Assume a long straight run of a HP/HT steam line requiring multiple expansion loops. Some things being considered for thermal expansion:
-location of supports, guides, limit stops, and/or anchors
-loads on supports, guides, limit stops, and/or anchors
-calculated stresses vs. allowable stresses
-long run could be on sleepers or pipe rack
-minimizing costs, i.e., expansion loops

Assuming stress allowables are satisfied and loads are tolerable, are there any other considerations? Particularly regarding maximum reasonable displacements that fall more into a best engineering practice as opposed to strictly code compliance?

thanks

Make sure your shoes are long enough (grin). We have some clients who insist on having guides at regular intervals I think in part to assure that long term the pipe stays where it's supposed to be. Having just worked on a steam system where the edge of the shoe was 1-1/2" from the support the shoe was supposed to be sitting on, I can see the value in this.... in some cases the client also has shoe standards that say "for XXX movement the shoe to be YYYY long"...

-Shannon

thanks,
we can use shoes 24" long, but i was wondering as a general rule, if it is practical to have axial displacements more than say 6" or less or more? i've searched all the literature available to me and have found no guidelines, not even client specs. i do know that if the shoe falls off the support steel, the longer the pipe, the farther away you want to be when it is pulled back!

Instead of "long shoes" or huge stanchion base plate.... put a piece of supplemental steel below the support location down the long axis of the pipe.... put a single solid plate across it and the cross beam to let the bottom support slide on and don't forget some high temperature grease!


Also cold setting 1/2 the movement is commonly done...

In general, it is accepted that axial movement be restricted to 6" as the standard shoe length most often is 12". Another consideration limiting the axial expansion is the spacing of pipes on racks and sleepers. If you can make sure that there is sufficient space between pipes close to loops or turns (in hot condition) you could go beyond the 6" movement.

regards,
Harish.

Ah ha!

Ive done this so i know.(I think)
As usually, on a rack system, stress is not the concern, but expansion and restraint forces are.
However, you mention steam here. The important think with steam is the distance between the dirt legs, or how long the condensate can travel before it is drained. On our last job, Spirax sarco recommended 35m between dirt pockets so as to minimise the risk of the condensate slug becoming too large or picking up too much speed. This figure between the pockets worked well for the distance between expansion loops as the expansion, stresses, forces and width of loop matched a 35m pitch pretty well. Therefore a tee was added to the front of each steam expansion loop and the system works bloody well!

my twopeneth.

Just completed engineering evaluation of over 30 year old P1, 42" steam line 1.9 km from end to end. Loops are between 40 to 100 m apart and elevated. No springs. Line runs East West. Loops towards North (4 off 90 degr plus 2 45 elb). In each long section one X, Z and +Y plus 2 North South guides with 50 mm gaps! (One support away from X,Z,+Y). Each loop has 2 North South restraints with 5 mm gaps. Absolutely nothing to stop lifting. Substantial number are lifting due to elevation differences in calcs. On site none.

Each long section has 2 drain pots. Direction of flow changes from time to time.

Line is in still there and fully operational. Problems noted: Some supports may be sinking (Bad soil), excessive spans create small condensate pockets and teflon pads are failing.

Dave,

I've designed many above ground high pressure steam pipelines (600, 900, and 1500 class, up to 24 NPS). They've all worked so far. Usually they are made with high tensile pipe (Q&T with yield of 448MPa/65ksi or 550MPa/80kis)

Things to consider:

The guide spacing near the loop:
-Assuming you're using "z-loops", where the lateral offsets section of the loop is at 45°, make sure the distance to the first guide on either side is far enough away to allow the loop to pull the pipe in the direction of the loop. Otherwise you can huge guide loads.

-The distance to the next guide should be far enough to prevent the prying or binding effect that occurs if they are too close.
-Make sure the guides are close enough to control the movement of the pipe in a predictable way. It's a balancing act.

I've had shoes as long as 1.5m to handle the growth. The anchor-anchor spacings can be 400m+ and the loops can be 12m x 9m + depending on NPS etc. Often we also have shoes of varying length ie., 300mm, 600mm, 900mm, 1200mm, and 1500mm, and the also offset them in opposit direction of growth. Sometimes we center them to make sure construction can't install the offset on the wrong side. Depends on the project.

Waterhammer - for some reason during start-up many operators find a way to induce waterhammer (steam bubble trapped in sub-cooled condensate collapses). Consider beefed up guides and anchors with fabreeka pads to lower dynamic load factor. The forces are HUGE so do everything you can to prevent this phenomenon.

If you are looking at plant piping most of that is irrelevant to your case.

In plants I try to keep the loops spaced to limit the growth to 6in/150mm but with longer shoes and clash checks you can go longer. Try also to place the lines nearest the edge of the rack so the supports can be braced more effectively at the anchors. A general rule to remember is that if you anchors are centered between loops, to add a 25% out of balance factor to the friction forces. Caesar doesn't do this automatically. I either do a hand calc or fudge the friction factor (-25%) on one side of the anchor. Avoid putting anchors at the end of a run.

in response to Paul Bond.

