25 replies to this topic

[jprocess]

Dear All
When the control valve failure case should be considered as a credible scenario?
How relief load should be calculated in case of control valve failure?
Thanks in advance.
Cheers.

[marthin_was]

The relief load will be calculated by using information max Cv value of the CV, then we can calculate the flow if the valve 100% open

CMIIW,

Dear All
When the control valve failure case should be considered as a credible scenario?
How relief load should be calculated in case of control valve failure?
Thanks in advance.
Cheers.

[JoeWong]

jprocess,
May be you turn-around and ask yourself, when can i not to consider CV failure is not credible ?

You may consider maximum CV upstr pressure i.e. PAH, 100% opening of CV, and downstr. pressure 110% of PSV setpressure ( assuming downstr. equipped with vessel designed according to ASME)

JoeWong

[pleckner]

@jp:

To add to what Joe wrote, IF a failed open control valve or regulator can overpressure downstream equipment then this becomes a credible relieving scenario. The relieving rate will be the maximum flow through the valve (maximum valve Cv). The pressure across the valve will be the maximum upstream pressure minus the downstream PSV relieving pressure. Note that the upstream pressure may even be a relief pressure. You need to consider this possibility.

[jprocess]

Dear Joewong and Phil,
Thanks a lot for your valuablr comments.
Dear Joewong:
I think when the function of control valve is not about pressure reduction between a high pressure source and a low pressure one,we may not to consider control valve failure case.Agree?
Dear Phil:
I describe what I found from your guideline to see if I get it right?
Volume basis formula for control valve capacity factor calculation for incompressible fluids is as follows:
CV=Q*√G/∆P
Where:
∆P=P1-P2
P1 : maximum upstream pressure i.e. PAH
P2 : downstream PSV relieving pressure i.e. 110% of downstream equipment PSV set pressure
G : sp.gr
CV : valve capacity factor at 100% opening
So we can calculate the Q and this is the relieving load.
Ok?
Can we use above mentioned equation for compressible fluids,too without any correction specially near critical conditions?
Thanks in advance.
Cheers.

[pleckner]

@jp:

It doesn't matter if the function of the control valve is pressure reduction or flow control. If by opening the control valve you can overpressure downstream equipment, then you have a credible relieving scenario. The calculation method is treated the same.

I strongly advise you obtain a copy of Fisher's Control Valve Handbook. They have worked examples for both liquid and gas/vapor control valve sizing. The equations used are all based on ISA. You should use the method described in the handbook to calculate the required relieving rate across a fully opened valve. You can get the handbook on-line, free. Just Google Fisher Controls. That should get you there.

[JoeWong]

@jp:

It doesn't matter if the function of the control valve is pressure reduction or flow control. If by opening the control valve you can overpressure downstream equipment, then you have a credible relieving scenario. The calculation method is treated the same.

Phil,
Bravo..., your explanation is absolute correct...

jp,
My appologize for my misleading sentence...

I strongly advise you obtain a copy of Fisher's Control Valve Handbook. They have worked examples for both liquid and gas/vapor control valve sizing. The equations used are all based on ISA. You should use the method described in the handbook to calculate the required relieving rate across a fully opened valve. You can get the handbook on-line, free. Just Google Fisher Controls. That should get you there.

Just for the benefits of all...FISHER HANDBOOK

JoeWong

[jprocess]

Hi,
Just one question:
How can we obtain valve capacity factor (Cv) at 100 % opening, without any vendor data?!
Without knowing this parameter we can not calculate the relief load!
Thanks in advance.

[Art Montemayor]

Jprocess:

This one is pretty obvious. You obtain the Vendor data. You call or write the vendor for this data; it's part of what you buy when you pay for a control or block valve.

You certainly (and no one here) are in no position to calculate it - although you can experimentally obtain it (which I doubt is cost-effective).

