There is a major push to install remote isolation valves on the suction lines to many of our pumps. This is necessary for safety. In order to try and hold costs down, the piping project managers are pushing hard to get me to accept butterfly valves for the larger sizes. These valves are sometimes installed in the common suction line before it tees off to the two pumps. However, often the isolation valve is located just downstream of the existing gate valve used for pump isolation. This would place the butterfly valve within 2 or 3 pipe diameters of the suction flange. I know this is a bad idea. I know it will reduce NPSH margin, increase turbulence and encourage vibration. What I am looking for is some real world experience from other engineers that I can use as a basis to resist this effort. Has anyone had a reliable pump turn into a "bad-actor" after a butterfly valve was installed? Has anyone been forced to go back at substantial expense and replace butterfly valves in order to address chronic reliability problems? Any examples or experiences related will be greatly appreciated.
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Butterfly Valves in Pump Suction 1
- Thread starter JJPellin
- Start date
Sorry I can't give you any referece for what could be a problem installation,especially with the valves so close to the pump inlet.
What do the valve manufacturers have to say?
Always a problem to go against your beter judgement but you have been around for long enough to understand the need to document your apprehension of using butterfly valves in this installation.
What do the valve manufacturers have to say?
Always a problem to go against your beter judgement but you have been around for long enough to understand the need to document your apprehension of using butterfly valves in this installation.
I don't think I would use a butterfly for an isolation valve. Is true isolation the objective?
BigInch![[worm]](/data/assets/smilies/worm.gif "[worm] [worm]")
-born in the trenches.
BigInch
-born in the trenches.I have used butterfly valves in the past for suction isolation (in a standby system) of chilled as well as cooling water pumps. In all cases, there was positive suction. Didn't see any adverse effects.
The data (all maximum values) for a rough idea
Pump flowrate - 300 m3/hr
Water temperature - 320C
Suction Pipe - 10"
Valve to suction flange - 30"
The data (all maximum values) for a rough idea
Pump flowrate - 300 m3/hr
Water temperature - 320C
Suction Pipe - 10"
Valve to suction flange - 30"
EdStainless
Materials
Do you have lots of head at the intake now? What about disolved gases? The only times that I saw problems was in systems handling hot water with lots of disolved gases. There the small head loss through the valve was enough to push the pump out of the safe operating intake range.
= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection
= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection
You had mentioned that valves are being installed for "safety" reasons. For whatever it is worth and depending of course on line size, if it were ever necessary to put workers in the line the choice of a butterfly valve could prevent worker or rescuer access at least past that point from either direction (in other words the "means of entry or exit" associated with a confined space could certainly be more "limited or restricted"). Also, if it were ever necessary or desirable to "pig" or inspect the interior of this line, for whatever reason with whatever device, this procedure would also be severely limited or prevented by the presence of the butterfly valve. Others have hit on other aspects.
Exactly. I wouldn't put anyone in front of a butterfly or behind a closed check.
BigInch-born in the trenches.
BigInch
-born in the trenches.- Thread starter
- #8
I should have provided some additional detail based on the questions I am getting. These valves are remote isolation to be used only in the event of a fire. If the pump is on fire and the operators cannot get to the block valves (gate valves), they need to have the capability to hit a switch in the control room, close these valves to block off the source of fuel that is feeding the fire. They will have no equipment isolation (for maintenance) or energy control use. The fact that the valves will be installed for this purpose is not negotiable. They are going to install valves on the suction lines to a large numbers of pumps. The only decisions that I may be able to affect are the type of valve and the exact placement. The actuators are pneumatic; air pressure to open, spring to close. They must be this way to be fail-safe. If the lines or wires burn through, the valve will close. Because of the spring to close requirement, they need to be quarter-turn type valves which drives us to ball or butterfly. Ball valves are much larger and much more expensive especially in the larger size. In the end, we don't expect to ever use these valves. If there is no fire, they will never be actuated. We will test them at every TA, rebuild them on some predetermined schedule but never actually use them to isolate the pumps (I hope). I plan to place the following conditions on the use of these valves in all circumstances:
1. Evaluate the NPSH margin taking into account the additional head loss imposed by the valve. With the valve installed, we still have to satisfy our specifications for NPSH margin.
2. Place the valve back as far away from the pump as possible. It is preferable to install a single valve in the common line to both pumps. This way any turbulence or maldistribution of velocity would have equalized by the time the flow reaches the pumps.
