Companies that source thermowells for oil, gas and petrochemicals applications will now need to consult the new, revised ASME PTC 19.3 (2010) standard, which has just undergone its first major revision in more than 35 years. This is likely to encourage engineers to seek out better, alternative, more innovative thermowell designs for process pipelines.
The old standard worked on a frequency ratio of f s < 0.8 f c/n but now this has changed to a more complex process where the cyclic stress condition of the Thermowell needs to be taken into account. If the Thermowell passes the cyclic stress then the ratio of f s < 0.8 f c/n is still applicable however if it fails then the ratio of f s < 0.4 f c/n is applicable. Also of concern to manufacturers and end users is that the standard is only applies to Thermowells with a service finish of 0.81µm (32µin.)
The new ASME PTC 19.3 standard has now grown from four pages up to more than 50 and so engineers need to be certain that they understand the changes involved. The 2010 standard addresses a number of new design factors that were not included in the original standard. These include in-line resonance, fatigue factors for oscillatory stress, effects of foundation compliance, sensor mass, stress intensification factors at the root of the thermowell, and fluid mass/density. This means the new standard should lead to a greater variety of thermowell geometries and discourages the use of velocity support collars, allowing designers to achieve faster response times than ever before in applications that call for a wake frequency calculation.
Chris Chant, business development manager at Okazaki Manufacturing Company (OMC) comments: “Today, petrochemical plants tend to use smaller diameter pipelines but with higher fluid velocities. This means that the design of the thermowell is critical. For example, the original ASME standard did not provide guidance on liquid mass, as the standard was originally developed for steam applications. However, for oil and petrochemical pipeline applications, Okazaki has always taken liquid density or mass into account when sizing thermowells. In fact, we are the only thermowell supplier who can provide customers with credible design alternatives to standard tapered, straight and stepped thermowells,” adds Chant.
Many thermowell suppliers incorporate a velocity collar on a thermowell in order to move the point of vibration or resonance. But adding a velocity collar means the thermowell needs to be manufactured to a very high tolerance (on the collar OD) and that the corresponding nozzle is similarly machined to suit. This tolerance must be an interference fit so that no resonance can occur.
If supplied and fitted correctly the collar only moves the point of resonance and does not solve the problem. While this seems to work, the extra costs incurred by the thermowell manufacturer and installation contractor are passed on to the buyer increasing the overall cost. The addition of the collar also increase the need for stocking specific spars for a single measuring point.
“Velocity collars are not always the answer,” continues Chant. “In effect, by adding a collar, you’re simply moving the problem somewhere else. What customers need is a genuine alternative, and that is why we’ve developed the VortexWell, a unique design of thermowell that incorporates a helical strake design, rather like on a car aerial or cooling tower fins.”
After extensive R&D using the latest CFD software, as well as independent evaluation, OMC was able to visualise and accurately compare the flow behaviour of the VortexWell helical strake design with a standard tapered thermowell.
In the analyses, the standard tapered thermowell showed classic shedding behaviour as expected, whereas the VortexWell demonstrated no signs of regular flow behaviour. The VortexWell helical strake design disturbed the flow sufficiently to interrupt the regular formation of vortices. Whilst a small vortex was observed in the wake of the VortexWell this was a localised stagnation point and didn’t shed.
However, the most significant comparison made was with regard to the pressure fields. For the standard tapered well design, an oscillating pressure field was observed around the structure. The VortexWell displayed a constant and stable pressure field, presenting no dynamic variations. As this pressure is the source of vortex-induced vibrations, it can be assumed that the VortexWell would experience a significant improvement in practise compared to the standard thermowell design.
In further tests, this time using FEA, OMC found that the ASME calculations used by thermowell manufacturers could be placing significant limitations on the safety of petrochemical applications. Using the ASME calculations gave the lowest natural frequency of vibration for the standard tapered thermowell to be 68.5Hz. However, OMC’s own FEA results showed a corresponding value of 90.3Hz, a difference of more than 30%. This highlights that the ASME calculations design rules include assumptions that can lead to considerable inaccuracies when designing thermowells for petrochemical applications. The risk of a thermowell failing due to under-engineering, or the extra costs incurred by the end user because of an over-engineered thermowell, can both be avoided if the buyer works with a reputable, experienced thermowell supplier.
Chant concludes: “Okazaki has been providing technical support and design services to the oil, gas and petrochemicals industry for many years now, working to Japanese standards, which are often more stringent than the ASME equivalent. Therefore, for Okazaki, the new standard simply reinforces that what we’ve been providing our customers with for many years is the best approach.”
