# Liquid Pipeline Surge

Pipeline Surge has been a hotly debated topic on "Cheresources". During one of these discussions I had said that I would try to compile a blog entry related to pipeline surge possibly with an excel workbook. Today's blog entry discusses pipeline surge sans the excel workbook. I expect the young process engineers to develop their own spreadsheet from this blog entry. What I am going to discuss is the fundamentals related to pipeline surge which also form part of the excel workbook I created for pipeline surge, as explanatory notes on the phenomena of surge. Let us begin:

Pressure Surges:
- Pressure surges in a pipeline are created by a change in momentum of the moving stream, e.g. by closing a valve, the origin of the pressure surge being at the point where the momentum of flow is changed.

- Because of the low density of gases compared to liquids, pressure surges are not of concern in gas lines.

- The theoretical maximum pressure surge that can be created in a pipeline would be caused by an instantaneous total blockage of the flow and would occur at the point of flow retardation, e.g. the valve.

- The maximum surge pressure is the sum of two components:
(i) The instantaneous pressure increase at the moment of total flow blockage
(ii) The subsequent gradual pressure rise due to the 'line packing' effect.

The magnitude of the instantaneous surge can be calculated using Joukowsky's equation:

Ps = ρ*a*Δv

where:

Ps = surge pressure, Pa
ρ = liquid density, kg/m3
a = pressure wave velocity, m/s
Δv = velocity change, m/s

The wave velocity "a" is given by:

a = square root(1 / ((1/K + (d/tw)*(1/E))*ρ))

where:

K = Liquid Bulk Modulus (refer attachment); water bulk modulus is generally considered as 2.2*109 Pa
d = pipe internal diameter, m
tw = pipe wall thickness, m
E = Young's modulus of pipe material, Pa (for commercial steel it is considered as 210*109 Pa)

Methods of Reducing Surge Pressure:
The primary method of preventing the generation of unacceptably high surge pressures should be the implementation and strict adherence to well formulated and clearly written operating procedures. Additional measures which may be employed to reduce surge pressures are as follows:

1. Slow Valve Closure:
By closing a valve over a sufficiently long period the surge generated may be significantly reduced. This also allows more time to trip the pumps and hence reduce the maximum pressure. This can be implemented either by slowing down the valve actuator or by installing a two speed actuator which reduces the valve closure speed over the (critical) last 10-20% of the valve's travel.

2. Installing a Pressure Relief System:
If the creation of an unacceptable pressure surge cannot be avoided using option 1, a pressure relief system can be installed as near to the point of surge origin as practically possible. The system would vent a quantity of product from the pipeline once a pre-set pressure limit is exceeded thereby limiting the final surge pressure. This can be implemented using bursting discs or rapid response relief valves.

3. Installing a Pump Trip:
If the advent of a potentially dangerous pressure surge is detected early enough, the tripping of the upstream pumps will generate a negative pressure wave which propagates from the pumps to the origin of the surge and can counter the positive pressure surge. The effectiveness of this form of surge protection depends on factors such as the pipeline length, amount of line packing, etc.

Readers of my blog, hope you enjoy this blog entry and I look forward to your comments.

Regards,
Ankur.

#### Attached Thumbnails

Qalander (Chem)
Dear Nice info
Thanks.
robjul
Indeed, a very informative info, simple and quite easy to understand.
Dayababu
Thanks for a good information.
Mark-TR
Hi,

I just would like to make a warning regarding the pump trip option; this option shall be carefully studied. As one of the main causes of surge in long line is pump trip.

The surge due to a pump trip happens because of check valve slam. Therefore, IMHO opinion this should be avoided unless it is carefully studied. Please also note that pump trip can also damage the pipeline because of the vacuum produced due to the stoppage of pump. Carbon steel lines of 12" or more do not like vacuum, so the increase of pipe schedule will be necessary under this scenario. Plastic lines may not like vacuum at all.

In water system a typical protection can be done considering a pressure/vaccum breaker.

For other systems gas or liquid accumulator’s maybe a better option. If the system is big enough ,a surge tank may be considered.

The protection may be also as simpler as changing pipe diameter, this will reduce the change of velocity and consequently the surge peak will decrease.

I also would like to highlight that to determine the maximum surge point location is very complex and shall be done with the adequate dynamic software such as AFT Impulse.

Kind regards
amit 1
Very informative and simple to understand. Thanks
sidibouziane
Hi,

my question is related to the installtion of pressure relief system. it sounds to be more technically feasible. to be sized, we need to know the amount of inventory to be released and size accordingly the container. comparing this to the frequency and risk of high opressure surge to teh additional cost of the relief system needs to be investigated.

if some can provide an answer how to calc the inventory relieved, would perfect.

regards
Thanks for the information,
Rubens
Relief valve on a centrifugal pump discharge? Even for a flat type rotor (I mean where the shut off is not much higher than operating pressure)? Maybe this phenoimena would be more important for very large flows and diameters, but for "normal" flows (~ up to 150 m3/h) I think it is not important. Am I correct?
zatish
Ankur,
Thank you for initaiting the interesteing topic. Would you please post some details (I mean mathematical equation) where it is shown how liquid pipeline surge is dependent on valve closing time. I requested this info from vendors supplying dynamic simulation package to calculate/determine surge. They kept quiet.

