Topic Options
#141 - 08/03/00 08:50 AM Replacing Cold Spring in an Existing System
Richard Havard Offline

Registered: 12/16/99
Posts: 58
Loc: Houston, TX, USA
I have a problem that is replacing part of an existing heater line. The section that is being replaced contained cold spring. I do not have any experience with cold spring, and I have a pretty specific question. Please bear with me as I try to set up the scenario.

The line begins at a heater where it ties 4 outlet lines together, runs south 15', east 38', south 50', west 20', and up a tower 20' to the nozzle. Two feet after the bend heading into the 50' run south, there is a support with an axial restraint on the south side of the shoe only. Approximately 8' further south, the line increases in size from 8" to 10" at an 8x10 reducer followed immediately by a flange pair. At the weld point of the downstream flange, there is 2.5" of cold spring specified. There are no other restraints on this line, according to the ISO and field notes, until it gets to the tower nozzle. There is a note on the ISO regarding installation of the axial restraint that states the bumper is to be welded in place before pulling the 2.5" of cold spring. This will effectively give a gap of between 0" and 2.5" on this restraint.

It is proposed to replace the line after the reducer to a point approximately 25' south, eliminating the previously mentioned flanges. Should the cold spring be put back in as specified since this is an existing system? I have my doubts that this was done properly in the first place, and, considering the seemingly unpredictable nature of the method used, am leary of duplicating this system.

Thanks in advance.

Richard P. Havard
Mechanical Engineer
Mustang Engineering
Richard Havard, P.E.
Piping Engineer

#142 - 08/03/00 09:18 AM Re: Replacing Cold Spring in an Existing System
Jim Wilcox Offline

Registered: 12/20/99
Posts: 46
Loc: Calgary, AB, Canada
Richard, I am by no means a cold spring expert, but my approach would be to model the system in CAESAR II, first without the cold spring, and evaluate the loads to see why the original designer wanted to employ it. Once you know that, you can consider other solutions (presuming there was a problem). If you conclude that cold spring is a preferred solution, you can then decide how much is appropriate for the new piping configuration. Burn a few runs and you'll figure it out.

#143 - 08/03/00 11:44 AM Re: Replacing Cold Spring in an Existing System
John Breen Offline

Registered: 03/09/00
Posts: 482
Loc: Pittsburgh, PA (& Texas)
Hello Richard,

You said:

“ I have my doubts that this was done properly in the first place, and, considering the seemingly unpredictable nature of the method used, am leery of duplicating this system.”

Well, so far you have done some things right. You have acquired a respectful skepticism, and you asked for informed opinions (I hope you get several). You recognize that “cold springing” should not be employed by an engineer who is not experienced in its consequences – that is a healthy realization.

There is a cliché that is often used in football: “when you throw a forward pass, there are 3 possible consequences – only 1 of those is good”. A similar comment can be made about “cold springing” piping systems – however, the odds are better for a negative outcome.

Let’s define terms. “Cold springing” is the practice of intentionally fabricating a piping system such that the pipe spools are made shorter than the nominal design dimensions (“cut short” by some frraction of the expected expansion) or they are made longer than the nominal design dimensions (“cut long” by some fraction of the expected contraction). “Cut short” would be employed in “hot “ systems and “cut long” would be employed in cryogenic (“cold”) systems. This practice creates a prestressed condition (hopefully in the desired direction) in the piping, and a preload condition (again, hopefully in the desired direction) at the attachment point of the piping to "strain sensitive" equipment. What do we mean by “strain sensitive” equipment. Well, there are practical limits on the loadings that can be accommodated by rotating equipment (pumps, turbines, etc.), fin-tube coolers, pressure vessels, and other types of structures. The API specifications for pumps and turbines, for example, tell the equipment manufacturers how strong they must build their equipment by defining requirements for the minimum piping loads that they must be able to accommodate. If the piping designer limits (by good piping design) the calculated forces and moments transferred to the pump or turbine such that they are less than the limits prescribed by the Standard, we should expect acceptable bearing life.

What are we trying to achieve when we “cold spring” a piping system? Not a reduction in pipe stresses. You will notice that the ANSI/ASME B31 Piping Codes will not allow cold spring to be considered as reducing the calculated stresses in the pipe. The advantage of a “well executed cold spring” (which in the opinion of many of my compeers, is a contradiction in terms) is to, at least (or most) theoretically, reduce the peak loading (forces and moments) on strain sensitive “terminal equipment”. That is to say, the (theoretical) effect of the “cold spring” is, on the initial temperature excursion, to effect a reduction in the peak loadings transferred to the equipment by the piping.

