How to deal with pipe D/t =160? 2

[DavidCR]

We deal with the design of pipe systems for steam and brine (geothermal gathering system) following B31.1 (we use Coade CaesarII to design) we have new applications that use bigger diameters, our worst cases are 60" diam, with a thickness defined by internal presure of 3/8" (D/t=160) and other of D:50" t:5/16" (D/t=160) with lenghts of the order of 300 ft or more installed in cross country.

Our design condition is internal pressure 131psig (9 barg) and 356°F (180°C) and we plan to use ASTM A515 Gr.70 plate.

We are looking for criteria about such dimensiones and how to define when D/t ratios start to need special considerations such as stiffner rings, use bigger thickness, etc., considering high loads on supports and high loads on "Ts" and "Ys" could be spected.

So I´d appreciate if somebody can give some hints on where to find criteria on this or somebody has been related with similar cases and if he has used special considerations, or anny comment on this.

Thanks a lot in advance, and blessings to you all from Costa Rica.

 

[MJCronin]

dave,

Based on your extreme D/t ratios, I would say that you are on the edge of what is defined as "piping". Based on what I have found, most piping systems are in a D/t range of less than 80....ASME B31.1/B31.3 cautions against using stress intesification factors in systems with D/t>100

It sounds like you are closer to the rules that define round ductwork.

Consider:"The Structural Design of Air and Gas Ducts for Power Stations and Industrial Boiler Applications" by the Air and Gas Duct Structural Design Committee of the Energy Division of the American Society of Civil Engineers (ISBN No 0-7844-0112-8).

http://www.amazon.com/Structural-Design-Stations-Industrial-Applications/dp/0784401128

This is an American Society of Civil Engineers Special Publication not a guidleine for piping design.

Your fundamental issue is, of course, how to design, size and space stiffeners.

Good luck

-MJC

 

[doct9960]

David,

Your pipe system is in steam service, right? Is there no vacuum (external pressure) condition? If there is, then you would definitely need stiffeners or thicker plate, or both; and you would need to follow the procedures in UG-28, UG-29 and UG-30 of ASME VIII Div. 1.

 

[JohnBreen]

DavidCR

Doct9960's point is well taken, either provide stiffening rings per the cited ASME VIII paragraphs or provide vacuum breakers in any section that can be isolated by, e.g., valves. Remember that with steam, condensation upon cooling WILL result in atmospheric pressure external loadings. Also, stiffening rings are often provided to extend the process pipes ability to span longer lengths between supports (they assure the pipe stays "round". In a bending failure, the pipe will have to "ovalize" before a plastic hinge can be formed so keeping the pipe "round" is important.

As Mike pointed out, you are on the "hairy edge" of what can be done in the way of structural analyses using Caesar II and ASME B31 Pressure Piping rules. We provide a note in Appendices "A" of the various B31 Codes that warns that the Stress Intensification Factors (SIF) provided by the Code cannot be applied universally to D/t ratios greater than 100. So, that is another reason why stresses calculated by beam theory will not be accurate.

You can do a general structural analysis using Caesar II (which employs beam theory) to get an approximation of the distribution of forces and moments (keep in mind that you will have to use the "pressure stiffening" option at the bends and "elbows" at it greatly affects systems with these D/t ratios) but you must be aware that beam theory will not necessary predict the highest stresses - especially at "boundary locations". You will likely have to use the Finite Element Approach (FEA) at least to calculate the stresses local to welded appurtenances and supports (Saddles). So, you should look at such books as the (cited by Mike) ASCE duct manual and also the AWWA M-11 (for ideas regarding the design of saddle supports). Also, the Lincoln Arc Welding Foundation's books by Omar Blodgett provide ideas on the design of saddles for large diameter piping.

I would sugggest that you have a look at FE / Pipe by Paulin Research Group - go here:

http://www.prg-software.com/FEPipe.aspx

You might model your entire piping system with FE / Pipe or you might use a combination of Caesar II (as described above) in concert with analyses of stress at such local areas as supports and welded appurtenances.

A horizontal run of pipe with D/t > 100 is analogous to a horizontal pressure vessel or horizontal tank. The stresses local to the saddle supports must be carefully evaluated. If you can find a copy of "Process Equipment Design" by Brownell and Young you will see an approach for evaluating these stresses. Also you could look for (google) the papers by Zick (CB&I Inc.) to provide yourself with a background of the stresses that must be evaluated.

You might find (by structural analyses) that you will have to provide reinforcement (reinforcing "pads", "doubler plates", etc.) at areas of your pressure boundary (pipe wall) where welded supports (trunnions, base elbows, dummy legs. etc.) are causing local pipe wall bending or "punching shear". It is prudent to remember to "follow the load path" all the way to the foundations (where the Civil Engineer takes over the responsibility) as failures of these components (due to unanticipated loadings) can have serious consequences.

