Closing the loop
2 Feb 2004
At its valve manufacturing facility in Cernay, France, Emerson Process Management recently commissioned a 'PlantWeb Dynamic Performance Loop' test rig. One of five such loops set up by the company at various locations around the world, the Cernay operation provides a simulated plant environment in which customers can see performance tests carried out to independent test specifications.
According to Geoff Timons, sales and marketing European v-p for Emerson's Fisher Valves Division, independent audit results have shown that between 15 and 30 per cent of all control loops tested have a direct impact on plant profitability. But perhaps more significantly, in over half of the loops tested poorly designed or maintained instrumentation contributed to a degree of loop variability that had a knock-on effect on profitability.
The PlantWeb loop at Cernay has been set up to prove that link between instrumentation performance and loop variability, and to demonstrate the benefits that a digital plant architecture such as PlantWeb can bring to loop performance.
The Cernay lab includes a flow loop (water at the moment, though there are plans for a gas loop) and the dynamic performance loop. Both are powered by PlantWeb through a DeltaV automation system. They are fully equipped with Emerson products such as Fisher control valves with Fieldvue digital valve controllers, Rosemount temperature, pressure and radar level transmitters, Micro Motion Coriolis mass flow transmitters, and Rosemount analytical conductivity and pH transmitters.
The flow loop is used to determine or confirm control valve sizing coefficients (on sizes from ½ to 10in), as well as testing for dynamic torque, stem force and sound levels. The dynamic performance loop is used to perform open and closed loop tests on control valves (from ½ to 4in) and other instruments.
Open loop testing involves introducing a step change in the signal to the valve and monitoring the resulting actual changes in valve position and flow rate. Closed loop tests, on the other hand, introduce a process perturbation to the loop, with the aim of assessing the ability of the various loop components to respond to and minimise the resultant variability.
While tests are running on the performance loop, on-line calculations can be carried out in the DeltaV system to show the degree of variability (the difference between the process variable and process setpoint) that is present. Variability is influenced by factors such as tuning, calibration and maintenance. It will always exist in a process, says Emerson, but it's the function of the process control equipment to minimise it to acceptable levels.
Emerson espouses a three-phase approach to optimising control loop performance — 'establish, maintain, retrieve'. The first phase is perhaps the obvious one of establishing what is actually expected of the valve. Reference specifications, such as those from EnTech ('The control valve dynamic specification; version 3') and the ISA standard SP75.25, provide performance criteria that can be applied to the selection of a valve to ensure it can minimise process variability.
Emerson's own knowledge base, built up from years of dynamic performance tests on control valves on loops such as that at Cernay, can also aid in the selection of the correct valve design and size to match the process requirements.
Having established the desired performance of the valve, phase two is all about maintaining it in operation. What the loops at Cernay demonstrate is how digital communications technology such as that in PlantWeb can be used to monitor control valves and alert operators when their performance starts to fall off. This is done using the diagnostic capabilities of the Fisher Fieldvue valve controllers, which can deliver captured diagnostic data via AMS Valvelink predictive maintenance software to operators and maintenance personnel.
If any of the key operating parameters of the control valve change, the operators, maintenance engineers and even Emerson's own diagnostic technicians can immediately be made aware, by either the PlantWeb's own alarm/ monitoring system, email, text message or even the production of a works order.
On-line testing
On-line testing can also be carried out on a routine basis to look at parameters such as friction and deadband. This data can then be compared with the initial performance criteria and also trended with historical data, so that potential failures can be identified before they become functional failures. If required, all these tests can be run remotely, using PlantWeb's communications capabilities.
If a potential problem is flagged up then phase three of Emerson's philosophy comes into play. This is identifying the actions that need to be taken to retrieve the valve's performance to its original benchmarked level. Problems can often be solved without taking the valve out of line. Diagnostic tests can be run when the plant is shut down or the valve by-passed.
Analysis of the test data, carried out either on site by suitably trained technicians or remotely at Emerson's European diagnostic centre, will then reveal what remedial action needs to be taken. And after the maintenance has been done, further diagnostic tests can be carried out to provide new benchmark information.
Geoff Timons acknowledges that there are still many control loops being run in manual mode when they should be on automatic, simply because their automatic performance is worse than the manual conditions.
'We need to establish and maintain valve performance,' he says, 'to know how it affects valve performance and when maintenance is required.' With its investment in the Cernay performance loop test rigs, Emerson can now demonstrate how that can be done.
SIDEBAR: Asset management and optimisation
One of the impressive features of Emerson's Cernay set-up is the way in which it highlights the capabilities of the company's asset management systems such as AMS Valvelink. The latter is just one of a range of diagnostic software that has now been brought together under the AMS Suite umbrella title. The family of applications now extends well beyond valve and instrumentation monitoring, taking in mechanical equipment, electrical systems and process equipment.
Condition monitoring techniques, such as vibration analysis, oil analysis and infrared thermography, monitor assets such as motors and pumps, feeding diagnostic data back through PlantWeb to optimise predictive maintenance, performance monitoring and general asset management.
All this asset information can also now be consolidated and summarised by new AMS Asset Portal software. This presents a web-based, single view into predictive diagnostics on critical plant assets, offering a new tool to the decision-makers in charge of the plant.
As the research director of the ARC Advisory Group, Wil Chin, says, 'the AMS Suite/Asset Portal should be welcomed by manufacturers managing a wide range of assets. It's the first real web solution that centralises the information in a common database to allow management of Emerson and third party assets from a common view, which is exactly what manufacturers are demanding.'