But let us consider the reality of construction. In most construction projects it will not be possible to obtain construction tolerances of plus or minus ¼ inch – more likely 3/8 to ½ inch. This implies that the designer should never ask for “cold springing” with dimensions less that ½ inch. It also implies that the designer should analyze what would happen to the equipment if greater or lesser amounts of “cold spring” were, in fact, built into the system (there is a real possibility of that). Now consider what must be done to create a “cold spring” in a system. When all the welds except the “final closure weld” are made, the piping contractor will be presented with a situation in which he must “pull” the gap together and do a “fit-up” prior to making the weld. If the contractor has enough chain hoists and “come-alongs” he might be able to pull the 2 pipe ends fairly close together. But then he must turn the 2 ends of the pipe (apply one or more moments locally) in order to get the pipe ends parallel for the proper “fit-up” – this is virtually impossible to do completely. The implication is that this is a relatively inflexible system - why else would we be trying to “cold spring” it(?). The fact is that it is virtually impossible to get the “cold spring” right. There has to be several stout structures close at hand to attach the chain hoists to and there has to be “elbow room” – and usually there is not. So the theoretical “cold spring” that was defined by the designer really is not perfectly applied as specified. If “cold springing” is “under-done” we do not get the total beneficial effect that we calculated. If “cold springing” is “over-done” we again do not get the beneficial effect that we calculated and it is likely that we have more “nozzle load” that we would have if we did not attempt the “cold spring” (can you see the football cliché manifesting itself (?)). Remember that the loading limits are the same in any direction so you cannot effect too much prestress or you will still overload the equipment. I think that your “doubts that it was done properly….” are well founded. Also, the accuracy of the gaps in line stops, limit stops and guides are also susceptible to the vagaries of construction. Also, since this system will go through another "shake down" cycle, those line stop (etc.) gaps should be examined in a couple of years to see if they still exist (adjustments might be necessary). Basically, “cold spring” designs are wishful thinking! I personnaly no not believe that there was ever a piping system built that did not include inadvertant "cold spring".

The “cold springing” of piping systems has generally fallen out of favor in recent years. It was very common in the design of power piping systems up to about 20 to 25 years ago. Remember power piping is large diameter pipe with very heavy walls – it is VERY stiff. Also, computer analyses used to be labor intensive and expensive (try to picture how hard it was to do proper modeling with no instant graphics). If we were offered a “panacea” – an instant cure which would preclude exhaustive reanalyses– we took it and we said, “thank you very much”. I think we now have a better understanding of the limited benefits of “cold springing” and the potential for negative results in the “real world”. It is an interesting situation really. COADE has provided us with wonderful tools to model process systems and we are getting closer to being able to calculate stresses, forces, moments, and deflections much more accurately. Instead of using these tools to model conditions that are unlikely to be created during construction, we can use them to find the extra flexibility that we need to control (or limit) the forces and moments applied to the terminal equipment in a way that is consistent with acceptable levels of stress in the pipe. If you are going to specify “cold springing” be sure that there will be adequate “field surveillance” provided to get it as close to the design as possible.

Some things must be remembered when modeling “cold springs”. Think about your load combinations and what parameters must be in effect – and which must be changed. When you are calculating pipe stresses, it is appropriate to use the cold (“as installed”) young’s modulus and to not apply the “cold spring”. It is conservative to omit the flexibility of the terminal equipment (e.g., vessel nozzles) in static pipe stress analyses.

When you have determined that the geometry (and support scheme, etc., etc.) will result in enough flexibility to satisfy Code stress requirements you are ready to analyze the “cold spring” effect – a very different set of load cases. The piping system must be analyzed both hot and cold to be sure that the “cold spring” won't be overloading the equipment in either condition. Be careful to properly model all your “boundary conditions” – watch the total travel of the spring hangers. It is appropriate to use the “at temperature” young’s modulus and include terminal equipment flexibilities in the “cold spring” analyses. Please recognize that any pipe stresses calculated in these analyses will not be "Code valid". By all means, reanalyze the entire system.

