Smart questions
Smart answers
Smart people

Member Login

Remember Me
Forgot Password?
Join Us!

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips now!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!

Join Eng-Tips
*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Donate Today!

Do you enjoy these
technical forums?
Donate Today! Click Here

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.
Jobs from Indeed

Link To This Forum!

Partner Button
Add Stickiness To Your Site By Linking To This Professionally Managed Technical Forum.
Just copy and paste the
code below into your site.

rsv (Mechanical)
30 Nov 05 18:22
I don't know whether this is a simple Q or not.
Why use ball valves on Gas pipelines instead of gate & or globe valves, am I missing something?

I am new to Gas pipelines having mostly worked in petrochem & powerstations where gate & globe valves are the standard for isolation because they are cheap & reliable.

I have used small 150# ball valves in Pharmaceutical plants but these are 30" 2500# and must cost way more than gate valves.
zdas04 (Mechanical)
30 Nov 05 18:33
Ball valves do cost more than gate valves, and they're darn sure worth it.

Gate valves take a lot of effort to cycle (an 8" gate valve requires up to 100 turns).  I started in steam plants where they were common and my first Oil & Gas project I specified them--some of the operators are still cussing my name.  Also, in natural gas service, gate valves are not advertised as bubble-tight, while ball valves (especially trunnion) are.

Globe valves are a good choice in throttling service in most fluids.  Any throttle valve will develop wear patterns on the seating surface fairly quickly and will not provide a postitive shut off.  Also the changes in direction within the globe valve create a non-zero pressure drop.  If you are using a globe valve in block-valve service then you're wasting compression (or pump) hp.  It may be small, but over time it adds up to significant wasted energy.

David Simpson, PE
MuleShoe Engineering
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

The harder I work, the luckier I seem

rsv (Mechanical)
30 Nov 05 19:11
I agree with the effort used by ops. but, I forgot to mention that the site is mostly automated so gates can use smaller actuators and take up less room.

Good gate valve manufacturers are offering resiliant seats which I assume will be as leak tight as ball?
zdas04 (Mechanical)
30 Nov 05 19:17
If it don't say "bubble tight" it aint.

30 Nov 05 22:01
Ball Valves versus Gates and Globes?
1) Rotating in the packing instead of reciprocating the process from within or the environment from the outside through the packing gland along with the stem.
2) Elimination of the bonnet gasket and potential leak path.
3) Sealing member rotates in it's own volume instead of having to displace a volume.
4) A full ball is a balanced trim. A half ball, gate or globe valve plug would be unbalanced. This would be important with actuators if you lost your gas or air supply.
5) No collection zone in the bottom of the valve. There has to be room for the gate to move to and may have difficulty displacing what ever has collected there.
6) Smaller overall envelop dimension.
7) Ease of automation - ninty degrees to go full open to full closed. Not multiples of turns.
8) Ball valves for control must be sized by Cv not line size. They can have the same capacity as a globe that is twice their size. In other words, they can be half line size. Ball valves have equal percentage inherent characteristics. So when applied to systems where 30% or less of the system drop is applied to the valve, will provide a "linear" installed characteristic.
Helpful Member!(2)  JimCasey (Mechanical)
1 Dec 05 9:32
It appears the general trend of this thread  is going in the right direction, but I will add my 2-cent's worth (after adjusting for inflation)

Gate valves were invented many years ago. They could be manufactured with the machinery and materials available in the early industrial revolution.They had a head-start on the market.  Ball valves were invented in 1949. This was made possible  by the availability of some of the polymers created in the 20th century.  Although PTFE was invented in 1928, its development and availability as an industrial material enabled the development of the ball valve.  
Ball-Valves shut off tight.  "Bubble-tight" means nothing, but API6D does.  

Ball valves are more compact.  Generally manufactured to the same lay-length as gate valves (ANSI B16.10), but look at the dimansions of the assembly: Gate valve has to have a "chest" or bonnet, equal to the valve diameter for the gate to occupy when the valve is open.  Then there's a packing box, and a yoke that has to provide for full diameter travel of the actuator stem.  Then there's the actuator which has to be beefy enough to provide the "Hammer blow" feature necessary to seat and unseat the gate. Then above the actuator it is necessary to provide room for one diameter of stem to stick out when the valve is open (frequently inside a tube to protect the grease on the threads).So a motorized gate valve is typically around 5 diameters tall, and almost always has to be installed with the stem vertical to reduce bending loads on the "tower". Compared to a ball valve which has a short packing box, a short bracket, and an actuator.

