16 replies to this topic

[ZAS]

Hi,

During a routine maintenance, a wrong isolation caused a power loss at our jetty which subsequently triggered foam skid and deluge skid at the jetty(which is at hihger elevation). This in turn initiated the main electric pump to start in order to pressurise the firewater ring main and followed by Diesel Pump, pressure rose from 0 to 16 barg in 10seconds. This sequence of events caused two bypass valves to split. We suspected a surge problem in the system. I have attached the pressure profile in this post

We are considering the following solutins:
1.Use vacuum breakers
2.Modify control system so that deluge valve will not open on power loss
3.Use surge drum
4.Modify pressure settings of Jockey and Main pumps

Can anyone please share thoughts and experience and if whether we undertood the problem and that the pros and cons for the solution. Also please kindly share with me other available asolutions.

Many thanks

Attached Files

•  Pressure Profile.bmp   949.27KB   98 downloads

[fallah]

Lack of adequate information for assistance!

Local power failure just in jetty? If so,why diesel pumps have been started? If there is total power failure how electrical pumps have been started?

A sketch could show your system details.The architecture of your safety instrumentd system isn't so clarified.

Submitting the complementary information (some of them are mentioned above) enable whom is going to help in this regard.

[breizh]

Hi ,
My advises will be :
Discuss with the company which installed your fire fighting protection and try to get support from the insurance company which is involved in the riks management of your facility . You cannot modify anything without talking to them .
NFPA should be a good resource .

Hope this helps
Breizh

[ZAS]

Many thanks Fallah and Brezh as I have not thought about the insurance. Apologies for my lack of clarity.
 FW.bmp   2.79MB   108 downloads  FW.bmp   2.79MB   108 downloads
Local power failure causes deluge valve to open. As pressure reduces from 7barg to around 1barg, the Main Electric pump kick in and raised the ring main pressure to 16barg in 10seconds. This resulted in damage to two valves. I have attached a sketch of the firewater system

The operations sequence of the firewater pumps is as follows:
If pressure falls below 5.8 barg, jockey pump starts
If pressure falls below 5.5 barg, second jockey pump, starts
If pressure falls below 4.5 barg, electric main pump, starts
If the pressure does not rise to 8.0 barg after 20s, the diesel main pump, starts.
If pressure rises above 6.5 barg, the jockey pumps stop.

I am serching for solution that would not allow this pressure surge.

Many thanks.

[Zauberberg]

We've had similar problems in one of the plants where I used to work, and the problem was solved by installing appropriate deluge valves (higher pressure rating) capable of withstanding surges which may occur when the ring main pressure drops for any reason.

In addition to that, your pumps may have a PCV on the discharge line that will ensure pre-set pressure not to be exceeded after the ring main pressure is restored. NFPA standards, as mentioned by Breizh, should be your reference.

[gvdlans]

Before you make a decision on any solution, I would recommend to have a detailed surge analysis done. I have done such studies in the past using Pipenet Transient software. What I think has happened here is that when pressure dropped a vacuum was created at the high points in your network. When the pressure came back up after the main pump started, the vacuum was violently filled with water again resulting in surge pressures at the valves resulting in valve damage. If that was the case, the solution may be to install properly sized surge suppression air valves, that let air enter when pressure drops. The air is (relatively) slowly released when pressure comes back again so that surge is prevented. See for example http://www.valmatic....uppression.html

[ZAS]

@Zauberberg: The PCV we have only allows flow from Jockey pumps to Firewater pond. We do not have relief system on the ring main, if the main duty pumps kick.

@gvlands: I thought there shouldn't be any air in the fire water systems. Or is that limited to pump suction? Do you have any reference of using Surge suppression air valve on Firewater systems?

We have done surge analysis which identify the problem as: Vapour pocket formed due to pressure drop in ring main and the subsequent vapour collapse due to sudden pressure rise(by main FW pump). My thoughs were,if I could eliminate(or at least minimise) vapour pockets and their sudden collapse, then that would solve the problem. But then how can I achieve that. The valves that got damaged were at a higher elevation of 53m AOD compared to the Jetty Deluge valves(15mAOD).

[gvdlans]

The results of your surge analysis are virtually identical to my analysis in my previous post. Note that what you call a "vapour pocket" is what I call a "vacuum". In reality it will indeed not be a complete vacuum but there will be some water vapour in it. Vacuum breakers/air release valves can be used in firewater systems, normally only valves that are FM/UL approved are used. For references I advise you to contact suppliers of such valves (Google is your friend).
I do not agree with your assumption that "there shouldn't be any air in the fire water systems." Take for example the spray nozzle systems that are normally not water filled (and therefore filled with air). Only when the deluge valves open, the spray nozzle system is filled with water...

[ZAS]

Many thanks for your invaluable responses.

[ZAS]

One more thought that leave me sleepless is this:
In our firewater design, when Deluge Valve opens( e.g. due loss of power ) it cause the main Firewater pump to start to maintain Ring main pressure. If the deluge valve were to inadvertently close for say 2-10seconds or more and the main fire water continues to operate, this may lead to pressure in the Ring main and associated piping to exceed design. In my case, a simulation result for this scenario far exceed pipe rating by up to 300%.

Has anyone come across this scenario or am I worrying about the imppossible. If you do what can I do to mitigate this if it were to happen.

Many thanks for your anticipated valuable response.

Amin

[breizh]

Hi ,
As per NFPA , you should have a PSV at the discharge of the pump to prevent such scenario !

Breizh

[ZAS]

As per NFPA section 5.18.3 on Location of PSV. The relief valve shall be located between the pump and the pump discharge check valve ....

Our system(see attachment in previous mail) is also designed as per this NFPA requirement. This provision only protects piping downstream fire main pump before the check valve but not the piping downstream the check valve. I want to know what sort of protection against overpressure is normaly provided for ringmain piping.

Many thanks

[breizh]

Hi,

*To protect your network from surpression you need to install Pressure reducing valves.

Breizh

[ZAS]

Many thanks.

[fallah]

ZAS,

It might be provided by using check valve with Non Slam characteristic.

Fallah

[ZAS]

Fallah,

Please clarify more on how check valve may protect against overpressure in firewater ring main piping?

[fallah]

ZAS,

Seems you mean overpressure due to surge pressure that may be created following fire water pump stopping.If so:

Non slamming characteristics provides conditions for a check valve based on its fast/slow closing (not intermediate),when the pump suddenly stops,the overpressure due to slamming in downstream piping (main ring) would be eliminated/minimized.

Fallah