D E S C RI P T I O N

The object of the present invention is a method and relative means for detecting fluid leakages in the sealing members, in the body or control mechanisms of a shut-off ball valve and then equipped with a double sealing means (or "seat") where each seat may be a unidirectional or bidirectional seat.

It is known that in valves equipped with unidirectional downstream seats an overpressure limit value (disengagement pressure Pd) exists between the valve cavity and the downstream duct above which the seat normally releases the fluid downstream.

There is a slight difference dp = Pd.eff - Pd.b between disengagement pressure Pd.eff verifiable during operation and disengagement pressure Pd.b previously measured at the test bench of a valve. This is irrelevant for the purposes of the methods applied by the invention; in any case, from now on, by disengagement pressure Pd it is meant the resulting disengagement pressure Pd.eff during operation.

The ball valves equipped with unidirectional downstream seat shall be hereinafter referred to as "pressure release".

If, on the other hand, the seat is bidirectional, the sealing thereof is theoretically perfect in both directions.

Several documents describe means for controlling the sealing of shut-off members. Document US 3 835 878 shows a shut-off valve the shut-off member whereof comprises a bellows and a compression spring. Such shut-off member can exert a variable pressure on the closing stop depending on the compression that can be exerted simultaneously by the spring, by acting on the stem, and by the gas trapped in the bellows. The closing stop, where the gasket presses, is crossed by a circular channel connected to a pressure gauge. The measurement of an anomalous pressure in the channel denotes an infiltration between gasket and stop. The device is not suitable to report leaks in the valve body (stem or flanges for closing the two half casings). The pressure detected by the pressure gauge in the case of gasket integrity varies with the squashing degree of the bellows for which it could be difficult to determine a pressure threshold below which there is a leakage from upstream or downstream and not a regular operating condition but with little squashed bellows.

Many documents indicate to interpose a pressure gauge in the pipe between two consecutive valves. In case of gate or poppet valves, such control methods can detect if there is a leak in the upstream or downstream valve but are ineffective for monitoring ball valves, in fact, if the downstream seat of the upstream valve is intact, it is not possible to detect leakages in the upstream seat thereof or in the valve body thereof and symmetrically, in the downstream seat or valve body of the downstream valve. It should be noted also that the inner volume of the pipe between two consecutive valves has effect on the pressure drop rate in the same pipe and can make the detection of fluid leaks uncertain or not timely.

"Pressure release" valves exist, that are valves in which, after their closure, there is provided a discharge device for the possibility/need of expelling to the outside the gas trapped in the cavity so as to let the pressure decreases to values close (but not necessarily equal) to the ambient pressure.

None of the documents mentioned provides teachings suitable for controlling the correct operation of the discharge device.

An object of the present invention is to eliminate at least in part such drawbacks by indicating means and methods applicable in a double seat ball valve where the downstream seat may be unidirectional, in order to determine, in the event of sealing defects in the line, that is, through the seats of the valve, if these are due to the upstream or downstream seat, or even sealing defects towards the outside due to elements of the valve body (stem gaskets or other).

In particular, the object of the invention is the automatic detection of such damages.

A further object of the present invention, in case said ball valve is equipped with a pressure release device, is to control the operation of said device.

A further object of at least some variants of the present invention is to gather information on the upstream, downstream and inner body pressure of a ball valve that can also be used for purposes other than the above.

Further features and advantages of the present invention shall be better highlighted by the following description of a method for the automatic detection of leakages and suitable means adapted for applying the method; the whole in accordance with the main claims and illustrated, purely by way of a non-limiting example, in the annexed drawing tables, wherein:

Fig. 1 shows a cutaway view of a ball valve with the ball in open position;

Fig. 2 shows a cutaway view of the ball valve of Fig. 1 but with the ball in the closed position and where pressure measuring devices and processing devices of data coming from said pressure measuring devices provided by the invention are schematically shown;

Figure 3 shows a detail of Fig. 2 consisting in a section according to a horizontal plane passing through the centre of the ball, with possible leakages of the fluid to be shut-off through the sealing seats of the spherical shutter of the valve;

Figure 4 shows, in a detail of Fig. 2, the zones of possible leakages of the fluid to be shut-off along the seat of the stem or through the junctions between the "closures" and the central body of the valve;

Figure 5 shows the same valve of Fig. 2 with the addition of a device for discharging to the outside the pressure of the fluid trapped in the cavity; Figure 6 shows, according to the same section of Fig. 3, the cavity of the inner body of the valve body and the flow of the fluid expelled to the outside by an expulsion device.

The following figures show charts of the pressure trend over time upstream, downstream and in the cavity of the body of a double seat ball valve.

