Steam Explosion Potential 1

[TexasPE]

I have been reviewing API RP 521 with respect to steam generation on depressurization in a hot crude oil separation system operating at above the normal boiling point of water. I am reviewing this because of a concern by one of my client’s safety engineers in which he references Section 5.9.1 of the Fifth Edition. His concern is that this section is stating that vapor generation is too quick to handle with normal relief. My interpretation of “hot oil” in Section 5.9 of API RP 521 is that this risk is normally associated with the mixing water with a hot oil heating media and is normally handled by operating the heat media at a higher pressure than the water being heated. In essence the risk of a steam explosion is due to the lack of vapor space in the heating media system and the remote nature of the expansion tank. My interpretation of the risk of steam generation in a crude oil separation vessel is one of 1) is there sufficient vapor space in the vessel to handle the increase in volume of the rapid transition of pressurized hot water to steam and 2) if the water is emulsified can the system handle the foam generation that can result?

API 521 Reference Section 5.9.1 Water into hot oil
Although the entrance of water into hot oil remains a source of potential overpressure, no generally recognized
method for calculating the relieving requirements is available. In a limited sense, if the quantity of water present
and the heat available in the process stream are known, the size of the pressure-relief device can be calculated
like that of a steam valve. Unfortunately, the quantity of water is almost never known, even within broad limits.
Also, since the expansion in volume from liquid to vapour is so great (approximately 1:1 400 at atmospheric
pressure) and the speed of vapour generation is essentially instantaneous, it is questionable whether the
pressure-relief device could open fast enough to be of value. Normally, a pressure-relieving device is not provided
for this contingency. Proper design and operation of the process system are essential in attempts to eliminate this
possibility. The following are some precautions that can be taken:
a) designing the water side to be at a lower operating pressure than the hot oil side;
b) maintaining minimum circulation of hot oil through equipment on stand-by in order to minimize collection of
water;
c) avoiding water-collecting pockets;
d) installing proper steam condensate traps;
e) installing heat tracing to eliminate condensation;
f) installing double-block and bleed valves on water connections to hot process lines;
g) installing interlocks to trip sources of heat in the event of water-contaminated feedstock.

Can any of you provide me with some clarification as to the scenario that the API RP 521 Committee envisioned when added this language to API 521? Was this intended to state that relief valves could not react fast enough to handle water vaporization if a hot crude system depressurized and caused the water in the system to flash? My calculations tend to support the concept that the water will not flash all at once csince there is limited heat available to supply latent heat for vaporization and auto refrigeration should limit vapor generation. I strongly susspect that this is the limited exception noted, e.g. “In a limited sense, if the quantity of water present and the heat available in the process stream are known, the size of the pressure-relief device can be calculated like that of a steam valve.”

Any feedback would be appreciated.

 

[psafety]

Can condensate from a failed trap or blow line allow a quantity of water to collect in the vessel?

We call it a hot stock eruption. I was next to a vessel that had X amount of condensate in it when an emergency shutdown diverted 390-deg oil to the vessel at app 700-gpm. This a food process and the vessel had a nitrogen atmosphere, thus a 8-inch PVVV (Groth 7613). The eruption almost ripped the PVVV from its hinge. Frying oil was dripping from a 40-foot ceiling for at least a 30-radius.

Don't know the "strength" of the vessel, but the flashed steam should not have more heat than the oil that flashed it, its just going to be an instantaneous rise. You'll never quantify the amount of water.

 

[TexasPE]

In our case we have water and crude oil at equilibrium conditions in the vessel at above the normal boiling point for water. Only by dropping the pressure can water flash to vapor. But if the pressure is dropped then vaporization will occur in the hydrocarbon phase to cool the water.

I believe in your case the issue was with sufficiently hot oil that enough energy can be transferred in very short order to fully vaporize the water and generate the massive volume increase. Sort of like adding water to a strong acid or a concentrated sodium hydroxide. Add the reagent to the water not the other way around or boom.

If I separate the phases and depressurize them to form vapor and then try to cross exchange them to heat them back up to raise the pressure I either violate the 1st law of thermodynamics by not conserving the energy or 2nd law of Thermodynamics with an decrease in entropy for the vapor.