Ebb and flow
For many years, flow measurement was, technologically, the poor relation of instrumentation. Even by the mid 1980s the majority of flowmeters sold were still differential pressure flowmeters, practically developed in the early part of the century. Terry Cousins, SGS, charts the recent surge.
As computing and electronic advances drove instrumentation onwards, flow measurement tended to wallow in a self-imposed suspended animation. Even by the mid 1980s the majority of flowmeters sold were still differential pressure flow meters, practically developed in the early part of the century, and positive displacement meters again largely developed in the 1930s. In spite of some very obvious deficiencies as measuring instruments – differential pressure meters not even having a directly proportional relationship of flow to output, but a square root relation, and positive displacement meters having often complex arrangements of moving parts – these types became strongly entrenched. It has only been the last 15 years that has seen a significant change.
The reasons are indicative of the practical problems of flowmetering. Checking the proper operation of meters on site is a difficult and expensive business. In fiscal terms, the cost of mobile or site calibration systems may be justified – in loading oil tankers, for example, 0.1% measurement uncertainty represents around $20,000 – but for most industrial applications the cost is not viable.
Meters cannot be placed into an arbitrary pipework configuration with the immediate expectation of good results. Only history and much testing has been able to provide a set of guidelines, often very loose, for meter installation.
The independent checking of meters involves very capital intensive flow laboratories, charging at high rates to calibrate each meter, giving a result that still does not necessarily apply to the actual installation.
There are many sets of variables that affect the requirements for meters for different applications. These include: pressure from vacuum to hundreds of bar; temperature ranges from cryogenic to thousands of degrees centigrade; sizes from millimeters to kilometers; erosive and corrosive materials; different uncertainty requirements; and fluid types (gases, liquids, mixtures etc.). Individual solutions, rather than single generic types, tended to be the way.
A panacea for all
Probably the biggest single factor retarding the advance of metering was the almost legendary overselling and failure of some products developed in the period 1960-1980. This problem can be seen from the product development graph shown below, which sums up the progress of almost all of the meters presently in service.
Almost every current meter developed in the last third of the century has followed this classic curve. The electromagnetic flow meter, for example, when first produced, was seen as a panacea for all measurement. Its good linear flowrange with negligible pressure drop made it a very attractive purchase. However, problems such as noise sensitivity, large zero drift and the fact that the fluid had to be highly electrically conducting soon became apparent. Support costs escalated and disillusion set in. Fortunately, development work continued, a new coil driving method was found that reduced noise sensitivity and zero drift, and a more sensible and pragmatic sales pitch was introduced. Further high quality production techniques have made for a very rugged, reliable and low cost meter, which is now a great success. The doppler ultrasonic flow meter, on the other hand, has not survived the initial trauma, and at present there are few signs that it will. The vortex meter is surviving, if in a more limited form.
Where are we?
So where are we with flow meters, and do we really need them? If we consider first the requirement, then there is very little doubt that good quality meters are needed more than than ever, for the following reasons:
- The need for more efficient systems often requires more reliable, more accurate measurement for control, and flow measurement is an essential measurement;
- Control of pollution, flare emissions, and in particular outfall control is becoming subject to very stringent regulation;
- Reducing levels of raw materials and fluctuating costs require constant monitoring and low uncertainty measurement for better accountancy controls;
- Stronger regulation of sales, for example accurate dispensing of beer, wine, petrol and gas, require good quality measurement.
A good example of the changes required for all of the above reasons can be seen in the water industry. Problems of lack of basic resource (or more particularly lack of knowledge of the level of that resource) and perceived leakage problems have presented a requirement for better bulk measurement. The Environment Agency is looking closely at both the abstraction of water and levels of outfall. Privatisation has required improved efficiency, reduction in manpower and more automatic controls.
The same is happening, and will have to happen, in most industries.
These factors have, for the last decade of the 20th century, promoted a concentration of flow meter development that has fundamentally changed both the types of meter used and the method of sale and of manufacture.
As can be seen in the comparison of the market shares of flowmeters shown above, the predominance of differential pressure and positive displacement meters of 1985 is being replaced by Coriolis, electromagnetic and, increasingly, ultrasonic time-of-flight meters. An interesting commonality of all of these is that they thrive on the improvements in computer and electronic technology. Coriolis and Ultrasonic flowmeters in particular have improved because of the improved materials available.
The picture painted by these charts will, I believe, continue. The big three meters will continue to increase their market share, pushing the other meters into more marginal applications.
The effect of this has been to change the state of supply of flowmeters. Historically, the industry has been made up of a large number of small and medium sized companies, many of the small companies producing specials for particular applications. Even companies like BP and Shell used “one man bands” to solve problems – individuals would solve a flowmeter problem, usually paid directly out of departmental budgets. A large number of highly innovative solutions were obtained.
The situation is very different now, with a few small companies in existence offering the relevant expertise. It is no longer a simple matter for departments in large organisations to channel funds discretely for problem solving, and, probably more importantly, the new meters have a much wider range of application. Now we are seeing, as a consequence, the formation of several very large flowmeter companies Ð Fisher Rosemount, ABB Kent Taylor and Fisher & Porter for example, formed out of a large number of smaller companies, beginning to dominate the market. Such companies are reducing the meter choice to improve their manufacturing and selling capability, pushing the industry more and more along the route already noted.
Is this good for the flowmeter buyer? There is very little doubt that the performance and quality of flowmeters has improved considerably over the last ten years. Modern production engineering techniques combined with expensive and well run development programmes have produced very competitively priced meters with good uncertainty of measurement for a wide variety of applications. The concern is more for the future. At present, prices and service are good, but there is still very broad competition. As this diminishes, however, driven as much by the larger customers such as the water companies taking on single source agreements as by the strength of the new conglomerates, will this continue? Also, the smaller specialist companies are disappearing without any obvious successors.
Whatever the outcome, the need for flow measurement will continue to grow, having become an essential part of industrial instrumentation. The limits of uncertainty will continue to be pushed back and, hopefully, reliability will become a by-word for metering, a feature that has not been a high priority in the past.