7 replies to this topic

[rd20]

Good day! Our propylene storage spheres are equipped with water flooding connections that are to be used to fill the spheres with raw water during emergency situations such as leaks in the flanges / nozzles located at the bottom part of the sphere. This is as per the design practices and technical specifications that are followed by our refinery. Generally, the faster the filling time, the better for us. However, it is not stated in the manuals the required / suggested maximum allowable filling time for safety purposes.

We need this info to check the correctness of specs of the currently installed booster pumps for the flooding of the spheres.

In other refineries with same pressurized storage, what is your practice about this?

Hoping to receive feedback fron you soon.

Thanks.

[djack77494]

You're in an area I'm not familiar with, but your query has generated some concerns in my mind. First I'd be worried about admitting raw water into any liquified gas storage vessel. LNG especially scares me. I must assume your storage is pressurized, but if the pressure should drop, all the water turns into ice. To a lesser extent, this is a possibility with other liquified gases as well. Also, these gases, even when liquified, have a relatively low density. Are your tank supports adequate for holding the vessel filled with water? I've seen photos of spherical tanks sitting on the ground (crushed) because this concern was not addressed. The ingress of pressurized water must be examined from the point of view of possibly overpressurizing your storage vessel as well. Please be more specific and indicate what hazard the water is intended to mitigate and how. I confess that this discussion is more likely to lead to my own education rather than me pointing you in the direction of a solution, but who knows?
Thanks,
Doug

[rd20]

When there are leaks in the bottom flanges and nozzles of the sphere, it is flooded with water, hydrocarbon will then float (due to difference in SG) and water will come out of the leak. flooding is continued while the problem is being addressed (i.e. bolt tightening, end cap replacing, etc). after fixing, water is then drained from the water draw-off connection until we are sure that there is no more water coming out.

With regards to sphere foundation, we usually design it on the assumption that the sphere is full of pressurized water (for hydrotesting purposes).

[Art Montemayor]

Rd20:

Pumping in water into a hydrocarbon tank to substitute water with the hydrocarbon as the leaking fluid and reduce the hazard is an old trick. I believe it was developed here in East Texas at the turn of the last century. The problem I have visualizing what you are trying to do (as Doug also has, I believe), is that you employed the term LNG and Propylene in your Post's title.

By definition LNG exists at -260 ^oF and there is no way on God's Green Earth that you are going to force water into such an environment. Additionally, Propylene is also frequently stored at semi-cryogenic temperatures and presents the same problem. It is LPG that is normally stored at ambient temperatures.

Are you now saying that your query only applies to Propylene and it is stored at ambient temperatures? If you don't give us ALL of the basic data you create confusion in attempts to interpret what you mean to write out - but leave out. Please be specific and detailed in your description.

[rd20]

I am sorry if I generalized all the pressurized storage spheres in my title. Actually, we have this setup in our LPG spheres also. Right now, we are also trying to do the same for our propylene spheres.

Yes you are correct. We are storing liquid propylene at ambient temperature (35ºC) and ~1500 kpag pressure.

In order to push water into it, we will be installing raw water booster pumps. However, I cannot find any document stating the required maximum flooding time for these spheres. Or, are there any other rules of thumb regarding this that I may use in evaluating the needed pumps?

[rd20]

any updates on this? thanks.

[djack77494]

any updates on this? thanks.

This is all very interesting and quite outside the area of my experience. You do seem to have addressed the concerns of my initial posting. Regarding your main question, however, I would think that the maximum water volumetric inflow must not exceed the capacity of your relief valve(s) to allow for the egress of displaced vapors. The PSV must be designed for this case, including having been deemed appropriate for liquid relief. Even with these precautions, I am somewhat nervous about this practice, especially as the possibility of ice formation is approached. I'd suggest that procedures be in place for lighter gases to ensure that the practice is limited in duration. You sure don't want ice forming in the only outlet for vapor/liquified gases.
Doug

[Zubair Exclaim]

API 2510 (A.1.7) says this flooding should extend outside the spill containment system ,plus the issues of water freezing should be considered if the water connection if made in the product line and not at the top of the sphere .Plus one should have adequate check valve and isolation on that supply line and spill system should be capable of handling this flow.

As far as the rate is concerned ,no practice actually speaks on the isse .I think you can have two approaches :

1- Consider bottom outlet lines including drain ,should all be considered leaked ,take half the size of the lines and see ho wmuch flow can gush out at pressure differential of (LPG storage + static head )VS (atmospheric pressure ).And use it as the flow rate of water .(there can be other possibilities like considering one flange leak at a time etc etc taken for design ).

2-Other could be to take in to account clause 11.1.2 which says PSV should discharge at a rate of 3% of vappour at a ful tank capacity .So this same capacity should be the flow rate .

3-Or just like clause 10.3.1.8 ,which says water for cooling should be activated within 60sec and fire water within 10 min .Use eithr of these times for the flow rate to fill height of the sphere from bottom, till there are connections or welding with possible leakage .

Well all these are my opinions maybe all are wrong or maybe right .But still its a intresting case ,and would love to know the outcome .

Also what kind of protection do you have at the bottom drain for preventing icing problem ?