More precisely, the charts shown refer to the case in which, following the closure of the valve at the initial time, the system reacts by keeping constant the upstream pressure P[N while the downstream pressure decreases continuously up to the value POUT:

- Figure 7 with double acting seat valve and perfect sealing;

in Figure 8 with unidirectional downstream seat valve and perfect sealing; in Figure 9 with unidirectional downstream seat valve and damaged downstream sealing;

in Figure 10 with unidirectional downstream seat valve and damaged upstream sealing;

in Figure 11 with unidirectional downstream seat valve and damaged sealing of the stem and/or valve body;

in Figure 12 with unidirectional downstream seat valve, gas expulsion device after valve closure and perfect sealing;

- in Figure 13 with unidirectional downstream seat valve, gas expulsion device after valve closure and damaged downstream sealing;

in Figure 14 with unidirectional downstream seat valve, gas expulsion device after valve closure and damaged upstream sealing;

in Figure 15 with unidirectional downstream seat valve, gas expulsion device after valve closure and damaged sealing of the valve stem and/or body.

A man skilled in the art shall observe that some of the charts just listed have the same trend as if the downstream seat was bidirectional but this is not relevant for the purposes of the invention.

Unless otherwise specified, in this report any possible spatial reference such as the terms up/down, front/rear, right/left etc. refers to the position in which the elements are represented in the attached figures; as well as the concept of upstream/downstream relates to the temporary origin of the flow (which may also reverse).

With reference to Fig. 1, of a ball valve 10 with two sealing seats, drawn in the open position, there are shown the upstream seat 11 and the downstream seat 12, the ball shutter 14, the stem 16 for driving the ball shutter 14, a first "closure" 19, a second "closure" 18 that are parts of the valve body among which the central body 17 is interposed and the sealing gaskets 15 between the central body 17 and the two "closures" 18 and 19. Both the downstream seat 12 and the upstream seat 11 may be both unidirectional and bidirectional depending on what specified below.

In Fig. 2 there are also shown the inlet IN of the fluid to be shut-off, a cavity V of stagnation of the fluid when shut-off, and finally the outlet OUT of the fluid to be shut-off.

The arrow indicates the direction of the flow of the fluid from the inlet IN to the outlet OUT (i.e. from upstream to downstream).

It is clear from the figure that the cavity V of stagnation of the fluid is the whole set of space portions communicating with each other and as a whole delimited by: central body 17 of the valve 10, outer surface of the ball shutter 14, seats 11 and 12 and finally "closures" 18 and 19 of valve 10. When and only when the latter is in the closed position, as may be well seen e.g. in Fig. 3, the cavity V as defined above also communicates with the through hole 20 of the ball 14. When the seals 11 and 12, both with ball 14 open and ball 14 closed, are intact, the cavity V remains completely isolated from the upstream and downstream pipe of the valve and, with ball 14 open, also from the through hole 20.

Fig. 2, where the ball valve 10 is shown in the closed position, further shows a first channel 22, and further two possible second and third channel 21 and 23 that indicate respectively a connection between the cavity V and a sensor sPy of the cavity pressure Py of the cavity V _; a possible connection between the inlet IN and an upstream sensor SPIN of the upstream pressure Pn^; a possible connection between the outlet OUT and a downstream sensor SPOUT of the downstream pressure POUT- The said connections symbolized by the second and third channel 21 and 23 are preferably but not necessarily made in the "closures" 18 and 19 of the valve 10 (as shown in the figure); alternatively they can be made in the means for the passage of the upstream and downstream fluid to the valve 10; e.g. in the pipe connected to the inlet IN and in that connected to the outlet OUT, respectively.

The said pressure sensors sPv., sP^ and SPOUT are symbolized as three U-shaped manometers but actually they are three pressure transducer sensors capable of sending signals representative of the pressure measured to an electronic processing unit indicated in the figure with CPU.

Figs. 2 and 3 also show, symbolized by arrows, the possible leakages from the inlet IN to the cavity V and from the cavity V to the outlet OUT respectively through the one and the other of the two seats 1 1 and 12.

Figure 4 shows in detail the possible leakages to the outside of the valve of the fluid to be shut-off through the seals 116 of the stem 16 or the sealing gasket 15. If the valve 10 is with pressure release, at a time t ₂ subsequent to the time ti in which the closing of a ball valve 10 actuated, its cavity V is put in communication with an environment in order to discharge at least partially the pressure thereof. Such communication is interrupted immediately after such emptying, at the time of emptying end t ₃ .

The connection with such environment takes place through a discharge device EV (such as a solenoid valve or equivalent device) which is shown in Figs. 5 and 6; the black arrows of Fig. 6 represent the flow of the evacuated fluid.

The said environment consists generally, but not necessarily, in the external environment at atmospheric pressure but, for the purposes of the invention, can be any environment that is at a constant ambient pressure P _amb , not influenced by the operation of the valve 10 nor by the discharge device EV neither, finally, by the operating conditions of the system in which the valve 10 operates. The ambient pressure P _amb is assumed of known value and equal to the atmospheric pressure or in any case lower than the pressures PIN and POUT- The method of detection of possible leakages according to the invention allows to assess whether there is fluid leakage and what the cause is among the three that, as will be seen, are identifiable according to the same invention.

The method also allows to detect an abnormal operation of the optional discharge device EV.

