Practical Evaluation of NRV slamming
I will do a review and a correction for some misleading data coming from a top tier manufacturer UK version datasheet.
The above examples calculation is right, but the comments and interpretation is totally wrong and dangerous. the valve closure delay is estimated to be 0.042 seconds, 42 msec. only!
as per the manufacturer statement:
"It is important to note that this surge pressure requires to be added to the maximum working pressure of the system – in this case, this would give a maximum system pressure of 14 bar."
Actually it is not the only misleading graph for NRV slamming, below is another misleading one coming from a USA top tier manufacturer:
The sorting of no, mild, or severe slam is based on test benchmark for 200mm valves usually, size and cross section area do matter, which is something good from the first graph and supplier to do DN800 to DN1600mm test.
Lets do the math to find out why reverse velocity or slamming pressure can't be enough to decide valves are non slamming or not
If I have a DN800 mm NRV and I get a 4 bar slamming. this pressure is not equal to the pressure while the pipeline system is running. usually pressure affects radially as forces on pipes wall with small axial forces at bends etc.
But when I have an NRV slamming I am getting all the pressure at the first milliseconds as impact force acting as axial on valve disc. this axial force can damage the NRV disc, otherwise it will break the weakest point such as NRV maintenance hatch or connecting flanges etc. but a big impact will be transferred to the NRV foundation. The effect of slamming is same as a train a car on the rail. if the anchors of the NRV strong enough the forces will be transported to the construction, if the anchors are smaller than the impact force the valves could move towards the pump damaging the piping and the pumps.
If we want to translate this slamming pressure to a value we can feel. we shall multiply the Pressure (Pa) x Disc_cross-section_Area(m2) to get the F(N)
F(N) = 400000*PI()*(800/2000)^2 = 201,061.93 N ,,,, 201kN on each valve.
If you want to evaluate it as forces to affect the station contraction and foundations we have to multiply it by the total number of valves (duty + standby), if we have 2 duty and 1 standby arrangement of pumps so we will have 3x201= 603kN = 60.3ton*g
to give you an example from a field test. we had 5 DN400 mm valves each facing 6 bar slamming, the resulting forces were 5*600000*PI()*(400/2000)^2 = 376991.1 N = 377kN circa, the pump station building was under big shock like an earth quack, I felt that vibration from outside of the facility while I am watching the data logger.
One more thing I will mention in brief. if i have 4 barg shock wave at the NRV down stream, it means it shall be -4 barg at the upstream of the NRV which actually -1 barg with a big pocket of cavitation at the valve up stream between the pump and the NRV valve, the result of this cavitation pocket collapse shall be serious too..
https://www.youtube.com/watch?v=X9UbzcanuDk&t=106s
Conclusion: conventional NRV is too old to be used in bigger sizes than 200mm and strategic pump stations. delay in mSec can cause big issues. Depending on valve manufacturers is not the best to do. and please think twice before ordering or selecting an NRV if you don't have the knowhow.