Bottom line. Today, when we have CAESAR II available, we can be more confident that we will have a reasonable degree of accuracy in our model and that we will have a good understanding of the response (to loadings) of our final “as constructed” system. Take a deep breath and try to make the system work without “cold spring”. Trade-off some “surplus” allowable pipe stress to create a geometry that is “kinder” to the terminal equipment. It might take a few more computer “runs” but it will be cheaper than trying to “enforce” an impractical “cold springing” scheme (some will say that they are all impractical). Remember, in a hot system the pipe will (over a period of time) “self spring” itself as it “relaxes” (elastic-plastic deformation of bends and elbows) and the result will be similar to “cold spring”. (So, don’t let anyone “fix” or “adjust” the piping system when a pump is replaced and the piping “jumps away from the pump nozzle when the flange bolts are removed.) The judicious “tweaking” of spring hangers can often help, but this is an art – you must be prepared to put your hands on the support hardware and “fine-tune” it yourself.

Good luck with your project. Best regards, John.

[This message has been edited by John Breen (edited August 07, 2000).]
John Breen

#144 - 08/03/00 08:03 PM Re: Replacing Cold Spring in an Existing System

Well it appears I'm late but I have a couple of things to add...

First cold spring if performed properly on a system operating below the creep range is beneficial. How so you ask? It allows a quicker "shakedown" to occur.

However I agree with the previous replies centering on the phrase "if performed properly".

I am kind of an old guy and I remember well the heady days of punch cards and green bar paper. In those days we might "cold spring" once in a blue moon. The few times I resorted to this stress range, slight of hand I did so on systems where between the two end anchors there was a loop available.

The advantage was that you could pull back on the pipe end using a pair of sturdy guides that would be left in place to help keep things straight. The other advantage was that rotations of the ends was minimzed.

Did I think it worked then and what do I think about it now?

Back then I doubted that I would get everything I needed in load reduction so I always left some room (25% off either + or - on the springing) for construction vagaries.

What about now a days? I currently wouldn't bet a nickle for sucess in this type of enterprise. Recently a bone head constructor decided that the tie bars I had specified for a set of Exp Jts were superfluos without asking me my opinion. Fortuanately they were going to hydro the steam system that the Exp Jts were in and nobody got hurt, when they became hyperextended. So if they ignore specifications written in english about what Exp Jt to buy what do you think your chances are? (See the pictures below of the ExpJts after the hydro.)

<font size="4" color="0000ff">Got Tie Rods?</font>

Note the yellow shipping bars were originally welded to the spool below. Internal pressure caused 8+ inches of hyperextension.

So there you go, if you do take the cold spring route design the system to make springing the ends together easier (reduce rotation reactions as much as possible). If you try it and succeed email me I would like your assistance in picking some lottery numbers!

Best Regards,

John C. Luf

[This message has been edited by John C. Luf (edited August 09, 2000).]

#145 - 08/08/00 12:14 PM Re: Replacing Cold Spring in an Existing System
KS Taylor Offline

Registered: 06/06/00
Posts: 3
Loc: Martinez, CA, USA
Over the course of my piping flexibility career (I started in the "heady days of punch cards & green lined paper"), I have developed a few realizations regarding cold spring:

1. Cold spring is a "paper solution" when applied to piping loads on equipment. It is hard to predict where the loads will actually be after a few cycles.

2. Don't even think of using cold spring to reduce stresses in the piping system.

3. The only remaining application where cold spring is defensible is in controlling deflections. Here I am talking about controlling large deflections (>>1") such as occur in pipe racks and pipeways. This can be reliably measured in the field when creating the cold spring where as a specified starting stress level is difficult if not impossible to actually measure.

Deflection control (i.e. creating enough room to allow piping to move without interferring with other piping in the rack) is the only cold spring application I would consider.


Moderator:  Dave Diehl, Richard Ay 
Who's Online
0 registered (), 36 Guests and 2 Spiders online.
Key: Admin, Global Mod, Mod
Su M Tu W Th F Sa
1 2 3 4 5
6 7 8 9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30
Forum Stats
11803 Members
14 Forums
16324 Topics
72947 Posts

Max Online: 303 @ 01/28/20 11:58 PM
Top Posters (30 Days)
Vanman 10
Kellyp1966 9
treysoucie 6
Michael_Fletcher 6
jwood 5