14 replies to this topic

[gvdlans]

Philip,

In part 4 of your articles series on Rupture Disks, you wrote that three types of backpressure are considered, these being constant, built-up and superimposed. Constant backpressure is the backpressure that exists when there is no relief. Superimposed backpressure is the increase in backpressure because of reliefs from other safety devices and built-up backpressure is the increase in backpressure because of relief of the considered bursting disk.

API RP 521, section 1.3 and API RP 520 part I, section 1.2.3.3 only define two types of backpressure: built-up and superimposed. Superimposed backpressure is defined as: "The static pressure that exists at the outlet of a pressure relief device at the time the device is required to operate. It is the result of pressure in the discharge system coming from other sources, and it may be either constant or variable".

In other words, the API superimposed backpressure is the sum of "your" constant and superimposed backpressure.

It can be quite confusing if people use different definitions for the same words. In my company we recently changed the process datasheets to bring them in line with API. For superimposed backpressure we normally fill in a range.

[gvdlans]

Philip,

In addition to the above, can you explain why the superimposed backpressure for RD No. 1 in Figure 2A can be 110 psig? I would say it is max. the pressure in KO drum (=5 psig) plus pressure drop over header (=built-up backpressure of RD No. 1 = 5 psi), so maximum 5 + 5 = 10 psig.
Pressure drop over the (now ruptured) RD No. 2 must be 110 psig - 100 psig = 10 psi.

[pleckner]

Somebody is actually reading the stuff!!!!

The superimposed backpressure is the sum of all the pressures imposed on a relief device (as you point out by basically quoting API). It is not a true backpressure in of itself. By this I mean that it is made up of a number of different types of pressures in the system, i.e. constant and built-up as well as the system operating pressure. Also, I'm am not really trying to give a definition as such to these terms. I'm just trying to explain to the uninitiated the parts that make up the whole. And besides, I take issue with the way API defines "backpressure" because built-up backpressure is most definitely a component of superimposed backpressure. Maybe one day I'll write to them and see how they justify what I think is a very poor definition.

Also, it's unfortunate that API leaves out the definition for constant backpressure in their list becuase it has a direct affect on the rupture disk system (not so much for relief valves). And, this is a true backpressure that needs to be considered. Just ask the vendors!

And just for the record, the statement I write is, "This total system pressure is being exerted or imposed onto rupture disk No. 1. This is called the superimposed backpressure with respect to rupture disk No. 1. " Again, I think API and I are defining the term the same.

And yes, I totally agree that you should fill in the range when specifying a superimposed backpressure on a relief deivce because it will vary.

OK, now for the next question. I guess it all depends on where you define the point where the two discharge lines tie into the common header. To make a point, I am assuming that the two are rather close together. Therefore, ruputure disk No. 1 will be exposed to the high pressure in Vessel No. 2. just as vessel No. 2 goes into relief. I admit that I could have been more clear on this. As a matter of fact, in one of my earlier revisions to the article I did mention that the two were in close proximity to each other. I guess I should have kept that in. (Slap in the head).

Incidently, the pressure drop is not over the rupture disk but over the entire pipe from Vessel No. 2 to the knockout drum. The rupture disk is just another piping component in the system. And, if you work out the math (which I didn't because this is all made up to make a point), you would find that you reach choked flow (and the corresponding choked pressure) somewhere near the knockout drum. Therefore, the pressure in the pipe will actually be considerably higher than what you would guess by just looking at the sketch. I'll have to give this some thought and if it warrants, I'll re-write this section and get an update posted onto the website.

By the way, if you think I'm bad, just take a look at the exaggerated example API uses in RP520, Figure 29 to make the point!!!

Thanks for taking the time to read the series and if I didn't answer your question/concern to your satisfaction, please don't hesitate to challange me again. And that goes for anyone else!!!!

[pleckner]

To all interested,

This is in reference to Part 4 of my series on Rupture Disks for Process Engineers, which can be found on this Web Site under "Process Engineering-As I See It". As a result of the questions/comments by "gvdlans" (see the posts below), I added some clarifications to the the discussion of built-up and superimposed backpressures and somewhat modified the supporting sketches (Figures 2A and 2B). I hope the clarification I added helps you to more fully understand the point I was trying to make concerning the affects of backpressure on the specification of a rupture disk.

To those of you who have not yet read the series (a few more parts still to come), I hope you will take the time to do so. Specifying a rupture disk is not as straight forward as you may have thought.

And, if you have any questions, comments, challenges, concerns, etc. please do not hesitate to post them here!

[gvdlans]

Philip,

Thank you for your replies, and for the updated Part 4 of the series on rupture disks. I am convinced the series is very helpful for anyone specifiying rupture disks!

On the backpressure issue, as I wrote before, in my company we recently revised the process datasheets. We used to have entries for constant, built-up and superimposed backpressure, but "took out" the constant backpressure to bring them in line with API 520/521. Better said, we now include constant backpressure in superimposed backpressure.
Also, I noticed within my company that many engineers confuse the meaning of the various types of backpressures. This can turn out quite costly when new valves have to be bought in a late phase of the project. Maybe that is why I am so sharp on the issue...