[jprocess]

Dear Art:
Thanks a lot for your reply.
My mean exactly refer to your statement:
"although you can experimentally obtain it"
This is a rough estimation of relieving load for control valve failure case and at this stage I have not any vendor data.
So I want to know that how can I experimentally obtain it?
Is there any rule of thumb in this field?
Thanks in advance.
Cheers.

[Art Montemayor]

Jprocess:

Whenever a problem or project starts to take on what seems to be a hue of complexity, I immediately change gears and go into pure, undiluted basics.

You now state that you “want to know that how can I experimentally obtain it (the C_V)”. Well, I would refer you to Crane’s Technical Paper #410, page 2-3, where you will find Crane Flow Tests and Figure 2-3, showing how Crane measures the pressure drop through valves. This is why I explicitly stated that I consider this experimentation as not cost effective for one simple valve.

Now refer to Fisher’s Control Valve Handbook (as my friend Joe Wong would say) and Section 3: Control Valve Selection. Every Chemical Engineer on this planet should have a copy of this Handbook. There is no excuse for not having it since it has been offered free of charge as a download on the Internet. On page 63 and “Valve Sizing”, you will find:

Q = C_V (deltaP/G)^1/2

Where,

Q = Capacity in gallons per minute;
C_V = Valve sizing coefficient determined experimentally for each style and size of valve, using water at standard conditions as the test fluid;
deltaP = pressure differential across the valve in psi
G = specific gravity of fluid (water @ 60 ^oF = 1.0000)

Note very carefully the definition of C_V. Fisher is saying exactly what I stated earlier. However, I would not resort to that. I would ask the manufacturer for the correct C_V for the subject valve. I don’t understand your reluctance to simply inquire with your valve’s manufacturer – but that is your decision. There is NO magical “rule of thumb” for estimating the C_V for a control valve.

But you can estimate a similar valve’s C_V. You do this by simply resorting again to Fisher’s Control Valve Handbook and look in pages 74 through 76. There you will find a listing of the C_V for specific types of valves. You select something similar and go with that value. That’s why the real smart engineers always specify Fisher Control Valves: you are backed up by years of expertise and secure and accessible information for the valves you buy from them. Customer service is the real strong basis for Fisher’s success in the control valve business. You haven’t told us the make and type of control valve that you can’t find a C_V for. You would do us all a service if you revealed this information and tell us all why you can’t obtain the basic information that you need and deserve to get. We all learn about good equipment manufacturers from other engineers. If there is a bad control valve manufacturer out there, I want to know about it.

[JoeWong]

Mr. Montemayor,
Thanks for treating me as one of your friend...a happiest news for my day...

Jprocess,
Mr. Montemayor has provided sufficient details on a Cv calculation and how you go about it.

The main objective is to get Cv of selected valve. Control valve selection is somehow required other information e.g. valve characteristic to suit process response, good operating point, trim type, valve cavitation consideration, etc. There is NO simple rule to select a control valve. Thus "There is NO magical “rule of thumb” for estimating the C_V for a control valve.". It required Instrument Engineer and vendor special knowledges, experience...

Control valve will seriously affect the relief load, PSV size, flare design, etc...Please get your instrument engineer in this design loop.

JoeWong

[Wouter]

A good set of postings on this topic already.
Just don't forget that it's not only fail-open valves which need to be considered as credible overpressure scenarios. The inadvertent (full) opening of a fail-close valve is always considered as a scenario as well!
Good luck

[pleckner]

I just re-read my post and it didn't come out the way I wanted it to. I should not have said "failed open" valve but just that the valve failed opened. @Wouter is correct but I will go one step further. We analyze all valves in that particular system and pick the one that if failed in the wrong direction will result in the largest relief.

