3 replies to this topic

[go-fish]

In order to provide sufficient NPSH to the transfer pumps in tank farm, there are generally the options to either elevate the storage tank, use a different pump type (vertical can pump, double suction) or increase the minimum liquid level of the tank. Generally, the preferred options are first two in order to minimize unused inventory in the tank. Has anyone ever experienced a situation where the minimum liquid level was increased instead of elevating the tank or using different type of pump?

Edited by go-fish, 18 May 2012 - 03:09 AM.

[kkala]

- In atmospheric tanks, atmospheric pressure is additive for calculating NPSHa, so problem of insufficient NPSHa is not anticipated. Of course suction line should be properly sized to minimize frictional ΔP.
It is a good practice that tank LLLL be higher than pump suction centerline and check suction line under this configuration (ensuring priming of centrifygal pump at any operative tank level).
- Concerning pressurized vessels, no case has been met concerning increase of vessel LLLL, evidently except the (rather small) adjustments made in detail engineering, when actual NPSHa of the pump gets precisely known. Two characteristic examples:
1. Concerning basic design of pumping from a liquid propylene sphere (1990), we elevated the sphere adequately for the pump to have enough NPSHa. See http://www.cheresources.com/invision/topic/14900-lpg-sphere-relief-scenarios - post No 20, as well as http://www.cheresources.com/invision/topic/11254-centrifugal-pump-npsh - post No 1, last paragraph.
2. Concerning installation of Byron - Jacson vertical pumps downstream an existing liquid ammonia drum (1977), we placed the pumps in an open pit to ensure NPSHa.
Of course this necessitated rainwater collection in the lowest bottom point of the pit, which had to be taken out through another pump (probably peristaltic in that plant).
Some cases might justify increase of drum LLLL, e.g. in a flare knock out drum, if its capacity after revision of LLLL is still considered adequate (possible, not met in practice). I assume these concern modifications on existing equipment.

Edited by kkala, 18 May 2012 - 01:41 PM.

[Fr3dd]

Well, you may have some problems elevating the thank if it's a really big one and it's actually worse because it takes a lot of unused inventory to gain a few inches in height. In this cases I have seen any types of ideas: since different pump types to lowering the pump's centerlines (putting them on a pit). If you ever find a mechanical engineer that accepts the latter, he will be one of a kind.

As kkala said, you may have to work in the suction line sizing in order to guarantee minimal pressure drop.

[Technical Bard]

I have seen designs where the low liquid level was increased to ensure pump NPSH, particularly in tanks of hot water - but I think it is a bad design. More effort should be put on pump selection, suction line sizing and routing and the opportunity to increase the elevation difference between the tank bottom and the pump suction. Wasted tank space is expensive both in the cost of the tank and the inventory being stored indefinitely.