Criteria for Selecting Design Temperature when not provided 2

[simon17]

Hello.

Is there a rule of thumb for choosing a design temperature when only an operating temperature is provided for analysis? On my current project, both design and operating pressure have been chosen, but only the operating temperature.

I remember hearing some years back about simply adding 50F, but most of my systems are below 200F and this seems too conservative.

For now, I have used 1.2xOperatingTemperature for my design case. Any suggestions?

Thanks,
Dmitriy

 

[EmmanuelTop]

Usually it is not that simple.

You also need to consider minimum and/or maximum ambient temperature, whichever is below/above your min/max design temperature for the process.

Good engineering practice dictates to look into all possible operating/shutdown/restart scenarios to see what could be the minimum/maximum temperature excursions. Depressurization of vessel which contains light liquid Hydrocarbons is one of the examples - you can get to very low temperatures after blowdowns.

There is a lot of information published on the net. Some more info at:

http://www.red-bag.com/engineering-...to-establish-mechanical-design-pressures.html

(Chapter 4)






Dejan IVANOVIC
Process Engineer, MSChE

 

[LittleInch]

As ever it depends on the situation and system.

Process Engineers tend to slap on a blanket 20C as a minimum on top of max operating, at least for process piping in plants. Max design temp in pipelines is often quite a bit closer as it makes a big difference, so 10C margin is common. Min design is often -10C , but I've seen zero - that one caused a bit of a problem....

Min design temp is often overlooked and only gets picked up in HAZOPs and such reviews, but a nominal -20C is fairly common unless you're in the artic or dealing with very high flash JT effects.

To me you need to see if the blanket approach takes ou over some sort of step limit for stress, type of material, reduction in yield stress (duplex is particularly bad) and then find some engineering reason to set a figure.

The argument that max pressure and max temperature can't exist a the same time might occur or might happen. Depends on your process.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[BigInch]

Check the process fluid conditions, boiling temperature of water, etc. May be no reason to go higher than that. If immersed in water, for example, 0-100C +/-10 could be pretty reasonable start. Check equipment ratings for max and min capability of heaters and chillers in the system. Check max and min ambient temperatures, wind chill and solar insolation.

 

[simon17]

Thank you for the tips. The red-bag.com link is helpful as well.

 

[EmmanuelTop]

Ambient temperature can be a big deal. In Middle East we have used +82 degC black bulb temperature for all non-sheltered equipment, even if the process temperature was max +30 degC. Arctic area is quite opposite. Depending on which codes you have to follow, you may end up with stainless steel (speaking from design temperature point of view, not from the perspective of material/process compatibility) where CS would normally be sufficient.

I also remember a project in Europe where ASME B31.8 was intended to be used for Dehydration unit piping, nested inside an existing compressor station. Design ambient temperature was -29 degC. While B31.8 was allowed to be used for compressor station piping, the certification body did not approve piping classes according to the same B31.8 for dehydration unit which sits on discharge of the compression trains. I still cannot really understand why the piping on my left side was made of one material, and the piping on my right side had to be designed by using another code (for the same process line). I understand the codes, but I could not grasp the logic and their physical meaning in this case.

Dejan IVANOVIC
Process Engineer, MSChE

 

[BigInch]

Pipeline code boundary limit includes compressor stations and all items within up to and including any piece of equipment having control of flow, pressure, temperature or safety on the pipeline. A dehy unit would most likely be upstream of the suction side of the compressor, be separated by a block and ESD valve going to the compressor units and have no real direct control of pipeline pressure or flow, other than possibly on/off of the feed stream. Pipeline pressure would be influenced by compressors, perhaps a control valve and surely an ESD between compressor and downstream piping leading to the pipeline itself. A dehy should not be exerting any direct control of pressure or flow on the pipeline nor have any influence on the pipeline's operating safety, hence it is not included within the pipeline code boundary. Where a unit is located inside or outside a fenced area isn't a deciding factor as to which code applies. Hope that explains the reason they considered it outside the pipeline' code scope. However it can get confusing at times, as lesser items are usually included, such as filters, separators, fin fans and the like, but those must be included in the pipeline code piping simply because they are on the piping route to the valves and compressors that do have direct control and influence on the pipeline and that's where the buck stops.

 

[SNORGY]

Seems to me, the number that the gray-haired guys used to tell me (before my own hair went gray and my own bald spot took shape) was that, in the absence of other data, design temperature was 28 C above operating temperature.

The 50 F is a real number with historic basis. Right or wrong, it's historically correct.

 

[EmmanuelTop]

BigInch, thank you for the reply. The pipeline compressor station was an existing facility and we were coming with the new TEG dehydration train inside the same facility fence. Actually the Dehydration unit Absorber was located in continuation of the compressor discharge header. So there was an existing B31.8 compressor header, then we had to install 20 meters of ASME B31.3 piping to the new Absorber, then the outlet piping of the Absorber (also B31.3) was tied back to the compressor discharge header which was, of course, B31.8 again. This header was continuing towards the pipeline inlet ESDV. Apologize for not posting a simple sketch. I have no means to produce a scan at the moment.

At that time, this was a big fuss. Compressor discharge side including the header was designed for 64 barg. Being forced to use B31.3 for the Absorber inlet and outlet piping, and given the fact that 400# ASME B31.3 is not a standard piping class in the company, the certification body proposed 600# class. Wall thickness of 600# piping was two times the wall thickness of B31.8 piping designed for 64 barg. 600# piping requires PWHT. You can imagine the blow. Ultimately, since the compressor discharge side was greatly overdesigned with regards to design pressure, we had de-rated the entire discharge system to 50 barg which allowed us to use 300# piping - similar wall thickness to B31.8 64barg and no PWHT required. To today's day I never understood why we could not use the same B31.8 piping for Absorber inlet/outlet piping.

Apologizes to the OP for diverting from the original subject - this just came in as a part of discussion on the design conditions.

Regards

Dejan IVANOVIC
Process Engineer, MSChE

 

[BigInch]

Quite a mess, but I would tend to agree in that particular situation the existing compressor station was captured, taken over and converted into part of a dehy-"process plant", shifting the usual division of function boundary. Pressure, flow and safety control of the pipeline should then be transferred to another control point, say at a pipeline inlet flange to a valve unit incorporating a flow/pressure control valve and perhaps a relief valve with ESD, located between dehy and pipeline.

But I have done the same thing (used B31.8) for sitting dehys between field gathering compressors and gathering system pipelines located in non-regulated remote gathering systems. Since it was a rather low pressure field, some compression was needed to get the dehy down to reasonable size.

Ya. Sorry dso88. Sometimes it's just too easy to keep going in the same thread.