches a temperature limit value (or respectively a density limit value), the method comprising, before filling the tank (1), a step (S1) of determining the ambient temperature (Tamb) at the filling device (10), a step (S2) of determining the pressure (Pini) of the gas present in the tank (1) and a preliminary step (S3) of estimating the initial temperature (Tini) of the gas present in the tank (1),the initial temperature (Tini) of the gas present in the tank (1) being a value estimated as a function of the ambient temperature (Tamb) and as a function of the pressure (Pini) of the gas present in the tank (1) before filling, said initial temperature (Tini) of the gas present in the tank (1) being greater than or equal to or less than or equal to the ambient temperature (Tamb).,

2. Method according to claim 1, characterized in that the initial temperature (Tini) of the gas present in the tank (1) is at least one of: a first calculated initial temperature (Tini_l) corresponding to a state of the tank (1) considered to be recently filled to a first initial density (pini_l), a second calculated initial temperature (Tini_2) corresponding to a state of the tank (1) considered to be recently drawn down to a second initial density (pini_2).

3. Method according to claim 2, characterized in that the first initial temperature (Tini_l) of the gas present in the tank (1) is included in a high temperature range having a determined lower limit (TLgas_max_l) greater than or equal to the ambient temperature (Tamb) and an upper limit (TLgas_max_2) corresponding to a determined maximum temperature limit, for example equal to 85°C.

4. Method according to any one of claims 2 or 3, characterized in that the first initial temperature (Tini_l) is determined from a first predetermined predictive curve “TPgas_hot” provided by a predetermined physical model simulating a reference filling of the tank (1) over the high temperature range.

5. Method according to any one of claims 2 to 4, characterized in that the second initial temperature (Tini_2) of the gas present in the tank (1) is included in a low temperature range having a determined upper limit (TLgas_max_2) less than or equal to the ambient temperature (Tamb) and a lower limit (TLgas_min_2) corresponding to a determined minimum temperature limit, for example between zero and -5°C.

6. Method according to any one of claims 2 to 5, characterized in that the second initial temperature (Tini_2) is determined from a second predetermined predictive curve “TPgas_cold” provided by a predetermined physical model simulating a reference emptying of the tank (1) over the low temperature range.

7. Method according to any one of claims 2 to 6, comprising a step (S4) consisting of modeling, during filling, a first temperature variation curve (or respectively a first density variation curve) of the gas present in the tank (1), having as starting condition the first initial temperature (Tini_l) (respectively the first initial density (pini_l)) of the gas present in the tank (1), and / or a second temperature variation curve (respectively a second density variation curve) of the gas present in the tank (1), having as starting condition the second initial temperature (Tini_2) (respectively the second initial density (pini_2)) of the gas present in the tank (1).

8. Method according to claim 7, comprising a step (S5) consisting of estimating, during filling, a temperature variation curve (or respectively a density variation curve) of the gas present in the tank (1), said curve being between the first temperature variation curve (respectively the first density variation curve) and the second temperature variation curve (respectively the second density variation curve).

9. Method according to any one of claims 2 to 8, characterized in that the first initial temperature (Tini_l) and / or the second initial temperature (Tini_2) is recalculated during filling as a function of the flow rate and as a function of the temperature of the gas present in the filling pipe, said flow rate and said temperature being determined by calculation and / or by sensors at the filling device (10).

10. Method according to any one of claims 1 to 9, characterized in that the temperature limit value (respectively the density limit value) is a determined fixed value, for example 85°C (respectively, for example 24.1 kg / m3 or 40.2 kg / m3) or a value given by a reference temperature “TPgas_max” curve (respectively by a reference density “RHO_cold” curve), said curve being provided by a predetermined physical model which simulates the thermodynamic conditions of the gas during a reference filling of the tank (1).

11. Method according to any one of claims 4 to 10, characterized in that the physical model is based on a system of equations comprising at least one of: - an internal energy balance equation applied to the gas present in the tank (1), - a mass balance equation applied to the gas present in the tank (1), - an energy conservation equation in a wall of the tank (1), - a heat flow continuity equation between the gas present in the tank (1) and the wall of the tank (1), - a heat flow continuity equation between the wall of the tank (1) and the ambient air, and - a flow rate equation linking a mass flow rate of the filling device (10) to a pressure difference between the filling device (10) and the tank (1).

12. Method according to claim 11, characterized in that the first initial temperature (Tini_l) (respectively the second initial temperature (Tini_2)) and the initial pressure (Pini) of the gas present in the tank (1) are obtained by solving said system of equations, for example using a tridiagonal matrix type algorithm.

13. Method according to any one of claims 1 to 12, characterized in that the ambient temperature (Tamb) of the filling device (10) and the initial pressure (Pini) of the gas present in the tank (1) are determined by calculation and / or measured by sensors at the filling device (10).

14. Method according to any one of claims 1 to 13, characterized in that the electronic control member (5) is configured to control the flow and / or pressure regulating valve (4) to produce a predetermined pressure curve or ramp during filling.

15. Method according to any one of claims 7 to 14, characterized in that the electronic control member (5) is configured to simulate and estimate the temperature variation curve and / or the density variation curve of the gas present in the tank (1) dynamically during the filling of the tank (1) and / or in advance, i.e. before filling.

16. Device (10) for filling a tank (1) with pressurized gas, comprising a gas source (2), a filling line (3) connecting the source (2) to the tank (1), a valve (3) for regulating the flow rate and / or pressure in the filling line (3), a set of sensor(s) (6a, 6b) configured to measure the pressure in the tank (1) and / or the ambient temperature at the filling device (10), and an electronic control member (5) configured to: - stop filling when an estimated temperature (respectively a density) of the gas present in the tank (1) reaches a temperature limit value (respectively a density limit value), - estimate, before filling the tank (1), an initial temperature (Tini) of the gas present in the tank (1),characterized in that the initial temperature (Tini) of the gas present in the tank (1) is a value estimated as a function of the ambient temperature (Tamb) and as a function of the pressure (Pini) of the gas present in the tank (1) before filling, the initial temperature (Tini) of the gas present in the tank (1) being greater than or equal to or less than or equal to the ambient temperature (Tamb).,

Images