Thermal Relief on the line

[chemprocess]

Hi all,

I have heard that for every 900 litres of liquid volume in a line, a thermal relief valve needs to be installed.Is there any standard which will indicate when to put the thermal relief valve on the line?

Please help

 

[Phil100]

Hi...it is not true to say that a thermal RV is required for any fixed volume.

Selection and sizing is based on design conditions of the containment system, anticipated temperature rise, hazardous properties of the fluid, trapped volume etc. There is no formal guide or standard that applies universally. There are alternatives to relief - expansion pots, not blocking systems in etc.

The Institute of Chemical Engineers in the UK sells a good guide to pressure relief by Parry. In this there is a useful flowchart as to when thermal relief is appropriate.

The guide is avaiiable through their website...

http://www.icheme.org

A healthy and peaceful 2004 to all!

Regards,

Phil

 

[Guidoo]

As far as I'm aware there is no such a standard. For example, API 521, 4th ed., section 3.14 does not mention a minimum amount of liquid. From this section in API 521 you may conclude that all pipes that can be "blocked in while they are filled with liquid at near-ambient temperatures and are heated by direct solar radiation" require a thermal relief valve (in most cases a 3/4 inch x 1 inch TRV will be sufficient).

An article in Chemical Engineering World of February 1998 contained a number of items that need to be considered:

- Is the pipe continuous in operation and thus not routinely isolated without being depressurised and drained? Most continuous process plant piping unit falls under this category.

- Is liquid highly toxic, corrosive or liquified gas?

- Is there a weak point (e.g. flanges) in the network where expansion could relieve without serious consequences?

As a rule of thumb, pipe containing more than 500 lts. of liquid or more than 45 m length (whichever is lower) which could be normally blocked in should be provided with TRV.

 

[aikmeam]

Try SHELL DEP. I remember something on this in their flare and venting system design section. If I'm not mistaken it's 500 litres of liquid volume and includes the considerations as Guidoo suggested.

 

[25362]

Depending on the liquid in the blocked in pipe, and the temperature in question, an approximate pressure increase per unit temperature increase, MPa/centigrade, can be obtained by dividing the cubic thermal expansion (CTE) coefficient by the isothermal compressibility (ITC), as tabulated, for example, in the CRC Handbook of Chemistry and Physics.

As an example, take methanol at 20 centigrades,

CTE, 0.00149/^oC; ITC, 0.001214/MPa, thus the developed pressure would be: 0.00149/0.001214=1.22 MPa/^oC =12.2 bar/^oC

Ethylene glycol, at the same temperature, would give:

0.000626/0.000364=1.72 MPa/^oC=17.2 bar/^o

These approximations assume the liquid-packed blocked in container -piping or equipment- is absolutely rigid.

 

[dcasto]

Instead of a relief valve, I use a 1/4" check valve around the valve that isolates the line. Th echeck valve can have a spring set to open with a 10 psi differential. If the isolated line pressure exceeds the unisolated system by 10 psi, then valve opens and allows a few ounces of fluid to exit and lower the pressure.

 

[1974vet]

If you have access, the following info should be helpful:
1. "Decide Whether to Use Thermal Relief Valves", F. Bravo and B. D. Beatty, Chemical Engineering Progress, December 1993

2. "Safer Relief Valve Sizing", W. Y. Wong, Chemical Engineering Magazine, May 1989

3. "Forestall Pipe Bursts", D. C. Cpenhaver and L. A. Coppari and S. G. Rochelle, Chemical Engineering Magazine, January 2001

4. "Save Pipes from Bursting with A Compensator", J. Boteler and D. Clucas, Chemical Engineering Magazine, December 1999

I can lay my hands on only #3. Item #3 indicates that pressure increase is a function of temperature increase, and is independent of pipe length. Assuming no leaks, only a small temperature increase is required to overpressure typical transfer lines. I can email my independent verification work of this reference if interested.

I have a related question: Is there accepted tabulated values of surface temperature of pipe (or metals) due to solar heating? This is different from the solar heat rate used for thermal expansion above.

Thanks

 

[EGT01]

1974vet,

In your reference
(1)"Decide Whether to Use Thermal Relief Valves", F. Bravo and B. D. Beatty, Chemical Engineering Progress, December 1993
the authors gave an example of maximum expected temperature based a location's typical maximum dry bulb temperature plus 15F due to sunlight. They referenced the Air-Conditioning and Refrigeration Institute, "Refrigeration and Air-Conditioning", Prentiss Hall, Englewood, NJ (1979).

I'm not familiar with the reference but maybe that would have what you are looking for.

 

[1974vet]

Thanks EGT01,
After some searching, I actually found the Refrigeration and Air-Conditioning reference! On page 280 there is Table 21B "Allowance for sun effect". The value selected in the CEP article, 15F, corresponds to the intersection of "medium colored surfaces", e.g., brick, red tile, unpainted wood, dark cement, and "Flat Roof". Looking over the table, this may be the best average value. The maximum value is 20F. These values are intended to be added to the normal temp difference between indoor and outdoor design conditions for building heat load calculations.
Thanks again. . .