27 replies to this topic

[venkatesh.b]

Dear sir,

I am trying to size a bypass valve for flow control valve. We received control valve sizes from vendor. But now how to decide the bypass size for that control valve?

Please guide for this problem

Thanks,
Venkatesh

[fallah]

Dear sir,

I am trying to size a bypass valve for flow control valve. We received control valve sizes from vendor. But now how to decide the bypass size for that control valve?

Please guide for this problem

Thanks,
Venkatesh

Venkatesh,

Please at first refer to the relevant spec in your project. The criteria for necessity and selection of the by pass valve and its type/size would certainly be there. If you haven't any spec in this regard, please specify the valve size/type and other important information of the valve. With having those information i would suggest the proper by pass size for it.

Fallah

Edited by fallah, 28 September 2011 - 02:35 AM.

[venkatesh.b]

Sir,

Control valve information-Flow control valve
6" Size
DP valve sizing=2.1Kg/cm2
Inlet pressure=4.6kg/cm2g
outlet pressure=2.5kg/cm2g
Inlet temp=40.3 C
Service=Sourgas
Flow in pipe=3413m3/hr

And also sir please provde information for calculation size,
Because we have other two control valves,we want to calculate bypass size for thos evalves also..

[fallah]

Sir,

Control valve information-Flow control valve
6" Size
DP valve sizing=2.1Kg/cm2
Inlet pressure=4.6kg/cm2g
outlet pressure=2.5kg/cm2g
Inlet temp=40.3 C
Service=Sourgas
Flow in pipe=3413m3/hr

And also sir please provde information for calculation size,
Because we have other two control valves,we want to calculate bypass size for thos evalves also..

Venkatesh,

Seems you haven't any project spec regarding your qury. Well....

If your sevice is continuous, the control valve should be equipped with block valves and by pass. Sizes of block valves and by pass would be in accordance with API RP 550 and for mentioned control valve would be as follows:
-Sizes of block valves : 6" (if line diameter is 6") and 8" (if line diameter is 8").
-Size of by pass valve : 6" (for line diameter of 6" and 8").

For other valves, please refer to mentioned RP.

Fallah

Edited by fallah, 28 September 2011 - 04:12 AM.

[venkatesh.b]

 CONTROL VALVE ASSEMBLY IN P&ID.doc   87.5KB   385 downloadsSir,

Tahnk u so much for fast reply for my query .

Sir i'm attaching my control valve aseembly which is suggested by my lead.Plz check and guoded me if any corrections

[fallah]

Sir i'm attaching my control valve aseembly which is suggested by my lead.Plz check and guoded me if any corrections

venkatesh,

Supposed you mean case-1 of the attachment, if so please finf below my comments about the configuration and its detail:

- Sizes of the block valves should be 10" instead of 12" to meet the API RP 550 requirement.
- Because mentioned flow control valve is used as SDV, the valve along its by pass (size 10") should be specified as CSC (Car Sealed Closed).

Fallah

[venkatesh.b]

Thanks for reply sir

Sir can u attach API 550 ,We don't have that standard,that is very useful for me

Your prompt reply is awaited

[fallah]

Sir can u attach API 550 ,We don't have that standard,that is very useful for me

Venkatesh,

Sorry, it is not permitted and you can buy it or ask from your company to provide.

Fallah

[venkatesh.b]

Ok thank u so much sirr

[paulhorth]

Venkatesh,

Your PID shows the control valve also having a shutdown function. This is considered bad practice, especially on a fuel gas line. Every client I have worked for would require a separate ESD valve in the line.
Consider the following emergency scenarios:
(1) The control valve fails to operate correctly, leading to a fire at the unit being fed by the fuel gas. How do you shut the plant down safely?
(2) The control valve is offline and you are running on the manual bypass. There is a fire (anywhere on the plant). How do you shut down the fuel gas safely while allowing the operators to escape?

These hazards and others would come to light if a HAZOP review is carried out.

Paul

[deltaChe]

From the Cv values seems that a globe valve one size larger than a conventional double
seated equal percentage valve would be needed to obtain a Cv between 100 and 200% of that of the control valve, as called for in the International Practices. For a gate valve two sizes smaller than that of the control valve would be needed to meet these criteria. Gate and globe valve Cv values are not well known and it is doubtful that a contractor would select sizes in accordance with the International Practices unless the criteria were specifically brought to his attention.

