SIL certification 2

[GB67]

I have customers that are asking for SIL (safety integrity level) certification on industrial valves.

I don't have experience on this filed
I read some technical notes and the reference ISO / EN (EN 61508, EN 62061, prEN ISO 13849-1) standards and not it's clear to me what is SIL.

By I still have some unsorted pratical questions:

1) how to calculate pratically the SIL parameters (in detail the PFD) for a industrial valve? It's necessary to make serial test or what?

2) the SIL certifcation must be issued by an indipended body or by the valve manufacturer?

3) they are already on the market some brands that offer SIL certification? I yes who?

Mi questions are realted mainly to
- gate valves NPS 2" and over
- buttefly valves as above
- check valves as above
- ball valves as above

 

[terje61]

Dear GB67

Prior to discussing certifications, understanding of the required Safety Integrity Level (SIL) is necessary. SIL levels are not reached by purchasing certificates.

In IEC61511-3 it is explained how to determine the required SIL, which then in turn also explains the SIL.
The IEC61508-7 B6.6.5 refers to the IEC61025 for the Fault Tree Analysis which I feel show very well what happens when you calculate the PFD of the system you are installing. Then the system PFD should be compared with the required PFD (according to required SIL).

1) A PFD should be calculated for the complete system and the applicable element (e.g. valve) only.
In order to calculate it the failure rate applicable for your specific situation is required. So the failure rate applicable for: your stroking time, your environment (like desert or arctic), your fluids (like contains sand or not), your process (like size / rating etc.).
A failure rate from serial test should not be used, field experience should be used.

You also need the testinterval. How often you want to perform a full stroke test?

2) The product certificate shall be issued by an independent authority. Dependent of the required SIL a verification of the complete system shall be performed by an independent authority as well. You shall also read and verify the report yourself for all SIL levels.

3) They exist on the market.
Some for the bare valve only. So then you don’t know what the combination with the actuator will do and there is one with actuator. You could check Mokveld Valves

http://www.mokveld.com

for large sizes, they are specific design (not gate or ball) but mainly for SIL2 and above (e.g. in HIPPS). For SIL1 relatively standard ball valves are O.K.

M.v.h.
Terje

 

[JLSeagull]

The ISA 84.00.01 standard reflects the IEC 61511 requirements as applied in the USA. Most SIL discussion applies to the process areas. However third party testing agencies such as TUV certify safety system components. Much of the SIL level discussion pertains to avoiding shutdown based upon high reliability with reduncancy and voting, etc. The TUV certification is to assure that the components fail in a safe manner that would cause a shutdown.

 

[Ashereng]

There are SIL rated instruments - such as pressure transmitters, that are available. They come with TUV certification, and can be used in a safety system.

As for the valves, I have not seen any vendors with a SIL rated valve. In fact, I have not come across a vendor that will certify their valve to API bubble tight once the valve leaves their testing lab.

Oh by the way Terje, I am talking to Mokveld regarding their TUV SIL certificates, and guarantee of bubble tight once installed. I will advise.

The weakness in most safety system chain are the final elemnents. In this case, your valve. That is why in saftey applications, there is an extensive maintenance and testing procedure in place. Also for SIL 2 and SIL 3 applications, it is common to see 2 valves in series.


"Do not worry about your problems with mathematics, I assure you mine are far greater."
Albert Einstein
Have you read FAQ731-376 to make the best use of Eng-Tips Forums?

 

[terje61]

Hello Ashereng,

It would indeed be very interesting to know the guaranteed bubble tightness in safety applications. Any information regarding this is highly appreciated.

Regards,
Terje

 

[NGiLuzzu]

Ashereng,
in my opinion, the weak link in a safety system chain is NOT the valve as a rule: as a mere mechanical device, the valve itself as less possible failure modes than the logic and the controls.
Of course, it also depends on the valve type and complexity... but it is not accidental that the standards about "safety integrity" were born and developed for electric and electronic systems (programmable and not) at first.



terje61,
how would a few bubbles affect safety? In which cases?

As far as I know, safety functions of valves involve operation (quick opening or closing on demand, relief, emergency shut-down, action in case of actuator failure, etc.) more than tightness; I believe that, when tightness is very critical for safety and/or process, DB&B's or two valves are required (see thread408-150132).



