7 replies to this topic

[papaya]

Hello guys I was told that the new ASME PTC 19.3 TW 2010 revision made a big change on the requirements for liquid flow, the max wake frequency ratio has been changed from 0.8 to 0.4, which is half the previous limit.

My question is due to the new limit if we can't get the insertion length we want without exceeding 0.4, what else can we do without paying too much. Making the tip & bore diameter will help with lowering the ratio but not too much. This is especially tricky when the pipe ID is big which requires longer insertion length.

[temperature lane]

You can run a quick, free check for thermowell compliance to the standard with SwiftyCalc ( www.jms-se.com ).

You note correctly that the 2010 ASME 19.3TW Thermowells standard is more conservative than its predecessor. This is due to incidents of field failure using the old standard.

One method not mentioned in your post of getting more insertion length is reducing the shielded length.

The better question is why do you want to go further into the pipe? I know that many folks want to use an insertion length of a half or a third of the pipe diameter as a rule of thumb, but once you are in far enough to minimize conduction error, what does the temperature at that one-third-of-the-way location tell you that you don't get from a shorter immersion length? Probably not much. And if the well fails or sensor fails due to vibration, even less.

One final note, the extra $ you spend designing and procuring a properly designed well is far far less than the cost of repairing and replacing a broken one.

[temperature lane]

Ps. Bulking up the root (and sometimes tip) will enable your well to go further into the process. The bore is not as significant a factor unless you bump into the standard's dimensional limits.

[papaya]

I was told that the new rmax of 0.4 or 0.8 is determined whether the thermowell & it's flange is one piece (not welded ie Vanstone), welded & machined or as welded. For Vanstone typical 0.8 can still be applied but welded will require the new 0.4 requirement which is really a big jump from 0.8.

Was recommended to use thicker straight thermowell instead of tapered. Is there any big impact on measurement using straight?

This is something new to me.

Ps. Bulking up the root (and sometimes tip) will enable your well to go further into the process. The bore is not as significant a factor unless you bump into the standard's dimensional limits.

[temperature lane]

Papaya,

That is not correct. There are extremely limited bases for exceeding the in-line limit of 40%.

The first is where you are dealing with a fluid that meets the standard's criteria for a low density gas. In this limited instance in-line resonance is suppressed.

The second (and this is where your Van Stone well may come into play) is where you well passes the cyclic stress test at in line resonance, but does not "park" within that zone. This is a very tricky option to implement as reductions in velocity or unsupported stem length can cause your well to park at in-line resonance which is strongly discouraged by the standard. This is counterintuitive as these types of reductions typically reduce the likelihood of failure. To avoid this possibility the well should be evaluated at all expected steady state conditions not just "max" conditions.

The other two bases for exceeding in-line resonance are extraordinary and are not likely a factor for your application.

By the way, the 2010 standard no longer uses natural frequency. It assesses installed natural frequency which is significantly different.

You will quickly see the difference when you run a SwiftyCalc analysis (correction to url: www.jms-se.com/swiftycalc ).

[temperature lane]

Ps. Straight is often stronger (if the root is the same as tapered) because tip diameter is a denominator in calculating wake frequency. The larger the denominator the smaller the wake frequency. This usually, but does not always, result in a longer permissible insertion length.

[papaya]

Thankyou for your kind inputs, the exemption to allow rmax 0.8 is abit unclear especially on the 2nd reason. Just want to make things clear, does the rmax 0.8 still apply to gas or for both Liquid & Gas flow are now 0.4 unless in special situation as you have described(coz i was told it applies to gas too)?

Because some vendors still claiming they are allowed 0.8 case even after i asked about the 2010 revision. With a 0.8 tolerance they hardly pass i wonder how they can make it under 0.4 when it's already quite short. Increasing the tip diameter would help but i don't think it can reduce it that much.



Just a side question can we use this standard to calculate wake frequencies for other installations that are not Thermowells (probes etc)?

[mxer911]

The first is where you are dealing with a fluid that meets the standard's criteria for a low density gas. In this limited instance in-line resonance is suppressed.

Hi Temperature Lane,

What would your criteria be on determining a low density gas? Typically in the past we have run it at any gas with a density at or below 1 kg/m^3. Any comments or suggestions would be grealty appreciated.