Time and effort spent at the front end ensuring that installation and meter choice are correct will always save money. Flowmetering, being an empirical discipline, requires experience to stand a chance of obtaining the right meter for the application.
The savings, in many cases, of good metering far outweigh any installation and maintenance expense. For example, the measurement of nitrogen gas supplied to a manufacturer of electronic components showed that 20% more gas was being manufactured than was being billed.
The meter used for the transfer was a square edge orifice plate, designed to BS 1042. It could have been that the edges were badly manufactured, the gas could have some chemical or erosive content, or the meter could have been installed incorrectly. As it turned out the meter was installed back to front, with the bevel upstream. The cost of this error, after two years operation, ran into several hundred thousand pounds!
Flow metering is still an empirical science, some might say art. Calibration is difficult, and often performance can only be checked by reference to other proven installations, or by interpolated data.
Design requirements for meters are wide ranging, and sometimes there is a tendency to overstate specifications to be competitive, particularly with turbine meters.
Back to basics
Basic rules are that the installation should be considered at the earliest possible stage, such that all installation options have been aired before finalising a design.
It is worth costing more substantial plant modifications to improve the meter uncertainty, and to compare the potential costs to savings in terms of absolute measurement, improvement in quality, safety considerations and also customer credibility. Unfortunately, reduced capital spend often wins the day compared to long term savings.
Ease of access is important. This is not only for repair and maintenance, but also for re-calibration, where on-site calibration is a useful facility. It is worth considering supplying insertion tappings for the chosen method at the design stage.
Next, ensure that the meters chosen are appropriate for the application. An application requiring good repeatability, such as batching or control, does not always require a meter of high intrinsic accuracy. It may be best to use a rugged meter that can be set up on site to optimise performance. Features such as good response and resolution may also be more appropriate.
Auditing new installations has been a revelation with regard to lack of forethought – and the implications of cost cutting on effectiveness.
In one installation, the reference meter type suffered severely from viscosity. This added an extra uncertainty of 0.2% to the primary meter. A new reference meter would have cost $400,000. As the meter monitors product worth $20 million per day, an improvement in uncertainty of 0.01% would have saved up to $700,000 per year!Further, acceptance testing can take many forms – a calibration certificate for the meter; witnessing the meter calibration; or the checking of all parts of the system.
Experience has shown how crucial system testing is. One system was audited where the process electronics met the flowmeters for the first time on site – and would not communicate!
The time, effort and cost penalty of trying to fix such problems on site often far outway any difficulties caused by the alternative – possible late delivery.
Another point is to make sure fluid properties are not overlooked. For example, positive displacement meters will be relatively immune to upstream pipe work and valves, but will have problems with particles in the fluid.
Most meters are calibrated under conditions of fully developed flow profile – that is a symmetrical turbulent profile formed from long lengths of straight pipe. Any deviation from that profile will effect the calibration, particularly the effect of fittings.
Where possible, ensure that the meter calibration is witnessed and related to a traceable standard. Typically, lSO 5167 covers most conventional d.p. meters. BS 3680 reviews open channel meters, and the American Gas Association (AGA) and American Petroleum Institute (API) standards deal with fiscal gas and oil measurement.
To conclude, always choose the right flowmeter for the right job, and be careful about becoming a guinea pig with the latest technological revolution!
* The Author is a consultant with SGS Redwood. The feature is based on a paper at the ITC C&I Show conference in 1996.