I would prefer to say "NO Butterfly valves" within 10 pipe diameters as a firm rule for the project. But I don't know if they will accept the higher cost of piping changes or ball valves without some justification. My engineering judgment is that the butterfly valves are a bad idea and may cause a reduction in pump reliability in some cases. But proof is hard to come by. API RP-686 states that no fittings (valve, reducer, elbow, etc.) should be located within 5 pipe diameters of the suction of a pump. But we have pumps all over the refinery (100's of them) that have an eccentric reducer and elbow right up against the suction flange. And most of these pumps run well. But that does not prove that it is a good idea to pipe them that way.
I was hoping that someone had gone down the path that I am being pushed down and that I could benefit from their hard learned lesson.
1. Evaluate the NPSH margin taking into account the additional head loss imposed by the valve. With the valve installed, we still have to satisfy our specifications for NPSH margin.
2. Place the valve back as far away from the pump as possible. It is preferable to install a single valve in the common line to both pumps. This way any turbulence or maldistribution of velocity would have equalized by the time the flow reaches the pumps.
I would prefer to say "NO Butterfly valves" within 10 pipe diameters as a firm rule for the project. But I don't know if they will accept the higher cost of piping changes or ball valves without some justification. My engineering judgment is that the butterfly valves are a bad idea and may cause a reduction in pump reliability in some cases. But proof is hard to come by. API RP-686 states that no fittings (valve, reducer, elbow, etc.) should be located within 5 pipe diameters of the suction of a pump. But we have pumps all over the refinery (100's of them) that have an eccentric reducer and elbow right up against the suction flange. And most of these pumps run well. But that does not prove that it is a good idea to pipe them that way.
I was hoping that someone had gone down the path that I am being pushed down and that I could benefit from their hard learned lesson.
Sometimes "good" just doesn't work out no matter how hard you try.
BigInch-born in the trenches.
BigInch
-born in the trenches.
MikeHalloran
Mechanical
I really, really, really hate that manager tactic:
"Find me someone who is dumb enough to admit that (what I'm pushing you to do against your better judgement) had a bad outcome ... or your tenure here will have a bad outcome", or words to that effect.
Do they teach that in business school?
Mike Halloran
Pembroke Pines, FL, USA
"Find me someone who is dumb enough to admit that (what I'm pushing you to do against your better judgement) had a bad outcome ... or your tenure here will have a bad outcome", or words to that effect.
Do they teach that in business school?
Mike Halloran
Pembroke Pines, FL, USA
I turn in my recommendation as I see it and if anybody wants to change it, they get to stamp mine rejected and sign off on their own.
BigInch-born in the trenches.
BigInch
-born in the trenches.Automate the existing valves. I've place ESD valves within 6 pipe diameters with no long term problems.
We have lots of posiseal butterfly valves that ESD to protect our equipment, and yes, we use them to isolate the same equipment. The service is cryogenic gas at 800 psig. If the equipment is take out of service, a blind flange is attached. It takes two sets of flanges, one for the valve to be set betwen, the other to blind the valve. Yes, we have a small gas leak going by the seals of the butterfly valve that we must deal with. If you can not afford some leakage, then go to a ball valve or DBB gate.
We have lots of posiseal butterfly valves that ESD to protect our equipment, and yes, we use them to isolate the same equipment. The service is cryogenic gas at 800 psig. If the equipment is take out of service, a blind flange is attached. It takes two sets of flanges, one for the valve to be set betwen, the other to blind the valve. Yes, we have a small gas leak going by the seals of the butterfly valve that we must deal with. If you can not afford some leakage, then go to a ball valve or DBB gate.
"1. Evaluate the NPSH margin taking into account the additional head loss imposed by the valve. With the valve installed, we still have to satisfy our specifications for NPSH margin.
2. Place the valve back as far away from the pump as possible. It is preferable to install a single valve in the common line to both pumps. This way any turbulence or maldistribution of velocity would have equalized by the time the flow reaches the pumps."
I would evalualte point 1. and make a recommendation on the valve location based on your own judgement and would not be prepared to sign-off an any other recommendation.
If the valves are too close and cause problems - you have a problem. If installed 10 pipe diameters upstream of the pump inlet and everything is OK the additional costs do not become a consideration.
2. Place the valve back as far away from the pump as possible. It is preferable to install a single valve in the common line to both pumps. This way any turbulence or maldistribution of velocity would have equalized by the time the flow reaches the pumps."
I would evalualte point 1. and make a recommendation on the valve location based on your own judgement and would not be prepared to sign-off an any other recommendation.
If the valves are too close and cause problems - you have a problem. If installed 10 pipe diameters upstream of the pump inlet and everything is OK the additional costs do not become a consideration.
Do the safety isolation valves need to meet some insurance underwriting criteria or be "fire-safe" design? Some butterfly valves can, but they are gasketed and have metal backup seats specifically for such use.