The method is applicable to at least three possible configurations of the valve 10:

- valve 10 with unidirectional downstream seat 12;

- valve 10 with unidirectional downstream seat 12 and with pressure discharge device EV;

- valve 10 with bidirectional downstream seat 12 and with pressure discharge device EV.

The method allows to identify and distinguish the following four possible sealing states of the valve 10:

- valve 10 intact;

- valve 10 with sealing defect in the downstream seat 12;

- valve 10 with sealing defect in the upstream seat 1 1 ;

- valve 10 with sealing defects of cavity V (that is sealing defects towards the outside of the seals 1 16 at the stem 16 or sealing gasket 15 or due to other causes such as micro-cracks; or sealing defects of the device EV).

The method achieves its purposes by the fact that, as will be shown later, the trend over time of the cavity pressure Py, at least during suitable time intervals defined below, in each of the above three configurations, has a characteristic profile which depends on the four possible sealing states.

The method involves the collection and analysis of the trend over time of the cavity pressure Pv _. in one or more of said suitable time intervals where "suitable" means time intervals within the set of which said trend over time appears to be unequivocally typical of only one of the above four possible sealing states. Such time intervals are comprised between the valve 10 closing time t _\ and a stabilization time which coincides with the moment in which the disturbance of the cavity pressure Pv due to the closing operation of the valve 10 and the optional subsequent actuation of the possible discharge device EV, indicated with t ₃ can be considered as ceased. By stabilization it is meant either that the cavity pressure Pv has substantially taken a typical constant value or that has taken a typical value ranging that has become a typical fluctuating value with a fluctuation (as shall be seen) of width substantially equal to Pa- Although the above analysis of the cavity pressure Py is sufficient to verify the integrity state of the seals of the valve 10, it can be useful to analyse, also or alternatively, the trend over time of at least one of the following pressure differentials:

- the upstream differential ΔΡ^ = PIN - Pv equal to the difference between the upstream pressure PIN and the cavity pressure Py,

- the downstream differential ΔΡουτ = Pv - POUT equal to the difference between the cavity pressure Py and the downstream pressure POUT- AS shall be seen, the trend over time of said differentials is strongly characterised by the state of the valve seals 10 and therefore simplifies the processing of the acquired data necessary to identify the sealing state of the valve 10.

It should be noted that also the variants of the method of the invention that provide for only the analysis of the upstream differential APnsj = PIN - Pv and/or of the downstream differential ΔΡουτ = Py - POUT, need the measurement of the cavity pressure Py, being these differentials magnitudes functions of the same cavity pressure Py.

The identification and reporting of the sealing state of the valve 10 make you of the processing unit CPU.

According to the invention, the processing unit CPU, with the techniques within reach of information technology experts, is capable of analysing the trend over time of the cavity -pressure Py or, more generally, of signals that are function of the trend over time of the same cavity pressure Py, understanding from such trend the sealing state of the valve 10 and also, in case of leakages, if they can be considered acceptable or to be reported for their elimination because excessive. By way of a non-limiting example, the processing unit CPU may perform at least the following operations:

- receiving from the pressure sensor sPy and optionally from the pressure sensors SPIN and SPOUT, in said one or more suitable time intervals comprised between the valve 10 closing time ti and the stabilisation time t ₄ , signals that are function of the trend over time of the cavity pressure Py as well as, optionally, of the upstream ΡΠΜ and downstream POUT pressures;

- analysing the trend over time of those among said signals which, depending on the configuration, is necessary or preferred to detect; i.e.: cavity pressure Py; upstream pressure PIN; downstream pressure POUT and _/ or directly of the pressure differentials ΔΡΙΝ = PIN - P _v (upstream differential) and ΔΡουτ = Py - POUT (downstream differential) received;

- processing values representative of the deviations among said trends over time acquired with other standards stored in it CPU and corresponding to the pressure trends in the same suitable time intervals with the same valve 10 free from leakages;

- having in memory values representative of the maximum acceptable deviations between said received trends over time and said stored standards;

- comparing the values representative of the deviations processed with those stored;

- identifying, by said comparison, the state of said valve (10) among the said four possible states;

- sending a signal representative of possible deviations which exceed the acceptable maximums, such signal being able to simply denote that there is fluid leakage or even indicate what the cause is among the three clearly identifiable ones according to the invention, that is: sealing defects of the upstream 11 or downstream 12 seat or, finally, sealing defects of cavity V. Such signal, purely by way of an example may be a simple visual and/or acoustic and/or, finally, electrical alarm signal to further devices.

This signal may be output even in normal situations to confirm that the seal is intact.

Finally, such a signal may be suitable to indicate the actual trend of the pressures and also for purposes other than those of the invention.

The figures starting from Fig. 7 show charts of the trend over time of pressures PIN, PV and POUT at the inlet IN, in the cavity V and at the outlet OUT detectable by the three pressure sensors SPIN, sPy and SPOUT starting at least from the time ti in which the closure of the ball valve 10 actuated up to when the effects of the closure of the valve 10 on the pressures ceased.

