How do you deal with Pipe Expansion Underground?

My (admittedly simplified) view is that pipes don't expand underground, because they are confined by soil weight and friction. Instead the pipe stresses change due to the temperature and pressure changes. If stress modelling shows that stresses will exceed allowable, or exceed the resistance provided by the soil, then there's a problem and the design needs to be modified.

Let me clear the air quickly. Metal expands when it is heated, underground, in the air, under insulation, wherever. It has no way to know where it is, it just explands at a known and predictable rate of growth per degree of temperature incremente.

Now, as to the answer to the OP, it will have to wait until BigInch gets on and enlightens us.

rmw

Well, I'm not BigInch, but I have laid a bunch of buried pipe.

The bottom of a ditch is never flat. It undulates considerably even in Kansas. Consequently, an expanding pipe will tend to be self-relieving at overbends. The stress calculations always work out to not need thrust blocks on buried pipe.

Second, even without insulation (which I would never place on a buried line since the ground is a very effective insulator), the temperature of the pipe remains very constant with regard to both time and position. It isn't like a pipe rack where you always have uneven heating and cooling. Stresses due to temperature changes also calculate out to be minimal for a "straight" run of buried pipe.

There are places on a pipeline where stress is a consideration, but they are almost always concentrated loads (a drip on the line for example) or a transition into the light (a dog leg or a pig launcher/receiver), but usually these forces can be reasonably constrained without any exotic steps.

David

Buried pipelines will throughout much of their length be fully restrained against movement due to soil friction forces. However, it takes a certain length / distance to develop sufficient soil friction to restrain the line so they will have areas where they are "unrestrained" and will move. These areas will be at the terminal points of the pipeline where it comes out of the ground and at locations where the pipeline takes a significant change in direction (because the soil restraint is no longer inline with the expansion force).

Normally, you would determine the tightest bend that the line could tolerate and still achieve full restraint and you would try to make sure that any overbends, underbends, and side bends were less tight that this. That would achieve full restraint through the buried portion of the pipeline. At the terminal points of the pipeline, or at locations where you have bends tighter than permitted to achieve full restraint, you have two options to deal with the situation:

1) You can provide a soft backfill (e.g., wood chips, peat moss, sawdust, Dow Ethafoam, etc.) material and allow the pipe to move underground (the underground equivalent to an expansion loop), or

2) you can install anchor blocks just prior to the bend or terminal point to act as rigid restraints to prevent movement.

Note: at the terminal points, if you are not going to install an anchor block then you need to install an offset so that the pipe has the opportunity to bend underground so that you don't impose a large force into the riser assembly.

With regards to insulation, having dealt with pipelines where it is critical to maintain temperature to prevent potentially serious accidents (e.g., hydrates in sour gas pipelines), I'd have to disagree and say the ground is not a very effective insulator for a hot pipeline. It is extremely common to provide insulation on buried pipelines and there are different products designed for this available from different vendors. Do a quick google search for one of the common ones, Shawcor Insul-8.

All this can be extremely critical on a hot line and one rule of thumb is that if the operating temperature differs from the installation temperature by more than 25 C, a detailed pipeline stress analysis complete with calculations on bend radius, soil restraint, anchor blocks, offsets, etc. is required. For pipelines that operate near at closer to installation temperatures often do not receive a detailed stress analysis as suggested by ZDAS04. At the temperatures you've quoted, your's is definitely a hot line that warrants detailed study by an experience and qualified pipeline stress engineer.

There is some good but old reading on this in an old article that I've uploaded.


ttp://files.engineering.com/getfile.aspx?folder=f95b2f8c-e8b8-43df-a182-a771d497affc&file=Part_2_-_Stress_Analysis_Paper.pdf

I can't add too much to zdas' and rneil's excellent comments and rneil was even good enough to post Peng's paper, which I started using the day it was published and have never stopped referring to it since then. I will try to add what little else I can.

Soil can set up restraints, virtual anchors, through friction and cohesion when sufficient contact length is possible. I've found that it generally takes between 500 and 1000 feet of buried pipe length to build up sufficient anchor force to restrain most pipe sizes, but it does vary with diameter and temperature, so be especially carefull where reverse angle changes in direction are at short intervals. Note that these anchors occur in both horizontal alignment and in the vertical profile too. At horizontal changes in direction pipe is forced into the sidewall of the trench and moves upward. In the profile, pipe is anchored at sag bends and tends to move upward into backfilled soil above the overbend areas. Where movements are extreme and soil cover does not provide enough weight to hold the pipe down at these bends, pipe can rise through the surface and become quite an embarrassment, so they are esp desireable to avoid. What I've done in the past is make an installation table, first taking the standard burial depth and then finding out what maximum change in direction can be made without overstress, then doing the same for increasing degrees of bend showing an increasing burial depth needed to prevent liftoff, until the line is reaching stress limits. Above that, special flexibility methods will be required, such as foam blocks or whatever, wrapped in geotextiles to try to prevent soil intrusion and compaction, etc. Be carefull where you place these soft areas, since they might make excellent farm tractor traps and you probably will want to fence them to prevent entry. Sometimes these tables begin at 7 to 8 degrees of overbend, which isn't really very much bend until extra soil depth becomes necessary. You can reach a maximum depth where soil forces on an unpressurized line can cause too much ovaling in thin pipe walls.

