Industry set for drift to smart metering

Flow measurement technology has advanced to a point where it is within an order of magnitude of national standards with respect to measurement uncertainty.

The next key milestone will not be improving on this figure, but instead making the equipment more cost-effective for the end user.

A gradual change is one that causes small cumulative changes in performance which may or may not be significant

Alongside improved data acquisition techniques, the digital signal processing (DSP) abilities have been greatly enhanced.

This has allowed the detailed monitoring of all the recorded data to be used as diagnostic tools to identify any problems within the metering system and to complete a ‘health-check’ of the meter in operation.

If diagnostics exceed accepted levels, then the system alerts an operator. These alarms can be time-dependent, which means any erroneous measurement will not be recorded as a fault until the software has confidence that the problem is real and not due to one instantaneous fault or error with the system.

Additionally, trending of the data over time can then be used to provide regulators and auditors with information on the present state of meters, with the aim of reducing the need for recalibration.

Once the meter is installed for use in a process stream, comparing the fingerprint with calibration values can ensure no change or shift from the calibration, providing confidence that the calibration is successfully transferred to the operating location and conditions.

Examples exist in industry where this evidence has been used to extend recalibration intervals for ultrasonic meters. This will lead to a condition-based monitoring recalibration timescale rather than a calendar based one.

In reality, meters shift in performance regularly and can either exhibit an instantaneous change or a gradual one.

A gradual change is one that causes small cumulative changes in performance which may or may not be significant.

An example of this type of shift could be caused by surface roughening of the internal pipe diameter. In this instance the shift may be small at first but over a period of months or years it could become significant.

Diagnostics can pick up these small changes and it has been shown in previous test work at NEL that in some instances the diagnostics are more sensitive than the meter performance itself to some sources of shift.

The next stage in making diagnostics more user-friendly and cost-effective is by combining the data with embedded expert technical knowledge. This will result in taking the somewhat daunting data produced from some meters and turning it into usable information.

Some meters generate thousands of diagnostic parameters which can be overwhelming if they are not fully understood.

Instead, the data can be processed further to tell the user exactly what to do; suggesting causes and solutions to the issue as opposed to a basic alarm.

This information is much more valuable as rarely will users have access to this level of detailed analysis on-site. This sort of information is available for some technologies and manufacturers but not all.

However, the diagnostic information can also potentially be applied to multiphase flows. Recent work carried out at NEL, supported by the UK Government’s National Measurement System tested the diagnostics capabilities of several commercial ultrasonic flow meters in two-phase flow.

The results show proportional correlations between diagnostic parameters and volume of gas present in the flow.

If confidence can be given to diagnostics and predictive models then potentially diagnostics can be used to provide evidence to the uncertainty of flow measurement.

It would also negate the need for re-calibrations, as any shift in measurement performance would be monitored and corrected for. This situation can only benefit the end user, as it will result in more measurement confidence and less effort in maintaining the system.