For situations in which a flow rate many times normal is needed for occasions such as filling a unit, a larger than normal or perhaps a second bypass valve may need to be specified. Conversely, as pointed out under “Choice of Failure Position,” for control valves, there may be situations that justify putting an upper limit on the bypass valve capacity.

[kkala]

In a locally designed condensate treatment unit, any case of (steam) bypass valve was specified as having same Cv (flow coefficient at fully open valve) to that of the control valve; this on the ground that there will be manual control through bypass in case that control valve is removed for maintenance. In that sense bypass piping was of same diameter as main piping (*). Downstream PSV flow was specified for fully open control valve plus half open bypass valve (1.5Cv in total), in other cases both valves are conservatively considered fully opened (2Cv). Check & necessary revisions according to real Cvs from supplier (it could be one for both valves, to my information) is realized during detail engineering. This practice for bypass valves was often applied in my working environment .
As Steven7788 has clearly indicated, size of bypass valve depends on the function that you require from it; this function can be different from valve to valve. There are general criteria, yet I think the intended function is predominant for the design.
I also agree with paulhorth, control and ESD functions should be implemented by independent instrumentation for safety. I have only heard discussions in case of a boiler, whether their corresponding PLCs could be combined (answer not known by me).

(*) Instrument Dept often selects one size smaller bypass diameter, especially for big diameters ( to some practices). For the specific case same diameter was judged proper by Process.

Correcting note, 1 Oct 2011: PLCs of ESD and startup to be combined, if codes permit, not ESD and general control functions.

Edited by kkala, 01 October 2011 - 01:06 AM.

[S.AHMAD]

Hi Ven

1. If we understand the purpose of having the bypass line for control valve, we should not have any problem in determining the size of bypass valve. Having said that, my review of your two cases, bypass valve is over-sized.

2. ESD as an integral part of basic control system is an old design which is no longer acceptable by current practice of process safety. Please refer and study the attached diagram for basic understanding of the system. The first scheme (BPCS) is considered as a single layer of protection (higher risk). The second diagram (SIS) os a two layer of protection and hence increase the safety integrated level (SIL).

Attached Files

•  SIS.pdf   28.31KB   439 downloads

Edited by S.AHMAD, 30 September 2011 - 04:02 AM.

[kkala]

Interesting to look into "SIS.pdf" attached to S.AHMAD's post. BPCS diagram is similar to " CONTROL VALVE ASSEMBLY IN P&ID.doc" (attached by Venkatesh). As I understand SIS diagram, high temperature closes EBV as well as FC valve through same temperature switch (THSD).
According to concept of independent ESD from control actions, THSD would close only EBV. If it was considered necessary for FC valve to close from high temperature (and this is extra safety), it would be realized through another control action, e.g. through a separate THSD.
A recent boiler process duty spec (2008, boiler now made by Vendor) required this independence between control and "ESD / startup" functions. On the other hand technology progress may have made "mixing" of control and ESD functions reliable enough in some cases.

Edited by kkala, 01 October 2011 - 12:59 AM.

[venkatesh.b]

Thank you so much for the help...

Ahmed sir

Can you provide any theory for SIS.pdf so that i can understand more clearly because i'm fresh process engineer.

[kkala]

As Venkatesh.b has said, brief concepts on which SIL / SIS is based would be helpful for a lot of us. Just a few lines by somebody knowing, if possible with some web reference. Here controls of a new refinery project were checked by a foreign consultant using the Safety Instrument Level methodology, but we have not been aware of it. Introductory seminars are occurring, e.g. http://www.cvent.com...a7f8334f3a.aspx.

Editing Note: Found http://library.abb.com/global/scot/scot211.nsf/veritydisplay/5c62194fa3f47c2dc1256ed70030f1b0/$File/3KDE010001R5001.pdf could give some insight.

Edited by kkala, 02 October 2011 - 03:07 AM.

[S.AHMAD]

Dear kkala, ven and others.

You can also get the details from following standards:


1. ANSI/ISA-84.01-1966: Applications of Safety Instrumented Systems for the process Industries
2. ANSI/ISA-84.00.01-2004 (3 Parts)
3. Technical Report: ANSI/ISA-84.00.02-2004 (5 Parts)

For an introduction, please find attached a copy of presentation slides that I normally used to introduce the subjects to participants of HAZOPS Training Workshop that I frequently conduct.