Environmental concerns.  Cycling a gate valve drags a diameter's worth of valve stem through the packing box, and drags a diameter's worth of environmental contamination ofn the valve stem back into the packing box. All the towering mechanism above the packing box puts a hopefully small but finite side load on the packing due to imbalanced mass. line vibration, etc.   Packing in a gate valve therefore wears very quickly.  Compare to a ball valve where the microscopic roughness on the valve stem moves packing particles through a rotational path that remains inside the stuffing box. Net migration of mass from/through the packing assembly in a ball valve is orders of magnitude lower than the gate valve.  With a less expensive, simpler packing design a ball valve can have zero leakage to the environment for much longer than a gate valve.  

Speed: A large gate valve requires minutes to close with a big electric actuator grinding away. Ball valves need speed controls on pneumatic actuators to slow them down to prevent water hammer.  Which would you want to isolate your piping in the event of a fire?

Gate valves are guided very loosely except when the gate is wedged into the seat.  If the operator tries to throttle with a gate valve all these loosely-affiliated pieces will jingle together due to flow eddies. Control will never be precise because the relationship of the trim pieces is not precise.  Wear will be rapid, and catastrophic failure is likely. Ball valve: Guided positively.  Small ball valves have the ball secured in an interference fit between two seats.  "Floating" balls don't float very much, and careful fitting of the stem tang into the ball slot ensures accurate positioning of the ball.  Larger trunnion valves are guided positively by top and bottom bearings (trunnions) so the ball is not deflected bt process effects, and the ball-to-stem joint is usually a positive connection.  

Conclusion: if you want a cheap valve that is seldom if ever cycled and can live with a small trickle when it is closed or if you have large maintenance apes with long cheater bars to try to get tight shutoff, then gate valves are OK.  If you cycle the valve with any regularity, or ever throttle with it, and want a high probablility of zero leakage either from upstream to downstream or from inside to outside, than a ball valve is a better choice.  
Ashereng (Petroleum)
1 Dec 05 11:44
Hi rsv,

From your question "Why use ball valves on Gas pipelines instead of gate & or globe valves, am I missing something?", I would like to ask you,

           "What do you want this valve to do?"

Depending on what you need the valve to do, you need to select the valve appropriately. The valve is selected "fit for purpose".

Having said that, there are usually multiple options for valves for each application. There are also many factors in valve selection (valve sizing, cost, etc. are each a component to be considered in determining the selected valve).

As with anything, go to the expert/professional for advise. If one needs information on pumps, consult a mechanical rotating equipment resource. If one needs information on valves,  I strongly suggest consulting your I&C resource (that's what they do) and ask them for further info and advice.

All of the above posting provide valuable information, insight, descriptions and value. However, they are not sufficient to select a valve as is, without further data and information.

One other note, the term "Bubble Tight" is sort of misleading.

ANSI B16.104-1976/FCI 70-2-1976, defines 6 seat leakage classs (note that they actually say "seat leakage" classes). All classes have an allowable "leakage rate".

Class VI is intended for resilient-seating valves (or soft seat), and provides that the leakage limits depends on valve size and ranges from 0.15 to 11.5 ml per minute for valves sizes 1" to 12", and 1 to 81 bubbles per minute.

The difference between the class include for different valve type, the allowed leakage, and testing conditions.

If you do not have an I&C resource available to you, this forum may be of help to you. I have seen other postings where people have provided sufficient information that someone was able to provide a solution.


stanier (Mechanical)
1 Dec 05 17:31
Ball valves can be fire safe if they have a secondary seat and that is a requirement for gas service. Some gases contain componenets that cause seats to stick and you would be better off with a ball valve as there you wipe the seat. If it is real sticky then a valve like the Orbit valve is used. This has a complex mechanism where the ball moves axialy away from the seat and then rotate 90 degrees.

Careful of the quick opening if your gas has condensate. A slug of condensate accelerated through a ball valve by opening too quickly can cause damage.

There a re few books around on the subject of valve selection. I havent had the opportunity to review any. Have other forum members found any of real value? Too many books around are just cribs of "old wives tales" and stuck together as an authorative work. Best use for them is a door stop. These are generally titled "Handbook of... or Guide to..".

Helpful Member!  jsummerfield (Electrical)
1 Dec 05 18:12
ANSI or ASME B16.104 is withdrawn.  FCI 70-2 was just recently updated to coordinate with the IEC.  Zero bubbles is a term within the API standards such as 598.  FCI is a control valve leakage rate.  Control valves leak.  Zero bubbles is bubble tight and clearly defined in the standards.  Typically API 6D ball valves are suitable for isolation unlike control valves.


