This invention relates to a method for generating a time-dependent profile of a parameter, the parameter being associated with the operation of a device at an underwater well facility, a method of determining the health of a device at an underwater well facility and an underwater well facility comprising means for generating a time-dependent profile of a parameter.

Underwater hydrocarbon extraction facilities, for example subsea well facilities, make use of subsea well "trees" (also known as "Xmas trees") located on the seabed. Such trees are highly complex constructions, which incorporate many different components and systems. Typically, different components and systems will deteriorate or fail at different rates, i.e. they have differing operational lifespans. Once such subsea well trees are deployed, there is currently no information available as to how devices or components in systems, such as those operating in a subsea control module (SCM) of the tree for example, change their characteristics over time and under different operational conditions. Thus it is not possible to predict the operational lifespan of the devices and in particular when a device is likely to fail.

An example of a conventional subsea well system is shown in Fig. 1. The well is controlled from a surface mounted master control station (MCS) 1, which may be located on a vessel, platform or at the surface. This feeds, via an umbilical cable 2, hydraulic power (i.e. hydraulic fluid) via a hydraulic line 3, electric power via current carrying line 4, and communications via a data carrying line 5, such as an optical fibre for example, to a subsea control module 6 which is mounted on a Xmas tree 7 located above a well head. The hydraulic line 3 feeds a manifold 9 at the SCM 6. Because the umbilical cable 2 can be several kilometres long, a hydraulic accumulator 8 is also provided at the SCM 6, proximate the termination of hydraulic line 3, to enable supply of hydraulic fluid at peak demands to the manifold 9. The manifold 9 provides a mechanical mounting for, and feeds hydraulic fluid to, the many DCVs required for operation of the well control system. For clarity, only one such DCV 10 is shown, though it should be noted that a typical well has a substantial number of DCVs controlling devices. Each DCV 10, when electrically energised, operates a subsea hydraulic device, here a hydraulic valve 11 for shutting off the production fluid flow through a production pipeline 12. Electric power via line 4 is fed to a subsea electronics module (SEM) 13, housed in the SCM 6. SEM 13 performs electronic control functions for the tree within its internal electronic processing 14, and bi-directionally communicates with the MCS 1, via a communications system 15 linked to line 5. In this way, operating commands from a well operator at the MCS 1 may be actioned by the SEM processing 14. The processing 14 is operable to output electric power via line 16, to the DCV 10, and thus to operate the valve 11.

Typically, electrical and hydraulic functions are monitored by sensors which conventionally connect to the SEM 13, and the parameter data is transmitted via the communication system 15, and the line 5 of umbilical cable 2 to the MCS 1. Typical functions monitored may include: input hydraulic pressure via sensor 17, DCV input hydraulic pressure via sensor 18, DCV solenoid current via sensor 19 and DCV solenoid voltage via sensor 20. Other or additional sensors may also be employed. All of these sensors 17-20 are within the SCM 6, and each of the sensor outputs is connected to the SEM processing 14. Further sensors for measuring well operation parameters are also installed outside of the SCM 6, such as sensor 21 for monitoring the production fluid pressure and sensor 22 for monitoring the production fluid flow rate. The outputs from such sensors are also connected to the SEM processing 14. Note that for clarity none of the connections from the various sensors to the SEM processing 14 are shown on Fig. 1, but simply indicated by the input arrow 23.

As in the example described above, operating parameters are typically monitored with appropriate sensors or circuitry within a subsea well system mainly within the SCM, since this is the hub that houses most of the well operating devices such as directional control valves (DCVs), which control the majority of the well production flow control valves and chokes used by the facility. The operating parameters that are monitored in existing well systems typically include voltages, currents and hydraulic pressures. The method of monitoring employed normally provides for monitoring of the profile, with time, of these parameters. Thus, for example, the current and voltage profiles of the electrical supply applied to the coil of a DCV can often be monitored simultaneously. Such data are typically transmitted to the well surface platform via a well communication system, and there selected and viewed by a well operator. The operator can therefore observe "live" parameters. In addition, the operator could try to establish trends by looking back through previously observed or recorded parameters. In practice this is difficult and is not implemented.

A condition monitoring system has been proposed in GB 0916421.1 which assesses the data received from the wellhead and their variation over time, by comparison with a model of their expected behaviour. However, a weakness of this system is that it is only as good as the model used. If the models are based on laboratory measurements of operational parameters for a device, there is scope for errors to emerge when used in respect of the device when deployed. Furthermore, an initial model used is static, and may not adequately model operation of the device over time.

It is an aim of the present invention to overcome this problem. This aim is achieved by directly acquiring diagnostic information of a selected set of parameters that can be used to capture how the characteristics / parameters of devices change with respect to time. This data is then used to identify patterns and failure models thus permitting prediction of failures and calculation of the useful life of devices. More particularly, the profiles generated in accordance with the present invention may be used to dynamically update the models used in a condition monitoring system.