In the slide: PFD = Probability of Failure on Demand

Hope the above info helps.

Attached Files

•  Module 10 SIS&LOPA Compatibility Mode.pdf   387.68KB   288 downloads

Edited by S.AHMAD, 02 October 2011 - 08:27 PM.

[kkala]

Thanks for the useful presentation (Module 10 SIS&LOPA Compatibility Mode.pdf), clarifying (among others) what SIL means. "Independent Protective Layer" picture also impressed me. Failure on demand probabilities have been seen here in books on Safety Studies (Seveso II), but these studies (as far as I know, local experience) do not include synthetic evaluation of probability for a serious accident, based on specific probabilities of partial events causing it. We simply report general preventive and suppressive / mitigating measures for them. Checking instrumentation and control through this method seems interesting, the challenge is to know the plant functions well.

[fallah]

The subject of this thread is "Control Valve By pass Size" and now we see the deviation to the subject from almost 10th post.....

Wasn't it better a new thread been opened with the subject of SIS, ESD,....,or something like it....?

Edited by fallah, 03 October 2011 - 07:07 AM.

[dwwjt812]

In a locally designed condensate treatment unit, any case of (steam) bypass valve was specified as having same Cv (flow coefficient at fully open valve) to that of the control valve; this on the ground that there will be manual control through bypass in case that control valve is removed for maintenance. In that sense bypass piping was of same diameter as main piping (*). Downstream PSV flow was specified for fully open control valve plus half open bypass valve (1.5Cv in total), in other cases both valves are conservatively considered fully opened (2Cv). Check & necessary revisions according to real Cvs from supplier (it could be one for both valves, to my information) is realized during detail engineering. This practice for bypass valves was often applied in my working environment .
As Steven7788 has clearly indicated, size of bypass valve depends on the function that you require from it; this function can be different from valve to valve. There are general criteria, yet I think the intended function is predominant for the design.
I also agree with paulhorth, control and ESD functions should be implemented by independent instrumentation for safety. I have only heard discussions in case of a boiler, whether their corresponding PLCs could be combined (answer not known by me).

(*) Instrument Dept often selects one size smaller bypass diameter, especially for big diameters ( to some practices). For the specific case same diameter was judged proper by Process.

Correcting note, 1 Oct 2011: PLCs of ESD and startup to be combined, if codes permit, not ESD and general control functions.

Dear fallah,
For PSV calculation of control valve failure, why downstream PSV flow should plus half open by pass valve(1.5Cv in total), in other cases both valves are conservatively considered fully opened (2Cv)?
thanks
Jia Tong

[jrtailor09]

Dear Venktesh,

There is a standard chart in EIL engineering guidlines with respect to control valve, block valve & bypass valve size is defined. This is thumb rule no need to go for any calculation.

Generally for the lower size of control valve around 3 " bypass valve size shall be same as control valve size & above that one size lower. Please follow standard table.

Regards,

Jatin

[fallah]

Dear fallah,
For PSV calculation of control valve failure, why downstream PSV flow should plus half open by pass valve(1.5Cv in total), in other cases both valves are conservatively considered fully opened (2Cv)?
thanks
Jia Tong

Jia Tong,

Attached from API 521 included the response to your query. If you need additional assistance inform me to provide.

Fallah

Attached Files

•  API_STD_521-Control Valve Failure.pdf   81.63KB   328 downloads

Edited by fallah, 14 November 2011 - 01:58 AM.

[kkala]

Dear fallah, For PSV calculation of control valve failure, why downstream PSV flow should plus half open by pass valve(1.5Cv in total), in other cases both valves are conservatively considered fully opened (2Cv)? thanksJia Tong

Jia Tong, Attached from API 521 included the response to your query. If you need additional assistance inform me to provide. Fallah

Attachment by Fallah speaks of two cases, as examples:
(1). When bypass valve is partially opened during operation to "assist" the control valve.
(2). When bypass valve operates instead of the control valve, removed for maintenance.
Case (1) could conservatively correspond to 2Cv, while case (2) to 1.5Cv. It is noted that Cv of control and bypass valve was considered same in the post of 30 Sep 11, where 1.5Cv & 2Cv was mentioned.
API can differ somehow to presented local practices, the latter being simplified and (in this case) more conservative.