For sidibouziane, I have the calculation details how to calculate surge volume.
Can anyone advise me how to upload/attach a file in this section?
Regards
ZK
ankur2061
Zatish,

Incremental increase in pressure with incremental closing of a valve cannot be provided as a simple mathematical equation. This would lnvolve complex equations and thousands of iterations over the incremental valve closure to predict the pressure profile as the valve closes. In a nutshell, the whole calculation for valve closure as a function of time and the corresponding pressure rise can be classified as transient or dynamic modeling of the surge process and better be done using a pipeline transient simulation software such as TLNET / TGNET or OLGA.

You can put in a separate post wher you can upload the surge volume calcualtion file.

Regards,
Ankur.
bahjat77
Thanks for the post. However, I disagree with what been mentioned for gases “Because of the low density of gases compared to liquids, pressure surges are not of concern in gas lines.” From a personal experience in one of the facilities I worked in, we had a significant damage to one of the gas pipelines when it was depressurized and then the isolation valve (SDV) opened quickly allowing a huge flow of gas from a 32 barg system.
ankur2061
bahjat77,

You are confusing two different scenarios.

The scenario you are mentioning is where an empty pipeline (atmospheric pressure) was filled up by opening the SDV and which caused a sudden large differential pressure leading to damage. You never pressurize an empty pipeline this way. Normally a by-pass (1" or 2" size) is provided to the SDV which is opened first for pressure equalization and once the pressure is equalized you open the SDV valve.

The surge scenario is where a flowing and operating gas pipeline having a block valve is closed abruptly. The pressure wave generated in the direction opposite to the flow in such a scenario is much lesser in magnitude then what it would have been for a liquid and the probability that the pressure surge exceeds the defined maximum allowable incident pressure (MAIP) for the line is very low.

Pressure surges have been studied in gas lines also, but in almost all cases cases the pressure was never found to exceed the MAIP. It is on this basis, the statement was made that pressure surges are not of concern in gas lines.

Regards,
Ankur.
Thanks Ankur for the insight.
zatish

Zatish,

Incremental increase in pressure with incremental closing of a valve cannot be provided as a simple mathematical equation. This would lnvolve complex equations and thousands of iterations over the incremental valve closure to predict the pressure profile as the valve closes. In a nutshell, the whole calculation for valve closure as a function of time and the corresponding pressure rise can be classified as transient or dynamic modeling of the surge process and better be done using a pipeline transient simulation software such as TLNET / TGNET or OLGA.

You can put in a separate post wher you can upload the surge volume calcualtion file.

Regards,
Ankur.

Dear Ankur,
But all simulation programs starts with mathematical model and all model starts with mathematical equations.
Base line of simulation is still the same.
Number of unknown= Number of equations.
I could not upload file for another member.
If you can help with steps please.........
Thanks and Regards
csample

Hello Ankur,

I have a case wherein we have static pressure of 15 ksi applied, valves open quickly to return the fluid (incompressible) to the reservoir. The fear is the return piping run is not sufficient to dissipate this pressure without overwhelming the piping, 2400 psi capacity.

Are the equations presented applicable for this situation or what is a better solution?

Kind regards,

ankur2061

Hello Ankur,

I have a case wherein we have static pressure of 15 ksi applied, valves open quickly to return the fluid (incompressible) to the reservoir. The fear is the return piping run is not sufficient to dissipate this pressure without overwhelming the piping, 2400 psi capacity.

Are the equations presented applicable for this situation or what is a better solution?

Kind regards,

csample,

Without looking at a sketch of your system i.e your system configuration I am afraid I am not in a position to make any comments. Please do that to get a more informed response.

Regards,

Ankur.

csample

Ankur,

I would like to share schematic details of my particular situation but unfortunately cannot.

However, if you could kindly consider your example by reversing the situation by having pressure being held, and then releasing it suddenly by opening a valve. You would still have a pressure transient that could cause damage downstream.

I guess I am asking about the applicability of Joukowsky's equation in this case.

ankur2061

csample,

Yes you could create a pressure surge downstream by instantaneous opening of a valve. But the bigger problem would be of damaging the valve seat itself due to high differential pressure caused by opening of the inline valve.

There is a simple way of avoiding the pressure surge and the probable valve damage due to high differential pressure and that is to provide a small bypass to the inline valve which will equalize the pressure slowly upstream and downstream of the inline valve when opened thereby preventing any downstream pressure surge as well as high differential pressure.

Regards,

Ankur

aniljani
mech001@gmail.com

Hi Ankur , thanks for the posts , We recenly came across a problem where the client reported a  damaged Angle Stop check valve next to the Pump  due to Surge pressure. According to the client  when the Client Stopped one of the pumps & started the Other ( 1 working & 1 std by pump) , the Angle Stop check valve next to the Pump which was stopped got damaged . According to him it was due to surge pressure . The Total pipe length is about 40 metres from the Discharge point and the valve. Is there any solution that we can adopt to mitigate this problem ?

### Latest Visitors

• elsevier
Today, 08:40 AM
• John Gill
Yesterday, 11:48 PM
• Anurag Patel
Yesterday, 11:36 PM
• sers
Yesterday, 02:38 PM