In accordance with a first aspect of the present invention there is provided a method of determining the health of a device at an underwater well facility, comprising the steps of:

1. a) generating a time-dependent profile of a parameter, the parameter being associated with the operation of a device at an underwater well facility, by
   1. i) obtaining data relating to measured values of said parameter,
   2. ii) passing said data to a storage means,
   3. iii) storing said data to produce a set of data related to the parameter at a plurality of times, and
   4. iv) generating a profile of the parameter from the data set, characterised in that the method further comprises the step of
2. b) comparing the profile with parameter data obtained at a different period of time during operation of the device at the underwater well facility,

In accordance with a second aspect of the present invention there is provided an underwater well facility comprising means for generating a time-dependent profile of a parameter, the parameter being associated with the operation of a device at the facility, the generating means comprising:

• means for obtaining data relating to measured values of said parameter;
• a data storage means;
• means for producing a set of data related to the parameter at a plurality of times; and means for generating a profile of the parameter from the data set;
• characterised in that the generating means further comprises
• means for comparing the profile with parameter data obtained at a different period of time during operation of the device at the underwater well facility, wherein
• the data storage means and data set producing means are located within a subsea electronics module, and
• the profile generating means and the means for comparing the profile with parameter data obtained at a different period of time during operation of the device at the underwater well facility is located within a subsea electronics module.

With the present invention, data may be stored over time so as to build up a library of operating parameters, thus permitting any change over time to be observed. When such parameters are related to the eventual failure of a device it is possible to model the failure mode such that the potential for failure can be forecast, thus enabling preventive maintenance to be implemented.

In the case of AC power supplies, parameters monitored may include the power factor and load voltages and currents in conjunction with the operation of DCVs over time, to identify potential power supply failure. Monitoring of hydraulic pressure profiles around DCVs in conjunction with the electrical voltage and current profiles of DCV coil supplies can also reveal, from previously recorded data of a failure mode, the likelihood of a DCV failure.

The data recorded over time may also provide valuable information to designers to optimise designs for longer equipment life.

It can therefore be seen that the present invention provides various advantages over existing systems, including primarily that the well operator may be provided with information regarding the life of the well operating devices and the potential to predict failure.

The invention will now be described with reference to the accompanying figures, in which:

• Fig. 1 schematically shows a known well system; and
• Fig. 2 schematically shows an embodiment of a well system incorporating the present invention.

Fig. 2 shows a similar well system to Fig. 1, and wherever possible like reference numerals have been retained. Here, in accordance with the present invention, the data output by the sensors received by the SEM processing 14, are also passed on to a storage / processing module 24, a new module not present in conventional systems such as that shown in Fig. 1. High speed sampling of the data is used, such the data is passed to module 24 virtually continuously and there associated with a time component. The required sampling rate must be fast enough to detect a trend so that possible failure might be detected to give useful warning of a problem. Taking the failure of the SEM power supply for example, failure may take around a second, so to detect the trend it would be necessary to check about every one tenth of a second - thus a sample rate of 10Hz. In other words, the module 24 produces time-dependent parameter data, i.e. a profile of the parameter data. Processing module 24 stores these data profiles and compares them with previously-recorded, i.e. already stored, data to reveal operating changes and trends. If a previously recorded data profile is indicative of a correctly operating device, then a trend away from that profile may indicate faults or failures of the device. In this way, the module 24 is able to provide "failure forecasting" for devices. The storage / processing module 24 communicates with the MCS 1, and thus the well operator, via communication system 15 and line 5 of umbilical cable 2. The time dependent data output by module 24 can then be used to construct reference parameters or models for use in a condition monitoring system such as described in GB0916421.1.

Initial operating parameters for devices, for example a DCV, can be obtained from initial measurements made in manufacture at production testing, with failure parameters obtained from original measurements made in the laboratory at the design stage. This initial information may be used to initialise the models used in a condition monitoring system, and which would then be updated or replaced by the profiles generated in accordance with the present invention.

In an alternative embodiment, failure forecasting is programmed into the storage / processing module 24, operable to output an alarm to the operator if required, i.e. if a failure or imminent failure is detected. In this embodiment, the module 24 is operable to compare a recently generated profile against a historically obtained profile stored within module 24.

Preferably, the data received by module 24 is measured against the temperature at the device location, i.e. it is time-correlated with the output from a temperature sensor (not shown) proximate the device. In this way, the operational parameters sent to MCS 1 can be adjusted to allow for the different operating conditions subsea to that of the laboratory, e.g. a low operating environmental temperature, or obtained in the laboratory in an environmental chamber set to the subsea conditions.

The above-described embodiments are exemplary only, and other possibilities and alternatives within the scope of the invention will be apparent to those skilled in the art.

A modification of the above-described methodology would combine a plurality of parameters associated with a particular device, such that a composite profile may be produced. For example, the output from sensors 21 and 22, i.e. which measure the pressure and flow of production fluid, may be combined, with time information, to produce a composite profile of these production fluid parameters.