[jprocess]

Dear Phil and Wouter:
You stated that we should consider all control valves independent of their fail position.I confused that how do you define control valve "failure" scenario?What does "failure" here mean?
If we consider both FC and FO control valves for control valve failure scenario,so failure here does not refer to "loss of instrument air" but seems to be a kind of valve mechanical problems which put control valve in 100% opening position! am I right?
My another question is about pressure drop across control valve.
As I found we should consider high pressure alarm(or even relieving pressure) of vessel upstream of valve and relieving pressure of vessel downstream of valve for pressure drop calculation.
In my case study the control valve located downstream of a slug catcher with 67 Barg design pressure and upstream of a copressor suction drum with 40 Barg Design pressure.Following your guideline the pressure drop across cotrol valve will be about 67-40=27 bar! The resultant flow rate may not be handled by PSV of suction drum or it may not be practical to size that PSV for control valve failure.
The operating pressure of slug catcher is 18 Bara and I can not understand that why should we consider PAH or design pressure of upstream vessel?How the outlet gas from slug catcher with 18 bar operating pressure can reach the 67 barg pressure?
Thanks in advance.

[JoeWong]

jp,

Phil / Wouter both have clearly pointed out that analyses ALL valves in that particular system and pick the one that if failed in the wrong direction (could be…operator inadvertent open a normally closed valve, hydraulic/pneumatic supply failed causing valve full open, controller giving wrong signal causing valve full open, valve mechanically stuck-open, etc) will result in the largest relief,...regardless what the failure position of the valve, the main point here is FULL OPENING of a valve (out of ALL valves) result in largest relief load…

Read several times statement posted by Phil and Wouter, you will understand the meaning. I leave it to others who are good in explanation to clarify in details…

To answer your second question, you may consider MAXIMUM operating pressure upstream of Control Valve. (i.e. PAH, you may consider PAHH if you would like to take conservative approach). You need not consider DESIGN pressure as this is double jeopardy.

For your system, you may only consider PAH or PAHH, both (any set point lower than design pressure) set by yourself. Example…PAHH could be 95% of Design pressure and PAH could be 90% of design pressure…

The slugtacher design pressure of 67 barg could be set by other scenarios such as maximum pipeline operating pressure duirng packing mode, increase wall thickness to reduce blowdown rate to manageable flaring load, etc. Please check the basis setting the design pressure, then you will understand the possibility of having pressure higher than 18 barg.


Out of this topic :

I have some doubt with your design pressure setting. Can you advise your compressor discharge pressure? Have you calculated compressor settle out pressure ?


JoeWong

[jprocess]

Dear Joe Wong,

Thanks a lot for your valuable comments. Like the always I enjoyed your profession a lot.

1. At this stage I have not calculated the compressor settle out pressure yet.
2. This is a 2-stage compressor station. The outlet pressure from the 1st compressor is 34 bar and from the 2nd compressor is 64.7 bar and I think the 40 barg design pressure for this compressor station is logical.
3. What is your idea about sizing control valve for a pressure drop of 67-40=27 bar? As you know typical pressure drop for control valve is 0.7 bar.
4. You are completely right. The 67 barg is the pipeline design pressure. But I did not understand your statement "increase wall thickness to reduce blowdown rate to manageable flaring load"! According to API 521 blowdown start from design pressure downto half of design pressure or 100 psig whichever is lower in 15 min. So how increasing the wall thickness will reduce blowdown rate?
5. I want to explain more about my case study. In fact, the feed gas to compressor station comes from 2 reception facilities. For the 1st the design pressure is 47 barg and for the 2nd as I stated before the design pressure is 67 barg. So I think my "system" for control valve failure study comprises of 2 slug catchers and compressor suction drum. Each slug catcher have 2 control valves on gas and condensate outlet lines. Between these 4 control valves the one located at gas outlet of 2nd slug catcher (67 barg) seems to be the one that if fails will results in the highest flow rate.

So there are some option for suction drum protection:

Option 1: Increase the suction drum design pressure to 67 barg. This option seems not to be economic.
Option 2: Size the suction drum PSV for 2nd slug catcher control valve failure with a pressure drop of 67-40 = 27 bar that as I stated before may not be practical because of high resultant flow rate.
Option 3: Increase the suction drum design pressure up to the 1st slug catcher design pressure(47 barg) and set the PSV located on the 2nd slug catcher (with 67 barg design pressure) to 47 barg.