Hope this helps, 'NGL

 

[JLSeagull]

Typically shutdown valves are line class valves specified in compliance with ASME B 16.34 and API SPEC 6D (ISO 14313). API 6D addresses bonnet and stem leakage in addition to seat leakage. Soft seated valves are initially bubble tight.

"Leakage for soft-seated valves and lubricated plug valves shall not exceed ISO 5208 Rate A (no visible leakage).
For metal-seated valves the leakage rate shall not exceed ISO 5208 Rate D, except that the leakage rate during the
seat test in 10.4.5.5.2 shall not be more than two times ISO 5208 Rate D unless otherwise specified. The test
procedures for various types of block valve are given in 10.4.5."

 

[terje61]

Responding to Anegri's remark:

Some of our safety shut down valves protect a part of the pipeline that has a lower pressure rating. In case of a calamity the pipeline will be blocked in several locations. When the SIL-rated high pressure safety valve is seriously leaking the pressure might rise in the blocked-in low pressure piping, resulting in a new potential dangerous situation. Up till now I have always installed a small pressure relief valve in the low pressure area. During 3-monthly tests we have not found leakages, but we are not measuring small flows in the short test periods. Personally I would be interested to know if the downstream pressure rise is a real scenario and if we must put more emphasis on testing the relief valve.

M.v.h.
Terje

 

[Ashereng]

Terje,

I am still waiting to hear back from Mokveld - I left a msg for them to call.

I too would like to know of the guranteed bubble tight in safety applications.

in my opinion, the weak link in a safety system chain is NOT the valve as a rule: as a mere mechanical device, the valve itself as less possible failure modes than the logic and the controls.
Of course, it also depends on the valve type and complexity... but it is not accidental that the standards about "safety integrity" were born and developed for electric and electronic systems (programmable and not) at first.

anegri,

Evidently, we disagree on the matter of what is more reliable - the valve or the safety system. I for my part, base my opinion on what I have seen, and the failure rates published on valve and SIS failure numbers.

With regards to why SIL rated SIS was developed first. One possible reason may be that it is easier to make the "brains" to a better quality and reliability and safety. Like I said earlier, I do not know of any valve manufacture guarantee the bubble tightness in a safety application. Which is why I really would like to talk to Mokveld. One reason for the lack of manufactures guranteeing their valves may be because it is very difficult to do so.

Anyhoot, that is my opinion. I welcome more information on this and the reasons for SIL SIS but not valves as well.

"Do not worry about your problems with mathematics, I assure you mine are far greater."
Albert Einstein
Have you read FAQ731-376 to make the best use of Eng-Tips Forums?

 

[GB67]

I got some information form internet.

The company ORTON also seems to get SIL 3 approval

http://www.truflointernational.com/company/newsDetail.asp?ID=151

Also Tyco annunced that they have doing "something" in this fild since 11/2006 but now still no any new.

It would be interesting to know te details of this certificate (operating conditions, valve range, scope of the certification -seat tigthness or simple operability-)

 

[Ashereng]

Terje,

I just talked to Mokveld regarding their SIL rated valves. Indeed, their valves have been tested, and certified for SIL 3 by TUV.

This is the first valve manufacturer that I have seen with a SIL rated valve (I haven't done a safety system in about 5 years).

Having said that, the SIL 3 rating certification is a test based probability certification.

Mokveld will NOT guarantee leakage rates, bubble tightness, etc. once the valve leaves their test lab.

In another words, once the valve leaves their lab, all bets are off if they leak or not. In many applications, there are only two states for a valve. Closed (meaning 100% closed - zero leakage) or not-Closed (meaning leakage).

I love this forum. Today, I leaned something new.

Thank you terje for the lead to Mokveld valve. I give you a star.

"Do not worry about your problems with mathematics, I assure you mine are far greater."
Albert Einstein
Have you read FAQ731-376 to make the best use of Eng-Tips Forums?

 

[NGiLuzzu]

Ashereng,
you talk about "valves (...) tested and certified for SIL 3", "a test based probability certification" and you say "once the valve leaves their lab".

