Sensible heat in Fire relief calculations

[Sawsan311]

Dear All,

For PSV sizing for the vapor fire relief case using the API 521 approach, I understand that we need to estimate the latent heat for multi-flashing cuts such that we can mimic the dynamic increase in relieving temperature during fire. However, as long as the liquid under fire in the sub critical region (less than critical pressure and below dense phase), do you agree that there is no role of sensible heat in estimating the vapor relief load?
ASPEN HYSYS V.10 PSV sizing tool allows the user to include or exclude sensible heat while estimating the fire relief load for fire case PSV,, I believe sensible heat shall be excluded in case the state is subcritical. What do you think?

Additionally, if the state of fluid becomes supercritical where the relieving pressure is more than the critical pressure, then one can say that sensible heat is the dominant term and there is no role of the latent heat since the liquid doesn't boil YET it expends by the increase in temperature causing the specific volume to increase all driven by the sensible heat added from fire. Sizing the PSV for this scenario can be using the numerical isentropic flashing method described in API 521.

Appreciate your views..

Regards,

 

[apetri]

I have not Aspen, I use instead Prode but I suppose the methods are equivalent, in dynamic simulation (with Aspen, Prode, Vmg , Simsci or another software) at each step you solve mass & heat balance including as input the heat from fire and (as output) the discharging flow, this procedure allows to solve all cases (with vapor & liquid), for piping and PSV there are several methods (to include vapor, liquid and two phases flow), many models (see as example HEM in Prode) can solve both two-phases and single phase in the subcritical and supercritical conditions you mentioned, see API for the details of numerical solutions...

 

[don1980]

I agree that sensible heating can be ignored. It has a small effect on the relief volume - smaller than the safety factor already incorporated into designs - thus it can be ignored.

However, I don't agree that one should apply the formula that API 521 publishes for vessel filled with vapor or supercritical fluid. That's because there's no technical justification for claiming that the PSV will protect such vessels from fire exposure. If a vessel doesn't contain a liquid (one which will boil at the relieving pressure), then the PSV doesn't provide any meaningful protection, regardless of its size. Note that the text this section of API 521 cautions the user that these vessels (vessel with no boiling liquid) will rapidly heat up and quickly approach the yield point. Yet, (inexplicably) API 521 continues to publish a formula for those that don't understand, or choose to ignore, this fact.

 

[Sawsan311]

Hi Don, thanks your reply I totally agree with you, the overtemperature risks associated with the fast heat pick up of unwetted vessels filled with gas due to their low thermal mass inertia makes the PSV unreliable for protecting against rupture due to the increase in overtemperature stress while the pressure inside the vessel can still remain much less than the PSV set pressure. Here is when the role of fireproofng using passive fire protection and emergency depressurization, equipment layout comes into picture.
However, considering the overpressure means in terms of relief devices is a code requirement, designers still tend to consider a PSV for every pressure vessel. Yet, ASME SEC VIII Append-X M also discusses the above mentioned mitigation measures for the fire scenarios.

 

[don1980]

Sawsan, ASME Sec VIII and API 521 are clear in stating that it's the user's responsibility to determine the sizing basis for relief devices. That's as it should be, because there's no way to write prescriptive requirements that apply in every case. Note that there's nothing in the main body of ASME Sec VIII or Appendix M that contradicts what I stated above. In fact, it points out this same risk (see M-13). Nor does API 521 state that users must use that formula for gas-filled vessels. API 521 simply provides a formula (one that has no technical validity) that users can choose to use or ignore. A relief device is required by ASME Sec VIII (unless UG140 is applied), and I'm not suggesting that it be omitted. I'm just saying that one shouldn't size that PSV for fire exposure if there's no boiling liquid in the vessel. Doing so is actively harmful because it misleads the owner into thinking that the PSV is providing some measure of protection from fire exposure when it's really not. And it deflects attention from the need to consider other protective measures - measures like those you mentioned which are effective at reducing the risk.

 

[apetri]

take care that in some cases (for example for wellhead separators with design pressure about or above critical pressure of mixture) sensible heat must be taken in accout when solving the mass & heat balance during dynamic simulation (see my previous post), anyway this is not a problem for the end user (these procedures can solve all cases with liquid and vapor),
said that, I totally agree with Don comments about the application of the formula that API 521 publishes for vessel filled with vapor....