On Figure 2A, with your additional explanation it now makes sense to me.

Thanks again! And you can be sure I will read the future parts as well...

[pleckner]

"GVDLANS"

Thanks for your input once more, it has been a pleasure conversing with you. And for those interested, I would like to comment one more time on this issue of backpressure.

Look at the catalogs from the major relief device manufacturers. They all talk about "constant" backpressure amongst built-up and superimposed. It seems odd to me that API would strongly recommend that the engineer consult with the manufacturer on issues but not include the terminology in their publications that are necessary for us to be able to talk to the manufacturer! That is one reason why, as I've said before, I do not necessarily and blindly follow everything that is written in API. They have been wrong in the past on a number of occasions.

Let me state it a different way. If I say that I want to design my relief device against a box load of pressure, well, wouldn't you like to know what is in that box? The "box" is analogous to "superimposed" since superimposed backpressure is made up of constant and built-up backpressure (built-up coming only from the flow affects of a different device somehow linked with the device in question). What about the relief pressure contribution from that other vessel in the system? Isn't this part of "superimposed" backpressure?

For a single relief device system, there is no flow until it relieves but there definitely is a backpressure that can mess up your design. Yes, even atmospheric pressure is imposed on the relief device and is thus a "superimposed" backpressure. However, it is really a constant backpressure. So why not describe what is in the "box"? After all, each component is different and each has its own magnitude. And, the manufacturers understand the terminology and are actaully expecting it in your specification, irregardless of what API writes! Who is the customer here, API or the manufacturer?

Now, I'm not advocating disregarding API!!! Heavens NO! But don't just blindly follow what doesn't make sense. You are allowed to question and deviate as long as you can show just cause to do so and in my opinion, this is one time to do just that. You aren't hurting anything by including "constant" backpressure in your specifications. You are not adding wrong information. You are just adding one more level of detail to describe what is in the "box". Oh, and unlike API, the rest of us understand the terminology just fine.

I will not get off my soap box.

[gvdlans]

Philip,

I see your point. However, I am not convinced that "unlike API, the rest of us understand the terminology just fine." With "the rest of us", I mean not just relief device specialists, but all process engineers, instrument engineers, relief device vendors etc. that have their part in specifying and selecting the relief device.

One advantage of the API approach is that it is simple. Superimposed backpressure is that backpressure that may exist before opening of the device, built-up backpressure is the additional backpressure that comes into place because of the opening of the particular relief device.

When I would add "constant backpressure" to my specification, risk is that someone down the line cannot be sure whether I already incuded the same amount in the superimposed backpressure or not. Of course, I can add notes to my datasheets describing what definitions I used, but it would be better to just refer to definitions in API 520/521.

So, if we are convinced that by not specifying the constant backpressure, a "wrong" relief device could be installed, we should indeed urge API to change their recommended practices!

[pleckner]

I've been doing relief system design for over 10 years and the data sheets I filled out never had superimposed backpressure written on them. They actaully just asked for Constant and variable backpressure, in those terms.

We're beating this one to death! The bottom line is, follow your company standard, plain and simple; as will I.

[sahas]

ASME PTC-25 -1994 states the following:

a) Superimposed Back pressure : the static pressure existing at the outlet of a pressure relief device at the time the device is required to operate. It is the result of pressure in the discharge system from other source.

Variable back pressure superimposed back pressure that will vary with time.

c) Constant Back pressure a superimposed back pressure which is constant with time

d) built -up back pressure : pressure existing at the outlet of a pressure relief device caused by the flow throughthat particular deviceinto a discharge system

e) Back pressure: the static pressure existing at the outlet of a pressure relief device due to pressure in the discharge system.

Thanks

[pleckner]

TO: Confused re. relief line losses

I sure wish you would have started a new thread for this one. I might have gotten to it sooner.

Anyways, your quesiton is usually asked coming from the other direction. That is, I usually get this question when someone is familiar with vapor/gas only calculations and needs to do two-phase calculations. It is irrelevant that the system is two-phase or single phase, you always need to consider the frictional losses n the relief line and especially when trying to determine choked flow.

Now, if I am misunderstanding your question, please repost with some more details; and please start a new thread?

[gvdlans]

To RVUser:

What you are writing is basically same as what I wrote in my original post.

However, as you may have read yourself, Philip Leckner does not agree with these API definitions: QUOTE "And besides, I take issue with the way API defines "backpressure" because built-up backpressure is most definitely a component of superimposed backpressure. Maybe one day I'll write to them and see how they justify what I think is a very poor definition." UNQUOTE

[pleckner]

To RVuser:

My series on Rupture Disks can be found on this website. Go to the home page (www.cheresources.com) and look at the menu items on the left. Find "Process Engineeing-As I See It". Click on this and you will find these articles.

On this backpressure issue, all I'm going to say is, use any standard you want. Use any definition you want. Just make sure you take all of it into account!