Do not overlook the possibility that in a geothermal steam system, piping that is designed for steam might on occasion become filled with condensate or brine. Obviously the additional weight of the pipe contents under these circumstances becomes a design issue.

Regards, John.

 

[DavidCR]

Thanks a lot.

We have no vacum.
In our case geothermal brine could develop pulse flow that may lead to ciclic forces.

You are giving good points and looking forward to more comments.

 

[zerosum]

DavidCR,

You don't have vacuum until the first process upset that allows that hot water to flash to steam and then condense.
I would certainly consider at least a partial vacuum scenario. There sure were a lot of collapsed vessels around the CPI before we started insisting on full vacuum design for anything anywhere near steam or other condensibles.

 

[DavidCR]

Yes I agree, some vacuum effect should be considered.

You give us some homework.

As a first step we are going to consider a Zick analysis and analysis as it were a vessel to compare with the results from ASME VIII and B31.1 due to internal pressure and other loads tipically considered by pressure vesel design software. And checking the referencies you mention.

Also the design of pipes near to Ts and Ys may include some thckness increments and FEA analysys. Also the FEA can be included in the design of the suports.

The aspect that I´m not sure how we are going to treat is the vacum effect, and possible we may contract somebody that can help us on this topics. Since our geothermal brine is contaminated with particles, precipitates, some vacum braker device may not be very effective due to the high possibility that it may be clogged.

Thanks again. And any other idea that comes up to your mind please post it.

 

[zerosum]

Apparently you are going to fabricate this piping more or less in situ and therefore possibly have a rolling device that could roll T sections the hard-way. I would use Compress, or lacking an appropriate Code application program, do it by hand, calculate the spacing required for vacuum rings (start with, say, ST2x3.85) to support an infinitely long shell (use 300 ft) at full vacuum with your diameter and thickness. Wild guess, based on distillation columns, every 20 feet or so. You will probably sleep better with a full vacuum design.

Is this considered power piping???

 

[DavidCR]

Yes it is power piping (B31.1).

Thanks for the hints on the vacum rings.

Briefly, geothermal brine is obtained from wells, and pipes gathers the fluids to ciclonic separators and then steam is piped to the power house, waste water is piped by gravity to be reinjected to other wells.

Pipe is made on shop and installed on site. We will have to use Ys for main branch connections and probably Ts will not be needed. I´m not sure what you mean with rooling T´s the hard way, but in general our Y´s are the simple welded type; cut and weld on shop or in situ depending on the complexity of the job.

For branch connections we use Coade´s Codecalc/PVElite to help design.

 

[JohnBreen]

We will have to use Ys for main branch connections

Who is doing the pressure design of these "Y's"?

The standard reinforcement schemes used in ASME Codes for branch connection reinforcement are not applicable to "Y's". I would VERY strongly suggest that someone review the design. For background, look at the AWWA publication M-11, "Steel Pipe - A Guide for Design and Installation". Also, rigorous weld inspection of the "Y's" is prudent.

Yes, the Scope of B31.1 includes geothermal piping but it includes many warnings about large diameter (relatively) thin wall pipe. Suggest you look into FE/Pipe as your primary component design tool to be used in concert with Caesar II.

Regards, John

 

[zerosum]

<I´m not sure what you mean with rooling [rolling] T´s the hard way,>

I wasn't referring to T's in a branch connection sense (Tees), but rather to structural steel T-sections (1/2 of a wide flange structural shape, A-36 or A-992) which should be the most efficient structural member for reinforcing your very large D/t fabricated pipe, either from vacuum or non-pressure imposed loads.

Good luck, sounds like an ambitious design.

 

[JohnBreen]

DavidCR

Do not underestimate the level of effort needed to design the branch connections (e.g., "Y"s et. al.) for large diameter, thin wall, piping.

Go here:

http://www.nwpipe.com/technical_resources_manual.html

Fill out the information and get the design manual. LOOK CLOSELY at the photos of the methods used in the design manual for the reinforcement of branch connections and laterals. Note that these reinforcement schemes for large diameter and relatively thin wall pipe differ greatly from the reinforcement methodologies shown in the piping Codes. NOTE ALSO THAT THE SCHEMES SHOWN IN THE DESIGN MANUAL ARE FOR FAIRLY COOL PIPE WALLS. There are other design issues when the reinforcement conducts heat away from the pipe wall. However, I am showing you this so that you will understand that you may have to use "design by analysis" methods (as opposed to "design by rule" methods as shown in the Piping Codes) to design the "Y"s and laterals you want to fabricate.

Regards, John