Reactor internal coil failure with respect to API521 guide on HE tube rupture 1

[chemks2012]

Hi all,

Hope you all are doing well. I need your help as always.

I have a reactor with internal coil and wish to know the coil failure case for reactor pressure relief sizing.

I read the definition of ‘corrected hydro-test pressure as: hydrostatic test [hydrotest] pressure multiplied by the ratio of stress value at design temperature to the stress value at test temperature

Reactor shell design pressure: 6barg
Reactor shell design temperature 0 to 250degC
Reactor shell hydrotest pressure: 12barg
Reactor shell hydrotest temperature: 20degC
For reactor shell, Ratio of stress value at design temperature to the stress value at test temperature: 0.8
From above definition, Corrected hydrotest pressure will be: 12 X 0.8 = 9.6barg

Reactor internal coil design pressure: 15barg
Reactor internal coil design temperature 0 to 250degC
Reactor internal coil hydrotest pressure: 32barg
Reactor internal coil hydrotest temperature: 20degC

Reactor coil contains steam at pressure between 14 to 15barg pressure.
Reactor coil is made up of 3”NB, schedule 40 pipe

My queries,

1) As you have gathered, I have applied tube rupture of heat exchanger from API521. Could I apply that analogy for internal coil as API 521 does not say anything about vessel internal coil?
2) Am I right in saying: As the corrected hydrotest pressure of reactor shell [low pressure side of a system] is 9.6barg which is less than steam pressure of 15barg, we need to consider pressure relief device?
3) Now main query- How do I decide whether it would be a pin hole or full rupture of a coil. I am asking as I did not get clear guide from API521.

Would appreciate your help/guidance.

Regards,
KS

 

[Latexman]

Is there any history of corrosion on either side of the coil? If so, what kind? Or, is this a non-corrosive environment. Is there any history or evidence that a catastrophic failure is credible? Water hammer?

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[don1980]

Comments on your questions:
1) Yes, one can apply the API 521 logic for tube rupture for this installation.
2) Yes, you are correct in saying that tube failure is a potential source of overpressure for the reactor shell.
3) Whether you base the relief sizing on a pinhole or a complete tube break is a risk decision that is left to the user. Risk analysis is a case-by-case assessment, and it's based on the owner's risk tolerance. That's why API 521 doesn't provide one-size-fits-all guidance in how to choose between a pinhole and a complete break.

Regarding this specific case, you have very good justification for using a pinhole leak as the design basis for the PSV. That's because the tubes are actually pipe (3" Sch 40) rather than conventional gauge tubing. Designing the relief valve for a complete failure of a pipe would be extremely conservative (excessively conservative in my opinion). Complete tube failure is generaly only applied to gauge tubing and not to pipe. For example, one generally doesn't consider complete tube failure in a double-pipe ("hairpin) exchanger. Instead, a pinhole leak is typically the basis for those type exchangers.

 

[chemks2012]

Hi Latexman, don1980

Thanks very much for your input.

It's new reactor and a new coil as well. As the coil will be used for cooling water and steam heating, I believe coil may see water hammering. No corrosion issues as such from reactor content.

I seem to agree with answer to point3 by don1980 though i.e. coil is a pipe and not tube. Agree on risk assessment bit too.

I believe there is a typing mistake in API521 too [see red text below of an extract taken from page 11 of 2008 edition]. It should be 'corrected hydrotest gauge pressure' not 'uncorrected'. Am I correct?

Regards,
KS


For example, an ASTM A 515 Grade 70 carbon steel vessel with a design gauge pressure of 517 kPa (75 psi) and design temperature of 343°C (650°F) has an allowable stress of 130 MPa (18800 psi) at these design conditions. Because the hydrostatic test is often performed at a temperature less than design temperature, the hydrostatic test pressure should be specified to account for the allowable stress differences at the two temperatures by multiplying the design pressure by the ratio of stress at test temperature to the stress at design temperature. At ambient temperature, the allowable stress of ASTM A 515 Grade 70 carbon steel is 138 MPa (20000 psi). If the pressure-design code requires the hydrostatic test be performed at 130 % of the design pressure, then the hydrostatic test pressure is as follows:

In SI units:
517 × (138/130) × 1,3 = 713 kPa (gauge)

In USC units:

75 × (20/18,8) × 1,3 = 103,7 psig
The uncorrected hydrotest gauge pressure is 517 × 1,3 = 672 kPa (75 × 1,3 = 97,5 psi). In this example, reliance on administrative controls as the sole means of overpressure protection might not be appropriate if the gauge pressure caused by closure of the outlet valve exceeds 672 kPa (97,5 psi). This assumes the overpressure occurs while the vessel is at design temperature.

 

[don1980]

The statement in red is correct. Hydrotesting, of course, is done using ambient temperature water. Since the design temperature is almost always much greater than ambient, ASME Sec VIII says to apply a temperature correction factor when determining the hydrotest pressure. That is, ASME says to perform the hydrotest at (MAWP)x(1.3)x(temperature correction). That means that the hydro is actually performed at a value slightly higher than 1.3MAWP.

In the example from API 521 5th ed, the MAWP is 75 psig at 650F. The uncorrected hydrotest pressure is 75x1.3=97.5 psig. To correct the hydrotest pressure for the 650F design temperature, one needs to include the temperature correction value in the equation (20/18.8). This results in a corrected hydrotest pressure of 103.7 psig.

 

[chemks2012]

don1980

Thanks for that.

Regards,
KS