Thank you.
Cheers.

[Art Montemayor]

Jprocess:

This thread is taking on an important contribution from its participants and the amount of critical basic data is now increasing on your part. However, it has been exceedingly difficult - if not downright tiring and taxing - to try to read your sentences when you have failed to leave the customary spaces between them! Any language is made of up of sentences, phrases, and thoughts that must be linked to form a logical message. If you don't leave spaces between your sentences, it is very difficult for your reader to try to make sense or understand what you are trying to communicate. That's why your keyboard has a SPACE BAR. Please try to use it and give our guys a chance to understand you and try to really help you out.

If you noticed, I have taken the liberty of redacting and formatting your last post.

[JoeWong]

ALL,
I appologize for raising some questions which move far away from main context of ORIGINAL thread.


Jprocess,
I believe you have got the main idea of your question earlier. We should stop from here...


Just not to disappoint you, i will drop some notes to your questions...

1. At this stage I have not calculated the compressor settle out pressure yet.

Point 1 noted.

2. This is a 2-stage compressor station. The outlet pressure from the 1st compressor is 34 bar and from the 2nd compressor is 64.7 bar and I think the 40 barg design pressure for this compressor station is logical.

As advised by Mr. Montemayor in many other posts, please state pressure unit either in "barg" or "bar a". Make this a habit. You will encounter disaster when you work on low pressure system e.g. storage tanks, vent & flare system, etc

My advice to you is do a quick calc on the settle-pressure for
(i) first stage alone and second stage alone assuming the check valve on 1st stage comp discharge works fine
(ii) both stages connected assuming the check valve on 1st stage comp discharge failed
You will have the idea if the suction scrubber design pressure is OK.

This may totally change your design pressure setting.

3. What is your idea about sizing control valve for a pressure drop of 67-40=27 bar? As you know typical pressure drop for control valve is 0.7 bar.

I am almost certain that the answer is NO for sizing control valve for a pressure drop of 67-40=27 bar in your case. You will probably under-sized the control valve.

I would suggest you establish the MAXIMUM and MINIMUM flow (includes MINIMUM turndown) and MAXIMUM and MINIMUM pressure drop across control valve (these figures may or maynot happen coincidently) from your production forecast and operating philosophy. Size your control valve for

- MAXIMIUM flow with MINIMUM pressure drop
- MINIMUM flow with MAXIMUM pressure drop

4. You are completely right. The 67 barg is the pipeline design pressure. But I did not understand your statement "increase wall thickness to reduce blowdown rate to manageable flaring load"! According to API 521 blowdown start from design pressure downto half of design pressure or 100 psig whichever is lower in 15 min. So how increasing the wall thickness will reduce blowdown rate?

Please ignore statement "increase wall thickness to reduce blowdown rate to manageable flaring load".
It is just the possible scenario to get high wall thickness, hence design pressure but totally out from this post. Answer this question will probably generate another 10 questions...Will explain when come to right time.

5. I want to explain more about my case study. In fact, the feed gas to compressor station comes from 2 reception facilities. For the 1st the design pressure is 47 barg and for the 2nd as I stated before the design pressure is 67 barg. So I think my "system" for control valve failure study comprises of 2 slug catchers and compressor suction drum. Each slug catcher have 2 control valves on gas and condensate outlet lines. Between these 4 control valves the one located at gas outlet of 2nd slug catcher (67 barg) seems to be the one that if fails will results in the highest flow rate.

So there are some option for suction drum protection:

Option 1: Increase the suction drum design pressure to 67 barg. This option seems not to be economic.
Option 2: Size the suction drum PSV for 2nd slug catcher control valve failure with a pressure drop of 67-40 = 27 bar that as I stated before may not be practical because of high resultant flow rate.
Option 3: Increase the suction drum design pressure up to the 1st slug catcher design pressure(47 barg) and set the PSV located on the 2nd slug catcher (with 67 barg design pressure) to 47 barg.