ChokeEng (Mechanical)
1 Dec 05 22:01
Better yet, because of the sheer mass of the valve, many 6D pipeline ball valves (>12" NPS) can pass API fire test requirements without having to rely on additional "firesafe" seals because it's extremely difficult to introduce enough heat into the test environment to compromise the soft seals during the test interval. 6D ball valves excel in isolation applications and can usually be supplied in double block & bleed should the application require it.

don f.
jsummerfield (Electrical)
1 Dec 05 22:27
A 30-inch Class 2500 ball valve is large.  So is everything else in the 30-inch system.  An automated shutdown valve offshore needs to close within 45 seconds.  As the valves become larger than 30-inch the actuators require quick exhaust and other tricks to meet 45 seconds.  A 30-inch rising stem valve could be very hard to automate and meet 45 second closure.

Ball valves don't suit all applications.  API 6D ball valves are typically limited to services below 400 degree F.


Ashereng (Petroleum)
2 Dec 05 11:43
Hi rsv,

jsummerfield is correct. API 6D does have a "bubble tight", "zero bubble" classification.

I would like to highlight that "bubble tight" is a factory test/verification. Once the valve is installed and put in service, the bubble tight may no longer be there as the valve will start to deteriorate. If a true bubble tight is required, then there would also need to be a maintenance program to verify that the valve is still indeed bubble tight, and to replace the valve if required.

In my above posting, I was thinking throttling. I work more with throttle type valves - characterised ball, globe, butterfly/disc - I guess the mention of cheap, in conjunction with globe and ball got me.

Sorry about that rsv.
jsummerfield (Electrical)
2 Dec 05 14:16
I too agree about factory verses the field installation.  Run that rusty hydrostatic testing sludge through those soft seats and they may begin passing sufficiently to be a problem.


MikeHalloran (Mechanical)
2 Dec 05 23:06
I'm no pipeline expert, but I conjecture one reason for using ball valves instead of gate or globe valves is that you can pass a pig through them.

Mike Halloran
Pembroke Pines, FL, USA

JimCasey (Mechanical)
5 Dec 05 8:45
Mike:  "FULL PORT" Ball valves are piggable.  MOST ball valves are "regular port" which coincides with having the port one line size smaller than the valve size.  E.g.: 3" port in a 4" valve. A regular port ball valve probably still has less pressure loss than a gate valve because the gate valve has a substantial discontinuity around the seating surfaces where there is a pocket for the edges of the gate, vestigial guiding surfaces, the bottom lip of the gate, etc.  Compare to a regular port ball valve that has gently tapered transitions from line size to port size.  That becomes pretty close to a large-beta venturi in service and causes almost no restriction.  Or, a Full-port valve, for which the custoemr pays a premium, has the same pressure loss as an equivalent length of straight pipe.  Maybe even less because the valve's internal surface finish is probably better than he pipe's.  
MikeHalloran (Mechanical)
5 Dec 05 9:39
Thanks, Jim.

Does that regular/full port distinction apply to larger valves too?  I.e., if you ordered a 30" ball valve, and didn't explicitly specify it, how big would you expect the ball bore to be?

Mike Halloran
Pembroke Pines, FL, USA

Ashereng (Petroleum)
5 Dec 05 10:35
Hi Mike,

When I spec my valves, I specify bore size. I feel that this is a good habbit, and it eliminates misunderstandings.
jsummerfield (Electrical)
5 Dec 05 13:07
Expect that the bore of a reduced port ball valve to be at least one line size smaller than the connecting flange size.


Sircrashalot (Mechanical)
5 Dec 05 13:33
I would expect manufacturer's of 30" 2500# valve to specify the bore size very clearly on their quote and drawings.  Also for a serious valve such as this, I would expect that the specification would clearly indicate what the bore should be.

zdas04 (Mechanical)
5 Dec 05 23:04
In large-bore valves, the "regular" trim is finally starting to not be the norm.  It is still a good idea to check, but when I've asked for availability of 16-inch and larger trunnion ball-valves for the last couple of years the on-the-shelf valves have been full port and the reduced-port valves were special order.  I've also noticed that the last quote I got had a lower price for full-port than reduced-port.  I can't think of any reason I'd install a reduced port ball valve in a new service.