For my company valves that do not meet "fire-safe" get fire proof insulatation and/or water spray per NFPA, to ensure fire does not freeze the valve or actuator. Rather than leave it to human factors (to actuate the valve), our valves have fail closed actuators with "fusible links" (a piece of pressurized tubing" that melts off releasing the pressure when exposed to fire.)
Whether the valves will cause flow disturbances depends on the suction energy. If there is a high suction energy and not enough suction head or straight pipe, caviation can occur. I will try to get the information I have gathered tomorrow & post again tomorrow night.
G
For my company valves that do not meet "fire-safe" get fire proof insulatation and/or water spray per NFPA, to ensure fire does not freeze the valve or actuator. Rather than leave it to human factors (to actuate the valve), our valves have fail closed actuators with "fusible links" (a piece of pressurized tubing" that melts off releasing the pressure when exposed to fire.)
Whether the valves will cause flow disturbances depends on the suction energy. If there is a high suction energy and not enough suction head or straight pipe, caviation can occur. I will try to get the information I have gathered tomorrow & post again tomorrow night.
G
Since you have stated that you also have to select the "type" of valve, then do the NPSH margin calculations (remember that most pump NPSHR curves are optimistic) and on the pumps that are safe use butterfly valves, and on the ones that are close use ball valves. On the ones that are too close to call, have them repipe them-upsize the suction piping and then put in the BF valve. Yes, it will be costly but no more so than replacing pumps that are torn up by cavitation.
Where you can, use the butterfly valves to your advantage. Often pumps can use a little bias in the suction to offset piping errors or flow patterns. A butterfly valve passes full flow long before it is 100% open, so if some can be stroked so that they give a little 'tweak' in the direction that is beneficial to the pump, then set it up that way. Remember that butterfly valve shafts don't necessarily have to be vertical either. That 'tweak' might be to compensate for a concentric reducer that is not properly placed. So put them horizontal or in what ever position helps the pump the most.
If you are going to be made to do it, then do it your way.
rmw
Where you can, use the butterfly valves to your advantage. Often pumps can use a little bias in the suction to offset piping errors or flow patterns. A butterfly valve passes full flow long before it is 100% open, so if some can be stroked so that they give a little 'tweak' in the direction that is beneficial to the pump, then set it up that way. Remember that butterfly valve shafts don't necessarily have to be vertical either. That 'tweak' might be to compensate for a concentric reducer that is not properly placed. So put them horizontal or in what ever position helps the pump the most.
If you are going to be made to do it, then do it your way.
rmw
dcasto, If you take it out of service, have 800 psig gas leaking past the seal and need to fix a blind flange, I can't possibly see how you can say that a butterfly "is used for isolation"?
BigInch-born in the trenches.
BigInch
-born in the trenches.You didn't say just what type of pumps you are shutting off. Split case double suction impeller or Ansi style end suction. Either way you may have a problem due to turbulence. I did a start up checks on 26 pumps that the engineer allowed the inlet isolation B/F valves to be placed within 4 pipe diameters of the inlet. All pumps were Horizontal Splitcase, double inlet impellers. We found that during the installation the valves were not trimmed and adjusted to full 100% open. The contractor just bolted them in place and let them be. On numerous pumps we had vibration problems and bearing problems. Water being directed to either one side of the impeller or the other will cause an Hydraulic Imbalance and premature thrust bearing fauilure. All the B/F valves were then adjusted using the stem as a guide to set the valve at 0 & 100%. This still did not fully correct the entire problem. The end user ended up removing the B/F valves and installing relisient seated knife gate valves as their shut off & isolation valves with a true unobstructed 100% open flow path. Once this was done, all vibration and bearing issues disappeared.
BigInch, agreed, but we have to depend on the valve until we get the expander out and the blind put back on. I had posiseal give me all their data (not that I could get all the stress calculations myself) the shows that their valve can withstand 1480 psig differential and not break the stem or blow out the seal faces.
- Thread starter
- #19
Thank you all for the replies. These valves are proposed for a number of different pumps. The vast majority will be single suction, over-hung, single stage, radial split API pumps. I can definately see where a double suction impeller could have additional problems from uneven flow to the two sides. Suction energy is an interesting concept that I need to study more. If this can give me a flag to look for likely problems independent of Nss or NPSH margin it could be very valuable.
MMcLean
Chemical
- Jul 21, 2003
- 25
I've used butterfly valves on large pump suction lines many times, due to cost contrainst driving it that way. If they are an isolation valve for hazardous service or a service that crystallizes when cooled, preventing the valve from re-seating properly, that is the only problem I'd have with them. The location very close to your pump could have some impact on the pump operation, but betterfly valve are a common valve in large lines due to their low cost. Depending on the line size, you may stuggle to get a ball valves big enough.
Mark McLean
Mark McLean
- Status
Similar threads
- Locked
- Question