The charts show pressure profiles for valves 10 of different type both in conditions of perfect sealing and affected by leakages.

In fig. 7, as in the subsequent, the trends over time of the pressures PIN, P _V and POUT are drawn as well as the pressure differences P^ = PIN - P _V and ΔΡουτ ⁼ Pv - POUT-

The axis of the abscissas represents the times in a generic unit of measure which may, but not necessarily, consist in seconds. The trends over time of the pressures, in fact, while having the profiles qualitatively represented in the figures, have temporal development which depends, as it is obvious, on various construction features of the valve 10 and, on the operating pressures and on the characteristics of the shut-off fluid.

The axis of ordinates represents the absolute pressures in a unit of measure in which PIN = 1.

In the ideal case of perfect sealing of all the members and of a valve 10 with bidirectional seats (see Fig. 7), the cavity V, at the closing thereof, will keep its cavity pressure Py constant and equal to PIN which was the existing pressure in cavity V with the valve 10 open. As a consequence the pressure difference ΔΡΙ = PI - Pv is kept constantly equal to zero. The pressure POUT, instead, drops to a constant value dictated by the downstream system.

In case of damage to the sealing members of such valve 10, the pressure gradients drive the leakage flows through the same seals.

In these conditions it is easy to predict the flows of any leakages and the consequent variations of pressures in the three environments.

A sealing defect of the downstream seat 12 produces a leakage from cavity V towards the outlet OUT through the same downstream seat 12 (see Fig. 3) while the cavity pressure Py moves to the value POUT.

If the seals of cavity V are defective towards the outside environment there will be leakages from cavity V of the valve 10 towards the outside environment (see Fig. 4) while the cavity pressure Pv moves to the value of the ambient pressure

Pamb-

If, however, there is a sealing defect of the upstream seat 1 1 and the pressure P( is kept constant after the closing of the valve 10, there is no leakage from the inlet IN towards the cavity V nor the variation of the cavity pressure Pv because this remains constant and equal to PI exactly as in the case seen above of intact upstream seat 1 1. The only possibility of leakage is if the pressure PI varies; in such case Py follows its value while, in the case of perfect sealing, would remain at the value taken at the time of the last closing of valve 10.

Therefore the identification, by monitoring the pressures P[N and/or Py of a sealing defect of the upstream seat 1 1 , in the case of valves with bidirectional downstream seat 12 and where the release of pressure is not provided it is possible only if the pressure PIN varies over time and in such case is denoted by the fact that Py follows its value.

The above pressure changes in presence of leakages are not shown in any of the annexed figures as this brief discussion makes them evident to a man skilled in the art.

According to the invention, however, the identification of sealing defects by monitoring one or more of the pressures Py and POUT or their differentials ΔΡΙΝ or ΔΡουτ is possible in the following situations involving a valve 10 with unidirectional downstream seat 12.

The case in which it is provided that the cavity V remains under pressure after the closure of the valve 10 (absence of the discharge device EV) shall be examined first.

For valves 10 with unidirectional downstream seat 12, even in the ideal case of seals perfectly intact (see Fig. 8), there is a limit value of overpressure (disengagement pressure Pd) between the cavity V and the downstream duct above which the seat normally releases the fluid downstream. Thus, the cavity pressure Py that from the opening to the closing time of the valve 10 (from t ₀ to t ₂ in the figure) is kept equal to PIN, then follows POUT keeping a constant positive offset equal to Pd and the downstream differential ΔΡουτ stabilizes to the value Pd because as soon as the pressure difference astride of the downstream seat 12 exceeded the value Pd, there would be the release of fluid downstream from the cavity V up to restoring the dynamic balance between the forces/pressures acting on the same seat 12. Once this behaviour as POUT, varies has been characterised, which may occur during testing or with the valve 10 in operating conditions but certainly intact, it is easy to recognize some abnormal operating conditions attributable to damages to the seals.

The pressure profiles in case of damafie to the downstream seat 12 are shown in Fig. 9 where the curve indicated with "seal Py", as in subsequent figures, represents the cavity pressure Py expected in case of absence of defects of the downstream seat 12. Because of the damage, the profile of the cavity pressure Py over time may not reach the respective ideal values; the leakage of fluid downstream accelerates the decrease of the pressure in the cavity V; in particular there will no longer be the initial stretch at a constant pressure equal to P^ : Py drops with a weak head then, as soon as the downstream differential ΔΡουτ ⁼ Pv - POUT exceeds Pd, there is a quick disengagement of the downstream seat. Finally, once POUT has reached its operating value, it will be noted that Py drops slowly until the latter due to the leakages of the damaged seal; the consequence of this is that ΔΡουτ after a short stretch of positive value, drops and keeps to the value 0.

