CSA Z662 Low Temperature 3

[302Hugo]

I am a bit confused with CSA's notch toughness requirements and I think I'm having a hard time understanding it completely.

Just like ASME has their Fig 323.2.2A Minimum Temp Without Impact Testing and their figure for reduction in temp in regards to the stress ratio, CSA has their Table 5.1, correct? In 5.2.2.1 it states that notch toughness properties shall not be required for less than 50MPa operating stress.

For crude oil I can assume it's a LVP fluid, correct? Now for a design operating stress of greater than 50 MPa the minimum design temp is "All" (I assume it can go down to -200C if the notch tests says it's ok?) but then the notch toughness category column lists it as Category I. Isn't this saying that I can use Cat I pipe (Non tested) for pressures greater than 50MPa? So why is there the 5.2.2.1 clause? And if this is a table for notch testing requirements, why is there Cat I pipe listed?

Obviously I am missing something here, so if someone can help me out, it would be greatly appreciated! Thanks.

 

[rneill]

Crude oil is a LVP product. I'll try to summarize as best I can ...

Per Clause 5.2.2.1, proven notch toughness is never required for pipe smaller than 4", piping with wall thickness less than 6.0 mm or pipe with design operating stress less than 50 MPa. In these situations, you would use Category I pipe regardless of the design minimum temperature.

If you do not meet the conditions specified in Clause 5.2.2.1, then Clause 5.2.2.2 requires you to follow the requirements of Table 5.1. For your example, I'm going to assume LVP and that you will hydrotest. Consequently, Table 5.1 says that for all minimum design temperatures and all operating stress values you can use Category I piping.

Let's look at another example, perhaps a sweet gas production pipeline with a design minimum temperature of -30 C, with a diameter of 12" and where the pipe grade is 359. In this example, there are two possibilities (a) and (b). Let's say that in case (a) my design operating stress (based on MOP) is 275 MPa and that in case (b) my design operating stress (this time a lower MOP) is only 200 MPa.

Case (a) in this case the stress of 275 MPa exceeds the PTSV of 210 MPA (see column 2 of Table 5.2) so Table 5.1 requires me to use Category II material. In this case, if my design minimum temperature was less than -30 C then Table 5.1 would have required Category II pipe.

Case (b) in this case, the stress of 200 MPa is less than the PTSV of 210 MPA so Table 5.1 allows me to use to use Category I pipe.

Once you determine the Category of pipe specified in Table 5.1 then there are two possible further exceptions as described in the notes to Table 5.1. For Category II pipe where the design operating stress is less than the PTSV, you are permitted to substitute Category I pipe for pipe runs shorter than 50 m. For other than CO2 pipelines, you can substitute Category III pipe for Category II pipe in runs less than 100 m.

Note: Category III pipe would usually be where you were using low temperature ASTM materials in place of CSA materials. Low temp CSA materials are recognized as Category II but many low temperature ASTM materials are classed as Category III.

To answer your specific question, Table 5.1 is telling you that regardless of the design minimum temperature, and regardless of the operating stress, you can use Category I pipe (pipe without proven impact toughness) for LVP pipelines so long as you complete a liquid pressure test. Recognizing that Codes are not design handbooks, and that reasonable engineering judgement must always be applied, I'm not sure that most designers would accept Category I pipe for temperatures down to -200 C (and this probably was not intended by the Code committee). The cost impact of low temperature materials is not what it was 15 or 20 years ago and I always specify low temperature (Category II or III) materials for design minimum temperatures less than -30 C, even for LVP service.

 

[SNORGY]

Thanks rneill.

I knew your answer would be better than mine, so I waited for it before posting. I was not disappointed.

Your last paragraph is consistent with my thinking - and with a design tool I developed years ago (which I now see is conservative but not correct in its treatment of notch toughness requirements in LVP pipelines that are subjected to a liquid hydrostatic pressure test). I now understand the failing of my ways...

So, I'll add two things:

(1) Clause 1.3(a) in CSA Z662-07 states that the Standard does not apply to piping with metal temperatures below -70 C. So, lower than that and you are outside of the scope of CSA Z662-07.
(2) The "Student Guide" to CSA Z662-07 does clearly state that proven notch toughness is not required for LVP pipelines because such fluids would produce, upon failure of the pipe, a fish-mouth fracture that does not have the potential for long propagation, since they have comparatively low stored energy. The "Commentary" to CSA Z662-07 states that there is a history of successful operation of LVP pipelines using Category I pipe material. However, with that said, there is (or has been) a course ("Oil And Gas Pipeline Systems") taught on CSA Z662 every year or two by an industry expert where the interpretation in the Course Notes appears to be that:

(a) for DOS lower than PTSV(1) (Column 2) in Table 5.2, Category I is acceptable, then;
(b) For DOS between PTSV(1) and PTSV(2) (Columns 2 and 3) in Table 5.2, either Category III or Category II notch toughness is required, then;
(c) For DOS greater than PTSV(2) (Column 3) in Table 5.2, Category II notch toughness is required.