Edited by kkala, 14 November 2011 - 06:24 AM.

[dwwjt812]

Dear fallah, For PSV calculation of control valve failure, why downstream PSV flow should plus half open by pass valve(1.5Cv in total), in other cases both valves are conservatively considered fully opened (2Cv)? thanksJia Tong

Jia Tong, Attached from API 521 included the response to your query. If you need additional assistance inform me to provide. Fallah

Attachment by Fallah speaks of two cases, as examples:
(1). When bypass valve is partially opened during operation to "assist" the control valve.
(2). When bypass valve operates instead of the control valve, removed for maintenance.
Case (1) could conservatively correspond to 2Cv, while case (2) to 1.5Cv. It is noted that Cv of control and bypass valve was considered same in the post of 30 Sep 11, where 1.5Cv & 2Cv was mentioned.
API can differ somehow to presented local practices, the latter being simplified and (in this case) more conservative.

Dear fallah,
For PSV calculation of control valve failure, why downstream PSV flow should plus half open by pass valve(1.5Cv in total), in other cases both valves are conservatively considered fully opened (2Cv)?
thanks
Jia Tong

Jia Tong,

Attached from API 521 included the response to your query. If you need additional assistance inform me to provide.

Fallah

dear fallah,
Thank U so much. This is a big help for a fresh engineer like me.
Another question: in the all normal cases, I can calculate and seek out the corresponding Cv of control valve. The vendor of control valve can confirm its final size according to these Cv. For case of control valve failure, when calculate relieving flow of PSV, we should use the Cv of fully open control valve. But some time, this value is too great and accordingly, the relieving flow is too great. For this problem, I find some one use the double Cv of the normal case as the the Cv of fully open control valve to calculate the relieving rate. Is this method right? in the all normal cases, I can calculate and seek out the corresponding Cv of control valve. The vendor of control valve can confirm its final size according to these Cv. For case of control valve failure, when calculate relieving flow of PSV, we should use the Cv of fully open control valve. But some time, this value is too great and accordingly, the relieving flow is too great. For this problem, I find some one use the double Cv of the normal case as the the Cv of fully open control valve to calculate the relieving rate. Is this method right? in the all normal cases, I can calculate and seek out the corresponding Cv of control valve. The vendor of control valve can confirm its final size according to these Cv. For case of control valve failure, when calculate relieving flow of PSV, we should use the Cv of fully open control valve. But some time, this value is too great and accordingly, the relieving flow is too great. For this problem, I find some one use the double Cv of the normal case as the the Cv of fully open control valve to calculate the relieving rate. Is this method right? whether this method is related to the selective principle of the Cv value?


dear kkala,
thanks for your reply.
For case (2) you said, 1.5 Cv is more more conservative.1 Cv is just ok as API 521. Is my opinion right?
For some project, their manual valve in the by pass line is lock-closed and so they don’t consider the double relieving rate. For this measure, how do you think that?
regards,



Jia Tong

Edited by dwwjt812, 15 November 2011 - 02:08 AM.

[kkala]

dwwjt812, on 15 November 2011 - 08:40 AM, said:
dear kkala, thanks for your reply.
For case (2) you said, 1.5 Cv is more more conservative.1 Cv is just ok as API 521. Is my opinion right?
For some project, their manual valve in the by pass line is lock-closed and so they don’t consider the double relieving rate. For this measure, how do you think that? regards, Jia Tong
A correction on my previous post: Presented local practices (realized) are simplified but not generally more conservative compared to API. See "API_STD_521-Control Valve Failure.pdf", next paragraph to the one placed in red square.
- Case (2), bypass valve operating, when control valve is at maintenance: API says that you can base relieving capacity on single Cv (maximum of the two), on the condition that bypass valve is locked close (administrative control) during normal operation.
Also that only "locked close valve" may not be reliable, if inadvertent opening of bypass valve can cause overpressure exceeding hydrostatic test pressure (corrected, I assume for temperature) somewhere downstream, likely to cause a catastrophic accident .
Note: I have not seen this last clause applied, advice / examples by others would be welcomed. Risk assessment during hazop may be needed.
- Locked closed bypass valve, while control valve is operating: Same case as before.
Hope it helps, representing my understanding.

Edited by kkala, 15 November 2011 - 06:42 AM.