Do you mean that Mokveld certification is based on results of tests carried out at Mokveld laboratory?

This sounds quite strange to me... I thought that the SIL calculations should be based on the data from the field, according to a proven-in-use approach (in my opinion, results from brand new valves are not much significant... and it is very difficult to really reproduce actual service conditions in a Laboratory).



GB67,
as far as I know, Tyco (for VANESSA Series 30,000 tripple-offset metal seated valves, in particular) developed "something" about SIL since 2002 and updated it in 2004... unluckily the document is not available on their website, but probably it can be found through your local Tyco Sales Offices.



Hope this helps, 'NGL

 

[Ashereng]

anegri,

The SIL certification is based on tests, and field data.

The tightness or leakage rate is a lab test.

Once the valve leaves the lab (or manufacturer), the leakage is not guaranteed.

"Do not worry about your problems with mathematics, I assure you mine are far greater."
Albert Einstein
Have you read FAQ731-376 to make the best use of Eng-Tips Forums?

 

[JimCasey]

Worcester Ball valves carry a SIL3 cert issued by Baseefa. There is also a SIL3 rating assigned to the Worcester/Norbro pneumatic Rack and Pinion actuators

 

[GB67]

JimCasey:

do you have such certificate for Worcester BV?
Or it's avialble on th web?

It would be interesting so see one of these certificate to check the working conditions and the perfomrmaces certified therein.

 

[GB67]

I found here in the website some documents form Worcester
but still no copy of the official certificate

http://www.flowserve.com/eim/v/inde...010VgnVCM1000005e191eacRCRD&vgnextfmt=default

 

[JimCasey]

I have the Worcester cert.
I think I have a way to get it to you.
The policy of this board is not to post email addresses (Mine or yours), and private messaging is not supported. However when I am not being a valve weenie I am a radio-control airplane nerd and I post frequently on RCUniverse. Go there, post me a private message, and I will email the Worcester SIL Cert to you, unless the management of this board finds that strategy objectionable.

"It would be interesting so see one of these certificate to check the working conditions and the perfomrmaces certified therein."
There is no such information referenced in the document.

 

[terje61]

This subject is something I am trying to get a grip on therefore a lot of remarks:

The discussion does focus a lot on certification. Happily I read that you want to read carefully the certificates. This is really important.
I was shocked to see that some official certificates are based on a number of cycles and that this is even used for low demand mode. I fully agree with Anegri that the cert’s should not be based on this. An example I know of:

http://www.seitz.ch/display.cfm?id=101374.

I cannot believe that TUV has approved this certification. The failure data to be used should be based on field data (dependable as IEC calls them or proven in use), thus these certificates should have never handed out basically. Therefore be aware of using simply certificates.

The MIL standard (which is for electrical components) actually shows multiplication factors to convert failure data from the laboratory conditions to real life. Because the laboratory failure rates cannot be used in the desert or when a valve has not moved in 12 months.

I reread the Mokveld report because of this discussion and they actually show different field data in their report for failure to close or failure to be tight shut-off. They seem to have failure data for 3 states for their valve (not closed / closed but leak / tight).

The Flowserve valves (the website does not say Worcester but says Schmidt) are unknown to me. What surprises me is that it concerns control valves. The IEC 61508 and 61511 are quite clear that control and safety shall be separated. A control element is normally limited to SIL1. It is also not clear if this includes the actuator. I would say that you cannot simply add the failure rate of the valve and the actuator. The connection itself might also contribute.

As Anegri said we should keep in mind that not only the valves, but the complete loop is considered (including the logic, sensors, barriers, solenoid valves etc.)

Hope this helps.

Gr.
Terje

 

[JimCasey]

As you say, you are unfamiliar with Worcester valves.
Worcester make ball valves. Although they do make a control version, the majority of what they offer are isolation valves. Worcester also make their own actuators which also are SIL-3 rated.

 

[jdraf]

Another possibility for safety shut off valves is MAXON which offers FMEDA reports by exida, a third party independent tester, for SIL3 capable valves to be used for FCI Class VI seat leakage. These are mostly fuel shut off valves for boilers and furnaces.

See here...

http://www.maxoncorp.com/Pages/product-Valves-Shut-Off