The pressure profiles in case of damaRe to the upstream seat 11 are shown in Fig. 10. The profile of the cavity pressure Pv over time initially follows the respective ideal values, up to the stabilization of POUT; then the fluid leakage from the upstream seat towards the cavity V will gradually raise the same cavity pressure Pv until the downstream differential ΔΡουτ ⁼ Pv - POUT exceeds Pd and therefore there is the quick disengagement of the downstream seat and a quick drop of Pv to a value comprised between Ρ ₀ υτ and POUT + Pd. At this point the sequence repeats cyclically with a gradual increase of Pv up to the disengagement and subsequent sudden drop. Correspondingly, the downstream differential ΔΡουτ ranges between the values 0 and Pd.

The pressure profiles in case of damage to the seals of the body 15 and stem 116 of the valve 10 are shown in Fig. 11. The profile of the cavity pressure Pv over time will initially follow the same trend as in the case of damage to the downstream seat up to the stabilization of POUT (a weak head); then, as soon as the differential Pv - POUT exceeds Pd, there is the quick disengagement of the downstream seat). However, once POUT has reached its operating value, it will be noted that Pv will continue to drop slowly well below POUT up to tending to the ambient pressure P _am b ₅ this due to the progressive emptying of the cavity for sealing defects towards the outside of the body of the valve 10. As a consequence, the downstream differential ΔΡουτ stabilizes to negative values. In the case of "pressure release" valves 10, that is, in those gas applications in which, after the closing of the valve 10, there is provided the possibility/need of expelling to the outside the gas trapped in the cavity so as to let Pv drop to values lower than POUT, close (but not necessarily equal) to the ambient pressure P _am b, also the downstream sealing is subjected to a pressure difference from the outlet OUT towards the cavity V, resulting, from the time t ₂ , POUT > Pv≥ Pamb- The charts of the pressure profiles are more complex but always suitable for providing information on any damages to the seals. In the ideal case of perfect sealing (see Fig. 12), as soon as the discharge device EV is actuated at time t = t ₂ , subsequent to the closing of the valve at time t = t _1; it lets Pv drop rapidly from the value equal to PIN up to values close to the pressure of the outside environment P _am b, after which all the values of pressures remain stationary even after the re-closure of the discharge device EV (at time t = t ₃ ). As a consequence, if the seals 1 1 and 12 are perfect, the downstream differential ΔΡουτ = Pv - POUT after reaching a maximum positive value equal to PIN - POUT within t = t ₂ , drops rapidly and stabilizes to negative values. As for the upstream differential ΔΡΙΝ = PIN - Pv, it has a zero value until the time t = t ₂ then stabilizes to positive values.

In the case of damage to the downstream seal (see Fig. 13), after the closing of the valve 10 at t = ti, the profile of the cavity pressure Pv over time cannot keep the ideal values because between the start of the decrease of POUT and the opening of the discharge device EV (for t = t ₂ ) a gradual decrease of P _v (which in the case of perfect sealing is kept constant and equal to P _O M) may already occur. Once the discharge device EV intervenes, Py has a quick drop up to a value close to the ambient pressure P _amb . After the closing of the discharge device EV at t = t ₃ , the downstream leakage towards the cavity through the damaged downstream seat 12 produces a slow and gradual increase of Pv that will tend asymptotically to the value of POUT- The downstream differential ΔΡουτ - Pv - POUT will tend to zero from negative values. As for the upstream differential ΔΡΠ _Μ = PIN - Pv, it increases slightly from a zero value at time t] until shortly after t ₂ (while in the case of perfect seals it was kept to zero until t ₂ ), then it rises sharply until the time t = t ₃ to then have a strong head and then stabilize to positive values.

In the case of damage to the upstream seal (see Fig. 14), the profile of the cavity pressure Py over time follows the respective ideal values by keeping constant and equal to PIN until the actuation of the discharge device EV; then there is a quick drop of P _v up to the ideal value (close to the ambient pressure); from now on there is a leakage from upstream towards the cavity through the damaged seat 1 1 with corresponding slow and gradual increase of Py substantially as in the case of damage to the downstream seat 12. In this case the gradual filling of the cavity (following the closure of the discharge device EV at t = t ₃ ) raises the Py well beyond the POUT- AS soon as Py reaches the value of POUT + Pd the disengagement of the unidirectional downstream seat occurs with consequent reduction of Py up to values close to POUT, then the cycle of gradual filling and quick release of the fluid repeats. The downstream differential ΔΡουτ = Pv - POUT after time t ₃ , will tend to range between 0 and Pd. As for the upstream differential ΔΡΙ = PIN - Pv it keeps the zero value until t ₂ to then rise sharply up to t ₃ to then have a strong head and finally take a fluctuating trend of amplitude P _d -with positive values.

In case of damage to the seal towards the outside of the body of valve 10, which includes any leakages of the device EV if present (see Fig. 15) the condition is more difficult to detect because the profile of the cavity pressure P _v over time will follow the same trend as the case of perfect sealing except that, between the start the decrease of POUT (at time t = tj of closing valve 10) and the opening of the discharge device EV (at time t = t ₂ ), a gradual decrease of Py (which, instead, in the ideal case is kept constant) occurs. Once the discharge device EV is opened (at t = t ₂ ) there is a quick drop of Py up to the ideal value (close to the ambient pressure). Even after the closing of the discharge device EV (t = t ₃ ) the pressure/time profiles do not change anymore.

