Why are these pipes built in a strange way?
April 19, 2011 12:20 AM Subscribe
Why can't LNG (liquid natural gas) pipes have long straight runs?
Please have a look at this photo, scanned from my newspaper a couple of days ago. It shows a tanker carrying LNG (from Russia to Japan), just about to dock. In the background, we see the series of pipes that will carry the gas from the ship over to the storage tanks on shore.
At intervals in each pipe, it makes a series of four 90 degree bends - left, right, right, left - and then continues on course. This happens not only to the pipes on the pier structure, but to those on shore too. There are no obvious obstructions requiring such a diversion. Here's a closeup. Why is this necessary?
Please have a look at this photo, scanned from my newspaper a couple of days ago. It shows a tanker carrying LNG (from Russia to Japan), just about to dock. In the background, we see the series of pipes that will carry the gas from the ship over to the storage tanks on shore.
At intervals in each pipe, it makes a series of four 90 degree bends - left, right, right, left - and then continues on course. This happens not only to the pipes on the pier structure, but to those on shore too. There are no obvious obstructions requiring such a diversion. Here's a closeup. Why is this necessary?
I wonder if it is related to the water hammer effect. Particularly, there's this in the mitigating measures:
Shorter lengths of straight pipe, i.e. add elbows, expansion loops. Water hammer is related to the speed of sound in the fluid, and elbows reduce the influences of pressure waves.
posted by sbutler at 12:43 AM on April 19, 2011
Shorter lengths of straight pipe, i.e. add elbows, expansion loops. Water hammer is related to the speed of sound in the fluid, and elbows reduce the influences of pressure waves.
posted by sbutler at 12:43 AM on April 19, 2011
Doesn't it need to be pressured to stay liquid and not become a gas again, and won't it be much more difficult in a long straight pipe to keep up that pressure?
posted by ijsbrand at 12:45 AM on April 19, 2011
posted by ijsbrand at 12:45 AM on April 19, 2011
Best answer: Another reason pipes require deviations like this is that they must account for thermal expansion (contraction).
When you are not carrying a fluid, expansion joints can be used for this purpose, which is one reason why you don't see this phenomenon in every application where something has to be extended over a large distance. In a pipe, the joints must maintain a seal.
If you have thermal expansion (or contraction) in a steel pipe, the pipes length can increase by millimetres to centimetres, depending on the overall length. Now, imagine lining the pipes up end-to-end -- the effect is additive. Without elbows like this, you would soon have ruptures or buckling in your pipe.
Not nice, when you are carrying LNG ;-)
posted by rhombus at 12:54 AM on April 19, 2011 [4 favorites]
When you are not carrying a fluid, expansion joints can be used for this purpose, which is one reason why you don't see this phenomenon in every application where something has to be extended over a large distance. In a pipe, the joints must maintain a seal.
If you have thermal expansion (or contraction) in a steel pipe, the pipes length can increase by millimetres to centimetres, depending on the overall length. Now, imagine lining the pipes up end-to-end -- the effect is additive. Without elbows like this, you would soon have ruptures or buckling in your pipe.
Not nice, when you are carrying LNG ;-)
posted by rhombus at 12:54 AM on April 19, 2011 [4 favorites]
Best answer: rhombus: "Another reason pipes require deviations like this is that they must account for thermal expansion"
This.
Generically, they're known as 'compensators', although that term tends to be used for actual expansion joints / bellows. 'Loop expansion joint' is the specific term for the short 'square' bends. You see them in any long (otherwise) straight pipe where a normal expansion joint can't normally be used e.g. high pressure oil, gas, steam, etc.
Often, after every few loops one will have a steel bellows joint anyway; I'm not sure of the engineering reason why, but presumably there is one.
posted by Pinback at 2:43 AM on April 19, 2011
This.
Generically, they're known as 'compensators', although that term tends to be used for actual expansion joints / bellows. 'Loop expansion joint' is the specific term for the short 'square' bends. You see them in any long (otherwise) straight pipe where a normal expansion joint can't normally be used e.g. high pressure oil, gas, steam, etc.
Often, after every few loops one will have a steel bellows joint anyway; I'm not sure of the engineering reason why, but presumably there is one.
posted by Pinback at 2:43 AM on April 19, 2011
Best answer: rhombus writes "Another reason pipes require deviations like this is that they must account for thermal expansion (contraction)."
Besides thermal expansion a loop like this allows for movement. IE: the ground moving relative to itself at other places. I'd bet that terminal is built on reclaimed land and is likely to move a lot during an earthquake.
