API 650 Thickness Reinforcement Plate for Nozzles 5

[hamidun]

Dear All,

Is there any section in API Std 650 that rule the minimum/maximum thickness of reinforcement plate for nozzles (shell/roof nozzles)?

Please help!

Thank you

 

[Geoff13]

The value of peer review is priceless.

I'm new to Eng-Tips. In many places on the web making any comment just starts a flame war. I'm pleased to see this site is just helpful discussions amongst fellow engineers.

 

[JStephen]

rww88- The book is an intriguing idea. Thoughts below are on that.

One thing I notice in working with the ASME design is that where there seems to be any ambiguity, it also seems to be left up to the designer as to how to resolve that. So I don't have a lot of confidence that different experts would work the details exactly the same way, yet it wouldn't be obvious that one was wrong or right, either. Some examples of where this comes up- the diagrams illustrating the reinforcing areas are all laid out on straight lines, but many of the actual nozzles are on heads, so there can be some variation as to how all of that is laid out on an arc or on an elliptical surface. Where I notice this is calculating the "d" dimension for hillsided nozzles, my dimension is close, but never the same as, what our program calculates. Which is "right" or are both? No clue.

One issue I notice is that where the code gives minimum fillet sizes for standard slip-on flanges, our program applies those without adding a corrosion allowance on the inside dimension, whereas, I think it should be added (although not always possible!).

One issue that comes up is that where the "limits of reinforcing" as defined by the code extend into the knuckle, but the physical reinforcing pad does not, how is that treated? My understanding: As a nozzle in the knuckle. But, our program doesn't check for that, either. Along with that, several years back, I submitted a request for interpretation to the effect of, can you arbitrarily limit the Limits of Reinforcing where it is advantageous to do so? The answer given was "yes", but I'm somewhat surprised that I had to ask- surely others had thought of that? So either it had been asked and answered before, or everyone just assumed it was obvious and didn't need to be asked, or nobody ever thought to do that before. (Come to think of it, for an elliptical head, do you assume elliptical geometry or the 90/17 geometry, which is much simpler?)

It's been a good while since I looked into the external loads on API tank nozzles. When Appendix P first came out, I ordered the two papers it was based on. As I recall, one of those papers made the assumption that shell thickness was proportional to internal pressure. Which of course, is reasonable for large tanks, but doesn't hold true for tanks in the minimum-thickness range. When I asked about that, API modified that to limit the approach to tanks over 100'. Anyway, a couple of issues like that come up. One item- in the App. P, they had graphs with three straight lines. If you interpolated between the minimum and the maximum, you did NOT get the values on the middle curve- yet you had to interpolate the middle curve and one of the others, or extrapolate beyond, for any actual problem. Which raises the question, how do you interpolate those graphs? You can work things two different (reasonable) ways and get completely different numbers, since they are log-log graphs.

 

[rww88]

JStephen thank you for the extensive post that certainly generates much food for thought. During my research I have noted that many of the topics that you brought forth are not addressed by even the most popular books on the subject. I am going to take your points under consideration when composing my book's content.

I currently have two eng-tips forum members conducting critical reviews of my text and I am delighted and encouraged by their important feedback. This forum serves as a tremendous educational resource for which I fully intend to extend well justified literary credit.