In regard to setting up virtual anchors, its important to have good values for cohesion from tests on soil samples taken at frequent intervals along the route. Designate some extra samples at points of large changes in direction. The samples should be from whatever material is to be used as backfill, either natural soil from the trench, or local sand backfill sources, if natural soils are not desireable. Friction and cohesion tests should be conducted on remolded specimens.

I am somewhat distrusting of providing underground flexibility, since I think that after a relatively small number of temperature cycles, much flexibility can be lost to soil sag and recompaction, so at places near valve, pig launching stations, and pump stations, I put a premium on even slight above ground flexibility. Be very careful with anchor locations, hot lines need BIG anchors and those introduce BIG axial stresses.

Between everyone's comments above and Pengs paper, the only thing I may be able to add is related to the thermodynamics. As zdas mentioned, soil is a relatively good insulator and I've run hot crude pipelines for 150 km from 195F inlet temperatures and only dropped 75 degrees or less, about 1/2 deg F per km, so I'm quite surprized you need insulation at a length of 10 miles, but given you may have lesser flowrates that would not maintain a high soil temperature, a colder soil temperature, or perhaps a product that condenses, has wax, or hydrates, or some other sensitivity, I won't say any more than that in regard to providing insulation, or not.

Last, be careful with the startup and shutdown of hot lines. In many cases a product's viscosity, such as heavy crude, can be very sensitive to temperature changes and required inlet pressures at lower startup temperatures can be significantly higher. I've even had to change pump configurations on the fly from series to parallel as the soil around the pipeline finally heated up and reduced the crude's viscosity to a point where we were able to reach 4 and 5 times the initial flowrate. You might want to conduct a "restart time to full flowrate" after allowing for various shutdown and line cool-off times. Maybe for something like shutdowns of 1,2,3,5,7,10,14, 30 day or even longer periods.

Have a good project.


**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

As noticed by rmv, thermal expansion occurs;

dL = alfa* L* dT

dL = change in length, in
alfa = coefficient of thermal expansion, 1/°F
L = initial length, in
dT = temperature change, °F

Now, a not criminal approach suggests a preliminary geological analysis to check the feasibility of the project and get the evidence the soil is competent for such an installation. Usually the temperature of the fluid carried by buried pipes is comparable to the surrounding soil temperature and so there is a little or no driving force to determine thermal expansion. When the fluid has a temperature which can provoke thermal expansion, and it could be considered fully-restrained by surrounding soil, each straight portion of the pipeline will be subjected to an axial stress. In correspondence of bends or other fitting, and because of the finite soil stiffness there is a relative movement between pipe and soil and so the pipe will be subjected to both bending stresses and axial stress.

11echo:

I am not sure what Code Of Construction you are working to, but I have experienced some issues arising from literal interpretation of the CSA Z662 Code. I have found that at a differential of approximately 40 C between installation temperature and hot operating temperature, it will be challenging for you to have a computer model "pass" a stress check, regardless of what soil you have / assume / use. The reason is that when you algebraically manipulate the equations for allowable combined longitudinal and hoop stresses, they begin to govern in the specification of minimum wall thickness. With me, that has given rise to the following rule of thumb:

When [Th-Ti] > SMYS/6, with Tx in C and SMYS in MPa, you need a detailed stress analysis to achieve proper design. This isn't quite the number (25 C) put up by rneill - which is a good number - but it is approximately what falls out of the equations when considering a sour line in a Class Location 1.

Further, when you state:

"...and dealt with the expansion once the line made the transition to above ground."

I am surprised if that has not created considerable problems in the past. At risers, pipe, once expanded and cooled, never seems to sit back down in the same location. After a few cycles, sometimes it just walks its way off the supports altogether and flops around in the free air on the lease adjacent to - but not "on" - the rack.

The posts by the strong pipeline folks here (zdas04, BigInch, rneill) are, as usual, such that I can't add much of value to the information therein. I will say that it is my (unfortunate) experience that Clients typically *do not* do a geotechnical evaluation along the ROW until they are absolutely forced into it, so most "analyses" are based on the engineer looking at the survey plans and making some kind of judgement (e.g., "...plan says "bog" here...so soil is probably just like muskeg, so from here to here assume a low undrained shear strength..."), and this is a bogus approach driven by unsophisticated Clients who place little or no value on the correctness of the engineering. (Not that I am in any way bitter.)

I even once had a Client who wanted me to analyze 10 km of a hot (75 C) line, installed in Northern Alberta in January, with the constraints (1) do not use anchor points; (2) do not use expansion loops or Z-bends; (3) we cannot tolerate any movement whatsoever at risers above grade. Fortunately, in that instance, I was able to get someone more senior on the Client's side who was able to issue some revised constraints to work with.