You'll also see this where gas comes into a house, the connection between the meter and the house will have a a couple 90s on the same axis offset a foot or more. that allows the meter to move up and down relative to the house without the pipe breaking off. The 90s acts like a hinge.
posted by Mitheral at 3:05 AM on April 19, 2011
Besides thermal expansion a loop like this allows for movement. IE: the ground moving relative to itself at other places. I'd bet that terminal is built on reclaimed land and is likely to move a lot during an earthquake.
You'll also see this where gas comes into a house, the connection between the meter and the house will have a a couple 90s on the same axis offset a foot or more. that allows the meter to move up and down relative to the house without the pipe breaking off. The 90s acts like a hinge.
posted by Mitheral at 3:05 AM on April 19, 2011
Secondary side effect of a loop like that is that you can walk/drive UNDER the pipe at those points. Or run your herd of imported antelopes through. Or some camels.
posted by FauxScot at 4:42 AM on April 19, 2011
posted by FauxScot at 4:42 AM on April 19, 2011
I too am going with the expansion joint theory. "Water" hammer is caused by the momentum of the stuff in the pipe, not the shape of the pipe. It is solved by having pressure dampers somewhere along the line to absorb the momentary pressure spikes.
(Although water hammer effect might be one of the things the expansion joints are protecting against.)
Often, after every few loops one will have a steel bellows joint anyway; I'm not sure of the engineering reason why, but presumably there is one.
The compensators can only protect the pipes against movement in one plane. I forget the exact design specs. (having read it in passing on a website about refrigeration), but I believe they only protect the length of the pipe, and cannot be asked to absorb torsional stresses. The lengths of pipe that are perpendicular to the main run of pipe bend to absorb length variations. But if one segment moves up or down in relation to another segment, the little length of pipe that runs parallel to the main run will be subject to twisting stresses, which it cannot take. So the bellows only allow the pipe to bend along its length, not to expand in length.
If memory serves.
posted by gjc at 6:14 AM on April 19, 2011
(Although water hammer effect might be one of the things the expansion joints are protecting against.)
Often, after every few loops one will have a steel bellows joint anyway; I'm not sure of the engineering reason why, but presumably there is one.
The compensators can only protect the pipes against movement in one plane. I forget the exact design specs. (having read it in passing on a website about refrigeration), but I believe they only protect the length of the pipe, and cannot be asked to absorb torsional stresses. The lengths of pipe that are perpendicular to the main run of pipe bend to absorb length variations. But if one segment moves up or down in relation to another segment, the little length of pipe that runs parallel to the main run will be subject to twisting stresses, which it cannot take. So the bellows only allow the pipe to bend along its length, not to expand in length.
If memory serves.
posted by gjc at 6:14 AM on April 19, 2011
Agree with the previous answers. These expansion joints are a common design feature of long pipes.
In the photo I can see a lot of expansion joins on the pier, a lot more then what seems normal for the length of pipe. A possible explanation might be that this is a unloading facility. The pressure in the pipes is probably lower than the pressure in the tanks on the ship. This means that the gas will expand when unloaded, and temperature will fall. The pipes will shorten considerably.
Bonus: here's a picture of what heat can do on long stretches of metal. Of course, cooling down will straighten the rails.
posted by Psychnic at 6:53 AM on April 19, 2011
In the photo I can see a lot of expansion joins on the pier, a lot more then what seems normal for the length of pipe. A possible explanation might be that this is a unloading facility. The pressure in the pipes is probably lower than the pressure in the tanks on the ship. This means that the gas will expand when unloaded, and temperature will fall. The pipes will shorten considerably.
Bonus: here's a picture of what heat can do on long stretches of metal. Of course, cooling down will straighten the rails.
posted by Psychnic at 6:53 AM on April 19, 2011
I'm a refinery piping engineer, so I'll chime in.
As previously stated, these are for dealing with thermal expansion. Typically the pipes will be anchored in between the loops do direct the growth into the loop (or out of the loop if the line is cold.) The lines are also guided to keep thermal movements in the axial (along the pipe) direction.
To put these movements into perspective, a steel line will grow 1.4" per 100ft at 250F and 3.62" per 100ft at 500F.
posted by bajema at 10:21 AM on April 19, 2011 [1 favorite]
As previously stated, these are for dealing with thermal expansion. Typically the pipes will be anchored in between the loops do direct the growth into the loop (or out of the loop if the line is cold.) The lines are also guided to keep thermal movements in the axial (along the pipe) direction.
To put these movements into perspective, a steel line will grow 1.4" per 100ft at 250F and 3.62" per 100ft at 500F.
posted by bajema at 10:21 AM on April 19, 2011 [1 favorite]
This thread is closed to new comments.
posted by sbutler at 12:40 AM on April 19, 2011