1. Don't long shoes create large thermal gradients and hence crack and fall off.
2.Waterhammer in steam lines is usually due to the impact of a condensate slug on an elbow. the culmination of a bad layout and/or rapid start ups. (hence good frequency of dirt pockets in high velocity lines.


also, dont put a dirt trap in the middle of the steam runs. anyone who plays golf should figure.

The thermal gradient issue can be a major consideration at higher deltas.... Also the base material and its weldabilty can be an issue, go look at the P91 thread for an eye opener on that topic.

I find the idea of buying common plate and structural shapes more attractive than long shoes. In the end after judicious consideration there may be multiple good choices, and a rose by any other name etc.

If the pipe material is in creep range only support welds I would do to the pipe are small "shear lugs". External stresses or heating/cooling differences will crack other type of welds sooner or later. Shoes with clamps and lugs to hold axially will work for me in creep range.

Jouko.... yep!

Thanks to all who have responded so far (not to stop any additional valuable posts); all great posts!!! I tried to post question so more than just myself would benefit.
This particular application is in a chemical plant, but it is a large plant. Some ideas I had were confirmed and some posts brought up new issues to me; and this is great!
It is interesting the different variations described all working with success. I've also run into the slug/hammer issue: those usually fall into the "war story" catagory. - thanks again

Good points superpiper,

1. Yes you are right and the long shoes should have expansion slots. Most 'clients' have support standards showing this.

2. Yes, commonly that happens. It's commonely refered to as "slugging" and the forces are in order of rho*V^2. The event I'm more concernced about is the condesate hammer that occurs when steam condenses rapidily due sub-cooled condesate trapping it. The forces are in the order of rho*c*V, where c is the speed of sound (celerity) in the condensate. It's something like 100x the force from a slug hitting an elbow. We can design for the slugging phenomena easily enough, but the hammer phenomena needs to be prevented.

Check out www.kirsner.org. I just went to his seminar, it really opened my eyes to a lot of misunderstandings and lack of research done on the topic.

Ive learnt the hard way about condensate slugging and water hammer.
A while back i did an analysis on a steam line (the one mentioned above) and its condensate line. Now i'm no expert but i try to examine how the designers run these lines and correct any mistakes.
Because this line ran some 550m outside, the philosophy was to have the condensate expansion loops at every other steam expansion loop. From a stress point of view this works wonderfully, however, the raised loop in the condensate line (from the traps on the main steam line) meant the condensate header is constantly flooded.
This in itself would be ok if we'd reallised it at the time, but it went through the net, so that blast type traps were installed into the condensate system.
The water hammer this is causing in the field is allarming for a 1.5" line.
This rack is a very substantial structure and you can feel the energy realesed when you stand on it!!! the whole rack shudders.
I've made loud noises about replacing the traps for thermodynamics and adding dampers on the condensate lines, but to no avail.
very soon i expect to hear about pipe shoes falling off!!.

This though has a second story. i am paranoid about keeping stress critical line well supported. Most stress engineers seem to be alergic to hold down type supports. for me, hold down guides ensure a well supported and stable system. i digress,
Because the condensate line was one of my first analysis' i maybe was a little enthusiastic about the quality and quantity of supports.
However, it is my belief that my attitude to supporting this line is the only reason that it has'nt fallen down and hurt someone. I'm not sure what the moral is here, but i still like to have stress critical pipework well supported in ALL axis of freadom.

Tim

Well actually in general clamping lines even thermally active lines down in X,Y,and Z now an then is a good thing... it providses a certain degree of stabilty to the system.

Lisega even makes a holddown restraint for said purpose so more supports are indeed better and clamoing a line down here or there is good.

One other matter to consider in this discussion.... very hot lines may require more supports than your analysis indicates if sag is critical (B31.1 uses a 0.1" criteria). Why do I say this??? We use the cold modulus of elasticity for our code stress calculations and displacements are calculated on this basis but as E decreases the sag will increase so that if you think you have a 0.1" sag you may have much more than that!

On high temperature steam lines (creep range) calculation may show low stresses but sagging causes low points if the overal slope is small. During heating up of the line any point where there is condensate pipe wall will remain cooler. Result is that the top of the pipe has lower strength but higher thermal expansion than bottom and the sagging just increases step by step. At the end you have a serious banana and pipe has to be cut.

More supports and check slope and drainage in all conditions.

I have seen far less problems on plants designed using slide ruler compared to computer design

Spielvogel still rules.... I have a set of Spielvogel solutions I use as benchmarks when testing prerelease versions of CAESAR II....

These days people are computer stupid.... last week I had an owner ask me why I hadn't performed a "Stress Analysis" (Meaning CAESAR II model) of his relatively flexible, well supported system? Doesn't B31.3 require this he asked....???

He was shocked when I told him that the code did state any thing specifically concerning computer analysis, and that basically it was up to the judgement of the designer!

So there you go we are computer literate and maybe computer stupid as well....