David Simpson, PE
MuleShoe Engineering
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

The harder I work, the luckier I seem

rsv (Mechanical)
6 Dec 05 18:54
Thanks for all the replies there have been some interesting points.

Apart from the pigging aspect why not use a butterfly valve although they were not mentioned in my original post I would imagine they are a lot cheaper & initially as leak tight as a ball & there would be no need for cavity relief?

My initial post was because of thoughts of long term seat reliability which has been mentioned above.
Metal seated valves especially with HF are allowed very marginal leakage across the seats but I would not expect it to deteriate as much as a soft seated valve.
I know trunnion mounted ball valves have cavity plugs & seat / stem sealant injection ports on larger valves to compensate for the above but, that makes me think the manufacturers are expecting these valves to leak.

Good point about emissions but I guess you could use bellows sealed.

As far as automation goes I can see the point as far as speed goes but the size of a pneumatic actuator for a large ball valve is incredible compared to an electric actuator for a gate valve especially for parallel slide.

As far as getting rid of the bonnet gasket - it is replaced with a body gasket on most ball valves & to pass the fire proof tests most ball valves require graphite packing to the stem & seat assemblies.

So ball valves appear to be the new kid on the block because our machining has gotten better but, I still think they are expensive (not quite on target but 1.5" metal seated 1500# ball cost £200 comparable globe £40) or maybe I am just a little old fashioned.
JimCasey (Mechanical)
7 Dec 05 9:13
When you compare to buterfly valves, it's important to make a distinction: Rubber lined/High-performance (Double-offset) or Triple offset.

In general, butterfly valves are limited because the vane bisects the flow path. So pigs, slurries, etc. meet an obstacle, and the sealing edge of the vane is directly exposed to erosion.  Also, the vane generates a significant dynamic torque which peaks at about 70 degrees open.  It makes butterfly valves a little challenging to throttle with.  Since the vane is on a shaft, which can be considered a beam supported outside the pipe cross-sectional area, they are limited in the shutoff they can take.  ASME 150 and 300 are universally available in HP valves, ASME600 is seen less frequently but still not impossible to find, but by the time you get above class 600 the shaft diameter has to be so big it's almost hard to tell when the valve is open.  

Rubber-lined butterfly valves are typically rated pretty low in shutoff pressure...225 psi or less.  DOuble-offset valves must have resilient seats so once the application temperature exceeds the capabilities of available polymers, metal seats are used and leakage drops usually to FCI70.2 class IV(altho there are some class VI metal seats.) Metal seats available to around 1100F, but they ARE a bit fragile since they must be made of thin stock to be conformal.    Rubber-lined and Double offset valves are position-seated. They are as closed as they get when the major axis of the vane is at 90 degrees to the pipe axis.  If you go past this point, they begin to open again.  

Triple offset valves can use metal seats for zero leakage. They are torque-seated. Usually there is a conforming insert, but the amount of distortion it must accomodate is small and the sealing member can be quite robust.    To close the valve, run the disc into the seating surface, and torque like mad to cause the surfaces to conform to each other.

Butterflies are less expensive than ball valves of the same size, shut off better at low pressures, and are very compact.  Butterfly valves usually are available in 2" and larger, while ball valves from 1/4 inch are the norm.  

Please have the following etched on a granite slab and displayed in a place of prominence:


Type X valves are not universally better than type Y valves.  One valve type does some things better than other valve types.  Knowing what you want the valve to do, and selecting the right valve to do it is the key to succesful application.  
Ashereng (Petroleum)
7 Dec 05 11:41
Hi rsv,

I would like to add that in a butterfly/disc valve, the sealing elements tend to deteriorate "faster" (a relative term) than those in a ball valve. This results in leakage.

Your original post pertained to shutoff service, but did not give specifics to what level of shutoff is required, it is difficult to say whether a butterfly is suitable in your application.

My above posts have indicated that I feel once a valve is installed, bubble tight does/may exist - this is still valid. Having said this, a butterfly may be appropriate, it all depends on the condition, requirements and service.

My suggestion is to have your I&C (instrumentation and control) resource person size/select the valve fit-for-purpose for your application. That is what we do.

rsv (Mechanical)
7 Dec 05 19:11
Thanks for all the comments.
From what I have gleaned from the above there is nothing cut & dried, it is still a question of personal choice / experience and how long you think resiliant seats will last compared to conventional HF seats.

The reason the question was asked in the first place was because I joined a project 1/2 way thru' after the valve types had been decided, I would have probably gone for gate valves but I wasn't sure if that was because of my background.