This type of damage is then distinguishable from the case of perfect sealing if the time t ₂ of intervention of the discharge device EV is distanced from that ti of closure of the valve, in a way sufficient to appreciate the initial gradual decrease of Py in the range between tj and t ₂ .

It should be noted that in such situation the downstream differential ΔΡουτ = Pv - POUT, within t = t ₂ takes a maximum positive value smaller than in the case of perfect sealing, being Py <PIN due to the leakage. After the time t ₂ , ΔΡουτ moves to markedly negative values as in the case of perfect sealing.

As to the upstream differential ΔΡΙ = PIN - Pv from the zero value up to the time ti it increases slightly between the times t _\ and t ₂ (while in case of perfect seals it kept equal to zero in the same range) to then stabilize to positive values.

It should be noted (see Figs. 12 to 15) that in all modes of operation of the "pressure release" valve 10, regardless of the fact that the same valve is intact or subject to leakages of any kind, the trends of Py and ΔΡουτ have a marked head after the time t ₂ while ΔΡΙΝ has a strong rise; such trends are absent in the charts relative to valve 10 without release of pressure because characteristic of the successful opening of the discharge device EV. Therefore the analysis of Py or ΔΡουτ or, finally ΔΡ _Γ Ν, also allows to signal if after the closing of the valve 10 the opening of the discharge device EV has occurred or not, fact that, at least, allows to signal the operation failure of the same valve but also, if not know otherwise, what the current mode of operation is, that is, if with pressure release or not.

It was therefore shown how the temporal profiles of the cavity pressure Py and pressure differentials ΔΡΙΝ and POUT are different in all the cases examined and such, therefore, as to allow, through their interpretation, to signal if and where sealing or malfunction defects of the discharge device EV are present.

In most cases, however, the measurement and the analysis of all the other three pressures PIN, PV and POUT and their differentials ΔΡι _Ν and ΔΡουτ is not necessary and the substantially only qualitative study of their trend over time in the already said suitable intervals is sufficient.

For the various possible situations relative to a valve 10 with unidirectional downstream seat 12 where the release of pressure is not provided the profiles that take the cavity pressure Py and the downstream differential ΔΡουτ ⁼ Pv - POUT are hereinafter recalled.

1. Seals perfectly intact.

Pv from the opening to the closing time of the valve 10 is kept constant (equal to P^) then it drops to a second value which is also kept constant (equal to POUT + Pd).

The downstream differential ΔΡουτ stabilizes to the value Pd. 2. Damage to the downstream seat 12.

Pv _, from the opening to the closing time of the valve 10, drops with a weak head then more sharply to a second value which is kept constant as well (equal to POUT).

The downstream differential ΔΡουτ after a short stretch of positive value, drops and stabilizes to the value 0.

3. Damage to the upstream seat 11.

Pv, from the opening to the closing time of the valve 10, is kept constant (equal to ΡΓΝ) then it drops and takes a cyclical pattern.

The downstream differential ΔΡουτ stabilizes in a cyclic fluctuation between the values 0 and Pd.

4. Damage to the sealing of the body of the valve 10.

Pv, from the opening to the closing time of the valve 10, drops first with a weak head then drops more sharply and stabilizes to P _amb - The downstream differential ΔΡουτ stabilizes to negative values.

It should be noted that for a valve 10 with unidirectional downstream seat 12 where the release of pressure is not provided the analysis of the trend of the downstream differential ΔΡουτ alone is sufficient to distinguish the four previous situations in which it is markedly different. In fact, it is sufficient to acquire the value ΔΡουτ at time t ₄ where Py may be considered as stabilized.

However, also the trend of Py is distinguishable in the various cases provided that the pressure measurement is sufficiently accurate to appreciate the presence or not of a weak initial head.

The application in which the pressure release valve 10 is provided (where the shut-off fluid must be expelled from the cavity V after the closing of the same valve) is now summarized.

The profiles trend of the cavity pressure Py of the downstream differential ΔΡουτ and upstream differential ΔΡΙΝ = PI - Pv is recalled.

5. Perfect sealing.

Py drops quickly from a constant value equal to PI to a constant value substantially equal to P _am b-

The downstream differential ΔΡ ₀ υτ = Pv - POUT from the initial zero value reaches a maximum positive equal to PIN - POUT within t = t ₂ then stabilizes to negative values.

The upstream differential ΔΡΙΝ = PIN - Pv has zero value up to the time t = t ₂ then stabilizes to positive values.

Damage to the downstream sealing.

Pv changes from a constant value equal to PIN to a value substantially equal to P _am b to then rise again tending to POUT-

The downstream differential ΔΡουτ tends to zero from negative values. The upstream differential ΔΡΙ = PIN - Pv from the zero value to the time ti rises slightly until just after t ₂ then it rises sharply up to the time t ₃ , then it has a strong head and finally stabilizes to a positive value. Damage to the upstream sealing.