The wording of the course notes is somewhat unfortunate because the context under which the above interpretation applies, in retrospect, probably relates to gas service although this is not explicitly stated.

So, I may have a design tool to modify and edit after stumbling across this. Meanwhile, in the worst case, the result has been that I would hitherto have been inclined (perhaps in error) to specify Category III for PTSV(1) < DOS < PTSV(2) and Category II for DOS > PTSV(2), even for LVP service.

While rneill has implied that his recommendation may be similar, I have to give a star to Crittenden for the original post that has allowed me to see, and if necessary fix, an error in a design tool that I am otherwise rather proud of.

Regards,

SNORGY.

 

[rneill]

Thanks, I had long forgotten about the -70 limitation to Z662.

I like to maintain the Category II / III materials in all cases with a design temperature less than -30 since I virtually always have a bunch of B31.3 piping connected to the system and it keeps things very simple - consistent materials for both the Z662 and B31.3 systems (and for both gas and liquid services) thus helping to preventing any chance of screwups.

The KISS principle - "keep it simple" for the field construction folks and warehousing of inventory.

Not sure about the justification stated in the "Student Guide" since it is only addressing the issue of fracture propagation and not the issue that it would still be a brittle fracture occuring with little advance warning. Generally, we like to have piping systems that fail in a ductile manner so that there is plenty of advance warning for personnel.

Note: I recently had a client use Category I materials in an application permitted by CSA Z662 but the regulatory authority in that Province did not accept this. I don't know the ultimate resolution but I know that at one point they sent samples of product from every heat used for impact testing to try and resolve the issue.

 

[SNORGY]

As a follow-up, I researched my background information for the design tool I developed. It was based on Z662-03 and on the "Oil And Gas Pipeline Systems" course notes from about that era. In that earlier version of the course notes, the wording was such that it suggested, in an even stronger manner, the interpretation of Notch Toughness Requirements per my 2(a), (b), and (c) above. So, I don't feel quite as stupid as I did earlier this afternoon. While I think I will nonetheless check my rationale against the other fluid services (HVP, GAS, CO2), I am not as inclined to "fix" anything if the only "error" being made is specifying some measure of notch toughness for LVP pipelines where DOS > PTSV(1), irrespective of what hydrostatic test medium is used.

I guess I just hate being wrong...

We also see lots of ASME materials above grade marrying up to pipeline risers. The problem does sometimes arise that we have a let-down valve between, say, a wellhead and the pipeline, and the J-T effect gives rise to prediction of temperatures below -45 C. So, per the letter of Table 5.3, it becomes difficult to "accept" such materials under the scope of Z662, whereas they could otherwise be accepted via B31.3 323.2.2. Sometimes it seems like there is too much "Code Interpretation" and not enough "Science" at play, since the pipe is the same.

Regards,

SNORGY.

 

[rneill]

I guess one possible out might be that you really only need to consider the metal wall temperature which is not necessarily the same as the process fluid temperature (which is what the J-T calculation gives you). Analysis to get an estimate of the actual metal wall temperature is often performed in gas plant blowdown scenarios in order to avoid designing for extreme low temperatures.

However, now that I think about it, plant blowdown is a very short term situation whereas a pipeline let down would be continuous so at some point I would expect the metal wall to approach the process fluid temperature.

Note: in my original reply, I made an error and the last sentence under my Case (a) scenario should have been put with the Case (b) scenario.

 

[SNORGY]

I read that a couple of times and knew what you meant.

With respect to blowdowns, when the theory based on J-T alone (uncorrected by further analysis) predicts appreciable MDMT depression, I have been in the habit of suggesting skin temperature RTDs to "measure the reality" before recommending metallurgical upgrades to, say, austenitic. However, that recommendation is usually rejected. On the other extreme, I know of at least one plant where the blowdon lines to the flare header were all A-312-TP316 on the basis of an excursion down to -66 C at 350 kPag. Myself, I would have rationalized that the combined stress state was low enough that A-333-6 in combination with B31.3 323.2.2 would satisfy Code in that case. The problem is that stress analyses are normally not done on low operating temperature lines (clients and employers view it as a waste of engineering time), so the combined loading cases are never truly known and you end up "taking a flyer" at the 323.2.2 stress ratio as some multiple of hoop stress, or in the case of flanges, the ratio of the pressure to the flange rating. Certainly, then, a detailed calculation towards prediction of actual wall temperature is invariably discouraged. After all, since evaluation of the actual stress state is considered to be a waste of time, why would one waste more time calculating the actual pipe wall temperature - even if the evaluation of one depends entirely on the other anyway?

Upstream O&G: Cowboy Engineering at its finest. I'm so proud.

Maybe it's just that I haven't had my coffee this morning...


Regards,

SNORGY.