I have heard a lot of stories over the years about very early failures of soft seated ball valves are these just stories or is it still true?
Ashereng (Petroleum)
8 Dec 05 11:50

As with most stories, at one point, it was probably true, and now, probably less true.

Ball valves have come a long way "over the years", along with everything else (e.g. my car, laptop, TV).

At one point, ball valves were not really used for throttling. Now, there are many characterised ball valves designed specifically for throttling (and the characteristics can be engineered to give exactly the response you want).

With regards to the early failure of valve seats, companies like Orbit, have specific designs to minimise seat failure (they ball moves off the seat befor rotation). Other companies have improved the characteristics of the resilient seat with better materials. The use of trunnion ball valves with springs has also increased.

Also, there are many types of ball valves, and, the wrong valve in the wrong service may also give rise to these horror stories.

My suggestion, as stated above, is to have your I&C (instrumentation and control) resource person size/select the valve fit-for-purpose for your application. That is what we I&C people do.

You stated above "From what I have gleaned from the above there is nothing cut & dried, it is still a question of personal choice / experience and how long you think resiliant seats will last compared to conventional HF seats."

Your I&C resource will choose the right valve solution based on sound engineering reasoning and knowledge, and not so much on "personal preferences" (sometimes, there is a right answer) and "how long you think resiliant seas will last" (because they have actual data/information).

rsv (Mechanical)
8 Dec 05 17:18
Thanks Ashereng.

I don't think it would be a good idea for me to ask my I&C resource, as my job is to specify the valves - I don't have a problem with doing this.

The question was put to get other peoples opinions on why they thought ball valves were better for gas pipelines & I have found all the responses interesting.

As long as the valves are either actuated or not used very often it appears to boil down to the (perceived) holy grail of Tight shut off & zero emmissions to the enviroment & a smaller footprint.

I personally believe that all valves will leak probably sooner rather than later but, that the actual leakage rate although it sounds a lot compared to zero is not that bad.
If you are really really concerned about leakage then use 2 valves in series regardless of the valve type.

I would very much have to agree that the stem sealing arrangement on ball valves is very much better than standard gate / globe valves but, if it is a real concern or the process is toxic then a bellows seal gate will beat a ball valve on emmissions.

If you are using automated valves then the pros and cons are horivontal real estate for ball valves & vertical real estate for gate valves and this would probably depend on the type of plant.

I have known ball valves that could not be moved after long periods with no use because scale formed around the seals & ball and the actuator did not have the torque required to move the valve whereas a gate valve would move but may leak slightly across the seat?
Ashereng (Petroleum)
8 Dec 05 19:00

My apologies. No offence meant.

I too am in I&C, and I often ask my peers for help in areas where I deal less in, and/or am less experienced in.

You mention a good point though with valve failure (it won't close when you want it to, or vice versa). For almost all ESD applications, using the SIL analysis/method, there is a requirement to do regular checks/maintenance on the SIL valves. The reason is to avoid an on-demand failure. Any valve that is neglected will eventually fail.

In line with the ESD application - for SIL3 application, one of my client's standard is 2 parallel trains of 2 valves in series. This is very expensive. It does however have a probability of failure that is very very small, and allows for testing without taking the process down.
rsv (Mechanical)
8 Dec 05 20:01
No offence taken.

When it comes to SIL rating 2 in series does mean you can reduce the SIL level required of the  individual valve/actuator so I think it is a good idea.

I believe that people get really hung up with the SIL rating of the actual valves (It is after all an instrument safety system 2oo3 poling etc.)- I know of only two manufacturers in Europe that have had independent sil rating of there valves & 1 is not a ball valve but an Axial valve.
Even though most manufacturers have not provided fault tree analysis etc. to achieve sil rating, I believe that ball valves from reputable manu. would be OK.
Ashereng (Petroleum)
9 Dec 05 10:06

I have not come across any TUV SIL rated valves. I have on the detection side (pressure xmtrs, etc.), and of course, on the control system side.

For our SIL application, because the client was very large, they had maintenence data from which the valve availability/failure rates were mined. We were actually able to use actual data! Haven't happend since, but it was nice.

Another common practice is to use a non-SIL valve (perhaps a globe throttling control valve) up stream in series with a SIL rated ESD valve. This way, the shut-off requirement would be lower, since you are taking credit for the upstream control valve shutting down the flow somewhat.

Ooops - I think we should probably start a new thread for this.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!

Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close