Pv is initially kept equal to P[N then drops substantially up to P _amb , then rises back over the POUT and finally takes a cyclical pattern.

The downstream differential ΔΡουτ stabilizes in a range between the values 0 and Pd.

The upstream differential ΔΡΙ = PIN - Pv keeps the value zero until t ₂ then rises sharply up to t ₃ then has a strong drop and finally takes a fluctuating pattern of amplitude Pd with positive values.

Damage to the sealing of the body of the valve 10.

Pv prior to the actuation of the discharge device EV has a gradual decrease then, once the discharge device EV is actuated, it drops quickly to a constant value substantially equal to P _amb .

The downstream differential ΔΡουτ = Pv - POUT, from the initial value zero reaches a positive maximum equal to PI - POUT within t = t ₂ such a positive maximum is smaller than in the case of perfect sealing (due to the already mentioned head experienced by P _v ) then takes a negative value. The trend is very similar to that provided in the case 5. The upstream differential ΔΡΠΜ = PIN - Pv has zero value up to the time ti, increases slightly between the times t _\ and t ₂ and then stabilizes to positive values.

It should be noted that also for a pressure release valve 10 the analysis of the trend of the upstream differential ΔΡΙΝ alone or of the trend of the cavity pressure Py is sufficient to distinguish the previous four situations provided that the pressure measurement is sufficiently accurate.

It should also be noted that, for a valve 10 with unidirectional downstream seat, the analysis of the trend of at least the cavity pressure Pv and/or pressure differentials ΔΡΙΝ and/or ΔΡουτ allows to distinguish any situation and to report the operation failure of the discharge device EV.

In particular, and merely by way of a non limiting example:

a) the analysis of the trend of the downstream differential ΔΡουτ = Pv - POUT, even limited to when the cavity pressure Pv can be considered stabilized, is sufficient to identify the four possible sealing states for a valve 10 without release of pressure that therefore can clearly identify; b) the analysis of the trend of the upstream differential ΔΡΙΝ = P _I N - Pv is sufficiently different in the four possible sealing states for a pressure release valve 10 that therefore can clearly identify;

c) the analysis of the trend of the cavity pressure Pv alone, allows to distinguish all eight situations.

Now that it has been clarified what the effects of the unidirectional downstream seat 12 and of the discharge device EV are on the cavity pressure Py it can be observed that even the sealing of the upstream seat 1 1 of a valve 10 with bidirectional downstream seat 12 is controllable if provided with a discharge device EV that opens and re-closes at times t ₂ and t ₃ .

The complete identification of the sealing state of the valve 10, also recalling what already said, takes place after the opening and re-closing of the discharge device EV and in the following way:

- if the cavity pressure Py moves to the value POUT, then, without waiting for the opening and re-closing of the discharge device EV, it is possible to conclude that there is a sealing defect of the downstream seat 12;

- if the cavity pressure Pv moves to the value P _amb? then, without waiting for the opening and re-closing of the discharge device EV, it is possible to conclude that, there is a sealing defect of the seals of cavity V;

- if the pressure PIN varies over time and the cavity pressure Pv follows its value, then, without waiting for the opening and re-closing of the discharge device EV, it is possible to conclude that there is a sealing defect of the upstream seat 11 ;

- if the cavity pressure Pv remains constant and equal to PI then it is necessary to wait for the actuation of the discharge device EV after the re-closure whereof at time t ₃ , if Pv after dropping to P _amb rises again (and correspondingly ΔΡΙΝ tends to zero), then there is a sealing defect of the upstream seat 11, otherwise if Pv remains equal to P _amb (and correspondingly ΔΡΙ tends to a positive constant value), there are no sealing defects.

Therefore, also according to such a method, the analysis of the trend over time of Pv alone is sufficient to identify the sealing state of the valve 10.

For greater accuracy in the analysis of the trends over time and/or a simplification of the same analysis procedure, it may be useful that processing unit CPU also receives signals of the opening/closing times of the valve 10 and/or of the discharge device EV although these can be identified by the same trends over time.

It is clear from the trend over time of the pressures in the commented charts that many other features may be chosen by a man skilled in the art, depending on the accuracy required and the operating conditions, to identify the state of the valve 10 without departing from the scope of invention.

For example in some cases it may be appropriate a redundancy of the monitored pressures, and therefore also the pressures PIN and/or POUT beyond those that above have been seen to be theoretically sufficient depending on the configurations; in other cases the analysis may need to be not only qualitative for the study of the trend over time but also quantitative; in other cases, finally, the pressures Pt _N and/or POUT can be assumed as certainly known and constant and therefore not needing measurement.

In conclusion, according to the simplest variant of the invention, the study of the trend over time of the cavity pressure Py in cavity V, alone, is sufficient to provide indications on the sealing state of the valve 10.