Multiple inputs with different design pressure made your system even more complicated. Can't drop any notes without detail analysis of your systems...

Another advice to ALL...Simple sketch tells million story...Engineer likes sketch more than words

Good day.

JoeWong

[pleckner]

@jp:

Another point (or clarification and summary) if I may. In calculating the relieving rate across a valve (control or otherwise) we must be careful in stating/defining the pressure we will use. We must differentiate between operating, relieving and design pressures.

1. We DO NOT use design pressures in the calculation.

2. The valve is assumed to be in the fully opened position, maximum Cv, irrespective what caused it to be that way.

3. The downstream pressure is the relieving pressure.

4. The upstream pressure is the maximum pressure that can be expected. This can be a pressure at some alarm point or even be a relieving pressure in the upstream vessel (note that this would not necessarily constitute double jeopardy and thus must be considered if credible).

It doesn't get more complicated than this.

[jprocess]

@jp:

Another point (or clarification and summary) if I may. In calculating the relieving rate across a valve (control or otherwise) we must be careful in stating/defining the pressure we will use. We must differentiate between operating, relieving and design pressures.

1. We DO NOT use design pressures in the calculation.

2. The valve is assumed to be in the fully opened position, maximum Cv, irrespective what caused it to be that way.

3. The downstream pressure is the relieving pressure.

4. The upstream pressure is the maximum pressure that can be expected. This can be a pressure at some alarm point or even be a relieving pressure in the upstream vessel (note that this would not necessarily constitute double jeopardy and thus must be considered if credible).

It doesn't get more complicated than this.

Dear Art:
Thanks a lot for your advise.I will try to type my posts correctly.
Dear JoeWong:
The design pressures that I stated are all in Barg.
I prepared a PDF file from part of my HYSYS simulation file that can be used as sketch.
If you see the sketch, the Aboozar sealine and slug catcher design pressure is 47 Barg and
for Foroozan is 67 Barg.The compressor station as you see have 2 tarins.Is this sketch clear enough?
If so I am willing to read your comments about those 3 options.
Dear Phil:
I confused with your sentence:
"4. The upstream pressure is the maximum pressure that can be expected. This can be a pressure at some alarm point or even be a relieving pressure in the upstream vessel (note that this would not necessarily constitute double jeopardy and thus must be considered if credible)."
As I remember this is in contrast with Joewong's idea that the upstream pressure will not reach the design or relieving pressure of upstream vessel because it will be a double jeopardy!
As I stated before,I have no sense that how upstream vesselcan reach its design or relieving pressure!
If it reaches its design pressure the PSV will open and relieve the fluid to "flare network" and "not to downstream vessel".Am I right?
Thanks in advance.

Attached Files

•  HYSYS_SKETCH.pdf   28.04KB   252 downloads

[pleckner]

@JP:

My apologies for indeed being confusing and what I now read in my last post as possibly being contradictory (my point No. 4 can be a contradiction to my point No. 1).

I don't want to get into a discussion of double jeopardy in this thread for two reasons:

1. This thread is getting too long as it is.

2. If it needs to be discussed, it should be discussed in its own thread.

However, let me say this. My point No. 4 is indeed correct. Sometimes in a control valve failure or control valve 100% opened scenario, you need to consider that the upstream pressure may be from a vessel that itself is in relief. You need to work this out yourself for your system and this would not necessarily constitute double jeopardy. I will argue this point all day. But I repeat and please understand, you need to determine if this is a possibility for YOUR PARTICULAR SYSTEM. If it is not, then don't worry about it.

If you are still confused about double jeopardy, please first read my article on this website: http://www.cheresour...asiseeit2.shtml

Then post your question in a new thread.

[jprocess]

@JP:

My apologies for indeed being confusing and what I now read in my last post as possibly being contradictory (my point No. 4 can be a contradiction to my point No. 1).

