Hi there,
I have already gone through various topics discussed in the forum regarding the surge analysis. Those are indeed very helpful but there are a few unanswered queries regarding the fire water network that we are analysing for an offshore platform. Our network is Titanium with low allowable stress values and we are encountering multiple failures during the static analysis that we have performed before we move onto dynamic analysis.

1. The unbalanced dynamic forces from Pipenet have been forwarded by our Safety group on pipe segment basis with pressure wave speeds of the order of 900 m/s. Their model is very simplistic taking into account the pipe lengths and in-line valves/controls etc. We have input same value of the unbalanced static forces on all the elbow pairs defined for a particular pipe segment (with DLF=2).
2. The firewater network is very extensive with 3 fire water pumps and near the pumps the layout is very congested with scores of bypasses in a very limited space. There are many directional changes in each pipe segment and some elbow pair lengths are either fitting to fitting or around 2-3 meters only. The wave will take approximately 2.5-3.5 ms or shorter to pass through these pairs. Is this not a very short interval of time to be considered as a potential problem?
3. Is there any thumb rule or guideline for the minimum elapsed time interval above which the unbalanced forces should be considered as a problem? The reason I am asking is that by inputting unbalanced forces at all the elbow pairs we may land up in changing the layout or providing additional stoppers / guides etc. that could have been avoided otherwise.
4. Based on replies to point no. 2 above, Can we purge shorter elbow pairs and consider only a few elbow pairs in dynamic analysis also? Or do we need to input all the elbow pairs irrespective of the lengths?
5. Is it normal practice to take allowable stress values (for sus + surge) as 0.9 Sy?
(Reference paragraph 302.3.6 (a) of ASME B 31.3)

I look forward to replies.

3. calculate bend-bend time L/C, and apply forces only on bends where relative L/C is more than 3.5 ms. L is spool length. C is wave speed.
4. refer 3
5. yes (occ).


regards

Or do this :
You have 2 times
1. Pressure rising time Tr (open or close time valve you dont have this time take r/c)
r- inner radius
c- speed of medium sound

2.Duration pressure time Td in the short pipe leg
L/c
L- Leg lentgh between 2 bends
C- speed of medium sound

In the case Td < Tr the force in this pipe leg cant to reach the calculated max force, because its not time enough.
So you can the max force with the quotient Td/Tr to multiply.
Force in this leg : max force * Td/Tr.
But the force is not zero.

Thanks for the replies but the basic query remains unanswered.
1.I have a doubt about the rise time, shouldn’t it be (πR/2) / C ? Where R=Bend radius, C=Speed of sound in the medium
For an 18” line with Sch 10 thk. the rise time as per (πR/2) / C is 1.23 ms (C=876.75m/s) and the equivalent pair distance is 1.077m (L/ C). Does that mean above 1.077m we have to put forces in all the pairs?
For the above conditions if we calculate the rise time as per (r/C), where r = inner radius (222.25mm) it is coming around 0.25ms which much lower than the previous value. So what’s the value to be taken for the rise time.
2.Is there a minimum Td say 10ms or so, below which the unbalanced forces are not considered a problem during surge?

Regards
Dmani

1. The rise time Tr (r/c) is a good assistance number if you havent valve close time.
r- is the inner pipe radius and not the bend radius.
2. In your case Td = 10 ms and rise time Tr= 0.25ms the condition
Td < Tr is not complied !
So the max force can to full develop in this pipe leg.
Wether this force a significant role in dynamic results give is a another question. Usually are the long pipe legs and not short pipe legs essential
in this dynamic cases, because the load to stay on longer time and the pipesystem have more time for a dynamic reaction.
Keyword : Inertia of pipe masses and impuls load time.

To be absolutely sure get the ramp time from pump manufacturer.


regards

Normally is r/c a conservative approach.