The pursuit of fairness
A central monitoring scheme for flow measurement is necessary and desirable, says Keith Flint of Pulsar Process Measurement. But, he argues, MCerts is not robust enough
As a principle, MCerts is a fine one. A central monitoring certification scheme provides the framework that water companies and other businesses need to comply with their discharge permissions and has to be welcomed.
If everyone knows where they are, they can comply with the requirements, and customers can have clear expectations of product performance, when that product has been purchased with an approval from the scheme. For effluent, the criterion laid down by the Environment Agency (EA) is that any site producing more than 50m3 of effluent has to monitor that effluent to within ±8% with a 95% confidence.
It would be fair to assume, therefore, that MCerts would almost by definition approve equipment capable of measuring to that standard. However, the equipment-testing regime in place (MCerts product conformity certificate) at present fails to prove that accuracy has been achieved. It is Pulsar’s contention that the powers that be know this and have failed to satisfactorily reinforce or review the standards and procedures, despite open criticism from much of the industry.
In addition, the sheer cost of having equipment tested to this flawed standard is prohibitive, especially when one considers that there is no potential outside the UK, as a unit tested in the UK does not become any more marketable in any of the other EU states.
In order to understand the issues in achieving the required accuracy within the outflow, it is necessary to appreciate the situation within the system and the sources of uncertainty within the overall measurement. About 95% of all wastewater treatment works’ outflows are open channel, of these perhaps three-quarters using flumes as primary measuring devices (PMDs), the other quarter being V-notch weirs.
There are three primary areas where inaccuracies can develop within the flow measurement – the flume or PMD structure itself, the measurement instrument itself, and then the capture and logging of the measured signal through the telemetry system.
Therefore the first complaint against the testing regime is that it tests the instrument under laboratory conditions rather than properly testing the system in the real world.
Historically, the greatest inaccuracies in open channel flow measurement have probably derived from poor flume construction and geometry. There has been more than one incident where the flume has been installed backwards.
More typically, the walls have been poorly constructed and are not parallel. Where walls are rough or damaged, there is both a restriction in flow due to the roughness of the material and a good surface for a bio-film to develop.
Any restriction in flow may also cause turbulence, which makes the flow dynamics less predictable and can introduce further errors. With modern materials and a greater understanding of the need for correct installation, these problems have diminished.
However, construction and installation are both very important, as is maintenance – a V-notch weir can drift by more than 4% in three months just through bio-film development.
There is a 67-page document on the EA website that details the test procedures for flow meters, so it is easy to see how the various factors that affect ultrasonic performance are evaluated. Pulsar is particularly concerned about the section that deals with the effect of temperature.
Temperature, both from the seasonal and diurnal variations (and the good old British weather), but more immediately from solar radiation (the sun shining on the transducer) has a marked effect on the accuracy of ultrasonic measurement. Temperature changes the density of the air, and it is the density, not directly the temperature, that provides the variation in the speed of sound.
For any given temperature, there will be variations in the speed of sound at different altitudes, for example. In our situation, the issue is that the sun will rarely heat the column of air between the face of the transducer and the liquid surface evenly.
Within the dark depths of the flume, the water, which may be very cold, will barely warm the air directly above the surface, while the transducer itself may be bathed in a balmy autumn glow. A temperature compensation unit mounted in the transducer head or as a separate sensor, will compensate for the temperature at that individual point, not for the stratified, varying temperatures through the column of air.
It is not uncommon to accurately sense air temperature at 30ºC,
1m above the liquid, and the air above the liquid is actually at 14ºC.
The single key point is that this solar radiation inaccuracy is, by itself, enough to create a variation in the measurement of more than the 8% limit, guaranteeing that the site is not measured within the terms of the EA policy.
Instrument manufacturers, with differing levels of success, have addressed the temperature issue in a variety of ways, but it is such a critical variable in accurate measurement that it is beyond belief that experts have devised a test that virtually ignores this important and dynamic effect. Within the EA document, the test consists of putting the equipment into a chamber and warming it up with a tungsten filament, which presumably looks a bit like the sun.
Even the most basic temperature compensation circuit will be able to follow a gradual variation in temperature from an external heat source, giving a completely false impression of the accuracy of the device, particularly when the test also allows significant time (ten hours) for the equipment to stabilise. Despite this clearly demonstrable failing in the process, the EA has continued to insist that the MCerts product conformity protocol is followed and there is a link to the Sira website from the EA’s site with a list of approved equipment.
It will not be a surprise to learn that the equipment listed here comes primarily from the large multinational suppliers. For the privilege of taking part in what we believe to be a meaningless exercise, we, as manufacturers, are being asked to pay a minimum of £5,500. Just
to reiterate, this is for approval of a single unit, for supply to the UK market only.
That cost does not include the cost of the test itself, a further four-figure sum, plus all the hidden costs of staff time, administrative effort and logistics. For a specialist manufacturer of flow measurement equipment the cumulative costs of entering several devices could be huge.
These costs are undoubtedly a handicap to the small and medium-sized manufacturers, especially when taken together with the view of ourselves and others that the flaws in the testing system render the exercise pointless. There are a number of manufacturers or suppliers who are equally frustrated by this technical and financial handicap.
Additionally, it is being seen as a ring-fencing the UK open channel flow market by having this MCerts equipment approval here alone. The large, multinational companies have these sorts of sums as small change, and some have clearly taken the view that they can gain an easy advantage over the rest of the industry.
Fair play to them of course, and in the same situation we may well have come to the same conclusion, but the fact remains that they have gone through the approvals process with perfectly respectable but standard equipment. The technical bar is, in effect, set too low.
The key industry complaint is that the product conformity standards have been constructed without proper initial consultations with either the equipment manufacturers or the manufacturers of the PMDs, the flumes and weirs. Since the introduction of the testing regime, the complaints from both industries have been loud and voluble. Yet the EA seems determined to sweep the problems under the carpet and carry on as if the standard was perfect.
In essence, the current MCerts product conformity certificate fails to offer manufacturers end users the confidence of measurement expected in real dynamic conditions where the products are to be used.
No doubt some readers will be saying, with a cynical smile: “What you’re really saying is that your kit can’t comply, so you’re just throwing stones.” This is not the case, Pulsar Process Measurement manufactures DUET, the most accurate ultrasonic open channel flow measurement device in the world, and one which directly compensates for air density, not simply temperature. Pulsar’s position is that it is not prepared to be held to ransom to join a process that is rewarding mediocrity, working against British business and British innovation, and is both substandard and unfair.