I don't want to get into a discussion of double jeopardy in this thread for two reasons:

1. This thread is getting too long as it is.

2. If it needs to be discussed, it should be discussed in its own thread.

However, let me say this. My point No. 4 is indeed correct. Sometimes in a control valve failure or control valve 100% opened scenario, you need to consider that the upstream pressure may be from a vessel that itself is in relief. You need to work this out yourself for your system and this would not necessarily constitute double jeopardy. I will argue this point all day. But I repeat and please understand, you need to determine if this is a possibility for YOUR PARTICULAR SYSTEM. If it is not, then don't worry about it.

If you are still confused about double jeopardy, please first read my article on this website: http://www.cheresour...asiseeit2.shtml

Then post your question in a new thread.

Dear Phil,
Thanks a lot for your valuable comments.
But I can not understand your mean from this statement:
"I don't want to get into a discussion of double jeopardy in this thread for two reasons:

1. This thread is getting too long as it is.

2. If it needs to be discussed, it should be discussed in its own thread."
I can not separate the discussion of double jeopardy from the main topic because these are related together.Double jeopardy is a general concept that should be considered here for this case study.I read all of your rupture disc series and also the part that is related to double jeopardy.
Also is there any problem for getting the topic too long?I believe the longer the topic the more useful points we gain.
Let's return to the main topic:
About your point:
"However, let me say this. My point No. 4 is indeed correct. Sometimes in a control valve failure or control valve 100% opened scenario, you need to consider that the upstream pressure may be from a vessel that itself is in relief. You need to work this out yourself for your system and this would not necessarily constitute double jeopardy. I will argue this point all day. But I repeat and please understand, you need to determine if this is a possibility for YOUR PARTICULAR SYSTEM. If it is not, then don't worry about it."
For my case the upstream vessel is a reception KOD with 67 Barg design pressure and a slug catcher with 47 Barg design pressure.I reviewd our preliminary relief load summary document for these 2 vessels and the considered scenarios are 1.Fire and 2.Blocked Outlet
If fire case occure for these 2 vessels the fire&Gas detection system will isolate these drums from downstream units and theire related PSVs will relief the fluid to flare network.So I repeat my question that how a gas stream with relieving pressure can flow to downstream unit?!
Also if blocked outlet case occure we have not any flow of fluid to downstream unit.
Could you please clarify this concern?
What is your idea about using HIPS?
Thanks in advance.

[pleckner]

@jp:

The problem with a thread getting too long is that people stop reading it, people loose track of what was said before and the whole meaning of the orginal quesiton is lost, especially if the thread goes off on tangents. A comprehensive discussion on double jeopardy should be left to a separate thread.

If you read and understood my article that discusses double jeopardy then you should be able to determine if using an upstream relief pressure (or a pressure close to the relief pressure) as the pressure into your opened control valve scenario makes sense for your particular system.

For your question, "If fire case occure for these 2 vessels the fire&Gas detection system will isolate these drums from downstream units and theire related PSVs will relief the fluid to flare network.So I repeat my question that how a gas stream with relieving pressure can flow to downstream unit?!"

If it won't, it won't. Have confidence in your analysis and conclusions. My concern is that you asked about an open control valve relieving scenario and we gave you things to look into, nothing more. If what we suggest you consider doesn't make sense for your system then OK. We're not trying to give you the scenarios, we are just making suggestions as to some things people fail to consider.

I've never done a system using HIPS myself but in general, and from what I've read, I don't have a problem with it. Using it or not using it is a choice that must be made by the facility. If it makes economic and practical sense for your particular system, then go for it.

[JoeWong]

jprocess,
Looking at the information you have provided...and the system is complicated.
...i have more than 10 questions to ask...I am sorry for not giving any further advice.

You have stated Fire & blocked outlet case...Please don't forget Control valve failure with pipeline in packing mode.

HIPS, there are many factors / requirements binding to a success implementation of HIPS. Try not to use it if you can.

Sorry for additional response. I think the thread should stop here.





JoeWong