Regards,

SNORGY.

What sort of insulation do you use that allows the friction from the soil to function as a restraint? Does the insulation require special cladding to avoid being crushed by the backfill?

Katmar. Great point! that we all forgot to mention. If there's coating or insulation, shear transfer would be limited to some value less than the shear stress of the coating and/or insulation. A good reason not use insulation and rely on earth only, or double wall the pipe with the insulation located in the annulus, or locate the insulated pipe in a box conduit.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

There are insulating poured-in-place techniques which do not foresee cladding. Powdered calcium carbonate, as well as perlite, are used as insulating materials in buried pipeline. The insulating medium is poured and compacted into containing forms placed in the trench.

HEY want to thank everyone for their responses, I'm finding it very interesting. Some/most of this "stuff" is over my head, but on occasion I can get my fingers on a few ideas, an I appreciate the opportunity to learn!
I wanted to submit abit more data that's been requested. The pipeline is built under B31.4 code. The insulation is a foam with a thick HDPE jacket. One of the main "issues" is delivery temp.; it can't be below the 140 Deg. mark. The commodity is heavy crude, so long down times (hours) would be an "issue".

My experience with hot underground piping (or "warm" if you consider 140 to 195 F not that hot) is that you install guides, expansions joints, anchors, loops, etc. just as you would above ground piping. I would insulate and heat trace it, especially for crude oil or other heavy oils or asphault, otherwise it will freeze up before it makes it to the destination, plugging the line, meaning the old line is abandoned in place and a new one installed properly next to it - you probably can find several examples at any oil refinery tank farm.

Try "Foamglas" insulation, expensive, but on a properly installed system, no corrosion on the pipe exterior, can accomodate expansion. I've put replacement systems in the ground that have lasted for over 20 years and didn't need further work, replaceing the previous systems (including some pour in place loose type insulation systems) that required significant repair/replacement every five years.

I've worked with steam, condensate return, hot water, black oil, and heavy oil undrground piping systems for refineries, heavy industrial sites, campuses, and chemical plants. Do the calculations for expansion, find the expected forces on the anchors and guides, see how much the soil will take, if necessary put in more anchors and expansions joints for lower forces at weaker soils, insulate and don't be cheap or try use lesser materials to save money. A bad break will be a catastrophic event that will blow away all the savings (again, my experience).

Good Luck!

In 35 years of pipelining, I have seen 1 incident where there was a failure because of thermal stresses on a pipeline. We laid a line that tied in to a mainline. The line was less than a mile long and on a huge Texas ranch.The line was tested and tied into through a hot tap.We did not cover the line in the hot summer sun. That night a cold rain hit and the line went from 130 to 60 in a matter of minutes and a weld cracked at the tie im. When we disconnected the new line it jumped back 3 inches from the tap valve. Upon inspection, the crew didn't follow proceedures and let the line snake back and forth in the ditch, they laid it straight and true like they do pipes on piperacks in plants.

Another resource is:

Excerpt taken from the website:

Design Guideline for Buried Steel Pipe

This guideline presents design provisions for use in evaluating the integrity of buried pipelines for a range of applied loads. Both new and existing welded buried pipe of carbon or alloy steel fabricated to ASTM or API material specifications and constructed in accordance with ASME B31 pressure piping codes are considered. The following load conditions are addressed: internal pressure, vertical earth loads, surface live loads, surface impact loads, buoyancy, thermal expansion, relative pipe-soil displacement, movement at pipe bends, mine subsidence, earthquake ground motion, effect of nearby blasting, fluid transients, and in-service relocation. The ASME B31 Guideline Committee currently is considering integrating the ALA-developed guidance into its standard.

In some work I have done using preinsulated ABS for district cooling in Dubai it was determined that the thermal movement of the pipe did indeed crush the insulation at changes of direction. In the scheme of things this did not affect the heat transfer overall as the percentage of damaged insultion was very small compared to the overall. Pipe overstress was not a problem.

District heating and cooling industries have had this matter to consider all their lives. The thermoplastic pipe industry also needs to consider these aspects particulalry where the likes of PE has a coefficient of thermal expansion 20 times that of carbon steel and much lower strength. So in your Google searches look at other industries and how they handle the movement or potential movement.

As long as we're on the subject of coatings, be especially careful with selection of coatings, as many start breaking down at 130F. FBE will work to your temperatures.

**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch

Installation temperature and other factors can also affect at least the magnitude of such eventual effects (notice the remarkable incident in Russia noted toward the end of past similar thread

here is another case study for you.

We had a ethylene pipeline operating at 2100 psi. A (still alive) person decided to set a new powerline pole and drilled a hole into the 8" line. The 20 segment was shut after about 15 minutes when the pipeline control system flagged a pressure rate of change. The 20 mile segment was isolated and the line depressured to zero psig in a few hours.

If you look at the PH chart you will see the line changed temperature by about 200 degrees F. It took 2 days to defrost the soil enough to cut and replace a 10 foot section. We pressured the line up and were back on line.