For a valve 10 without release of pressure, however, it may be sufficient or more useful the analysis of only the downstream differential ΔΡουτ ⁼ Pv - POUT- For a pressure release valve 10, however, it may be sufficient or more useful the analysis of the upstream differential ΔΡΙΝ = PI - Pv alone.

In general, however, and preferably, the device according to the invention shall have means adapted to monitor and analyse both ΔΡουτ and ΔΡΙ SO that it is usable without distinction for valves 10 with or without pressure release.

In addition, since, as said, the reading of the cavity pressure Py and/or of the pressure differentials Y^ and/or ΔΡουτ allows to know if the discharge device EV is present and operating, the processing unit CPU may also decide autonomously which of the pressure differentials ΔΡΙΝ or ΔΡουτ to use for its analyses.

As for the cases in which the analysis of ΔΡουτ = Pv - POUT limited to when the cavity pressure Pv is stabilised (and thus at any one moment even very distant from the closing time) is sufficient it is however preferable that the analysis is carried out close to the closing time in order to obtain information on the state of the valve 10 as soon as possible.

As for the pressure sensors usable, it has been seen from the above description that when only pressure differentials ΔΡουτ and/or ΔΡ _Ϊ Ν, are intended to be studied for the analyses, it would be sufficient to provide only a corresponding downstream SAPOUT and/or upstream SAPJN differential sensor for the direct measurement of the upstream ΔΡΙΝ = PIN - Pv and/or downstream ΔΡουτ ⁼ Pv - POUT pressure differentials. Nevertheless, it is preferred to have three absolute sensors for the direct measurement of PI , PV and POUT both because they provide more precise measurements and because their indications, even if they were redundant for the purposes of the invention, can provide information usable for other purposes.

It is unlikely that sealing defects are simultaneously present in the upstream 11 and downstream 12 seat and, again, in the body of the valve 10 or in EV because, of course, at each alarm signal the repair of the reported defect is assumed to follow. In any case, the occurrence of such an unlikely situation would correspond to a trend over time of the cavity pressure Pv and/or differential pressures ΔΡ^ and ΔΡουτ or abnormal to such an extent that it is not interpretable by the processing unit CPU or misleading (i.e. reporting a defect other than the actual one). Then, in case of anomalous trends that are not identifiable, it is sufficient that the CPU outputs a signal for an unidentified anomaly; in the case of misleading trends, instead, this could lead to a non- resolutory repair but the error signal will be repeated again when put back into operation, and this time it will provide the correct indications.

From the charts shown in the figures it is clear that the pressures trends over time are representative not only of the state of the valve 10 but also of the magnitude of the leakages. This is in fact inferable from the variation and stabilisation speed of Pv and/or, as a result, of ΔΡΙΝ and ΔΡ ₀ υτ· Therefore the analysis of trends over time of the pressures provides not only qualitative but also quantitative information allowing to decide if the defect is negligible or to be eliminated.

If the maximum acceptable deviations between trends of ideal and actual pressures are lower than those considered to correspond to dangerous conditions, the method according to the invention prevents any situation of actual risk, even before it occurs.

Once it is determined that the sealing state of valve 10 is assessed, according to the invention, by acquiring and analysing the trend over time of signals which are function of the trend over time of the existing cavity pressure Py inside cavity V, it is evident that many methods for detecting sealing defects alternative to the one provided by way of an example are possible according to the invention, that are even simpler and possibly not as accurate as those described. For example, all the methods of analysis described above occur after a closing operation of the valve 10 but it is also possible to know in advance (i.e. before the closure of valve 10) if the seals of the seats are intact and to have the guarantee that, once the valve is closed, there are no leakages from upstream to downstream.

In principle, since the cavity V remains always completely isolated when the valve is open, the cavity pressure Py that exists therein should always keep a constant value and any change thereof, therefore, should be an indication of sealing defects. In fact, the cavity pressure Py may have variable values due to previous opening/closing operations or even for imperceptible leakages that brought it to the ambient pressure P _amb or to the operating pressure of the fluid flowing in the valve 10. Variations of such pressure, therefore, cannot be considered sufficiently indicative of possible leakages or at least of leakage requiring interventions.

Having said that, if the discharge device EV is provided, and remembering that, with the valve open, the cavity V remains completely isolated from the upstream and downstream pipes and from the through hole 20, it is possible to provide a test to be conducted when the valve 10 is open, in the following way:

- the discharge device EV is opened and the cavity pressure Py detected _;

- the discharge device EV is again closed and the cavity pressure Py again detected.

It should be noted that not necessarily the cavity pressure Py drops to the ambient pressure P _am during the opening of the discharge device EV if the re- closing that follows is very sudden.

In the case that there are no leaks from the upstream 1 1 or downstream 12 sealing seats, the closure of the discharge EV must not be followed by a rise in the cavity pressure Py beyond the value reached with the opening of the discharge device EV.

This simplified method does not distinguish between leakages due to upstream 11 or downstream 12 sealing seats but has the advantage of being performed with the valve open, that is, without having to wait for a closing of the valve 14 or having to impose one that could be in contrast with the current needs of the system that requires the fluid.