10 replies to this topic

[CSNK]

Dear all,

 

I am checking the re-vamp conditions for the flare system of a unit.

 

May i know how the administrative controls and instruments credits can be taken so that some of the huge modifications can be avoided during power failure scenario.

 

Any reference from the experts is appreciated regarding administrative controls and instrument credits for flare system basically during the major relief i.e total power failure.

 

Thanks.

[Dacs]

What do you mean by "administrative control"?

 

If you mean the operator response time, then yes (and speaking for my experience only), we took credit on how quickly can the operator act during HAZOP.

 

And you can also credit instrumentation that provides redundancy (like 2 separate LTs working independently).

 

But that's something your HAZOP charman has to set at the beginning of study.

Edited by Dacs, 09 May 2013 - 01:37 AM.

[CSNK]

Yes you are right Dacs, Administrative controls means operater intervention (like running turbine drive always or during partial power failure run two pump around on different electrical control panels.)

 

If by taking credit for the instruments (i.e it will get reset to some desired value as soon as upset occurs), relief load reduces can i take credit of this instrument controls and yes there is no reduncdancy for instrumentation. Unless i provide new XV's to improve reliability of the whole system.

[Dacs]

Administrative controls means operater intervention

Just to guide you on the approach that we had during HAZOP. We normally take 5 minutes operator response time and if the upset can be alleviated within this time, we take credit for it, otherwise we don't.

 

If by taking credit for the instruments (i.e it will get reset to some desired value as soon as upset occurs)

That's fine and dandy but what if the instrument itself is at fault?

 

So I don't think you can take credit for it unless you have redundancy in place.

[CSNK]

Thanks for the prompt reply.

 

You mean to say i need to provide some redundancy for the single instrument i.e either provide new XV.

 

Just for curiosity, isn't failure of instrumentations along with power failure a kind of double jeopardy.?

[fallah]

Just for curiosity, isn't failure of instrumentations along with power failure a kind of double jeopardy.?

 

chiks,

 

No, because while you might not having instrumentation failure you cannot take credit for favorable instument response.

[Bobby Strain]

Owner/operator can do whatever they can get their insurance provider to accept. An engineering company, however, must recommend adherence to accepted guides and standards. I have never recommended load reduction by operator response. It is simple enough to use reliable and proven instrumented systems as safeguards to mitigate relief from a number of sources. Don't get inventive and make up rules for a hazop.

 

Bobby

[Dacs]

Bobby: Not that I disagree with you, I just shared what has happened on the HAZOPs that I attended. And yes, we all agreed that operator response (for some system) can be taken credit.

 

It might not have happened in your HAZOP study, but it has happened with mine.

 

For the OP, we can't make the decision for you. It's up to you how to proceed (and it's flare system, that has to be addressed very carefully).

 

EDIT: Just to add, I do remember we talked about it prior our HAZOP study. We concluded that while human error is always there in plant operation (and in fact it was one of the source of deviation in our study), a certain amount of confidence can be attributed in a properly trained operator abiding to implemented (and well documented) SOP (otherwise why would we allow people to operate a plant who aren't fit in the first place?).

 

Some people might feel differently on this (and for a good reason) but the team felt that not including operator action was deem too conservative.

I attached a snippet of one of the HAZOP worksheet that I attended to show how we took credit for operator action.

 

Again, take everything at face value and don't take my post as an endorsement on how you proceed with your own study of risk assessment.

 

EDIT2: To the OP, have you considered HIPPS in your study?

 

EDIT3: I attached another snippet of the HAZOP worksheet where we indeed took some credit on operator action. The 1st attachment was in fact shows that operator action cannot be credited.

Attached Files

•  attach.jpg   210.35KB   7 downloads
•  attach 2.jpg   255.6KB   2 downloads

Edited by Dacs, 09 May 2013 - 11:31 PM.

[Bobby Strain]

Dacs,

 I'll point you to the BP Texas City refinery where 15 people died 8 or 9 years ago as a good case study. This is extreme, but I think it points to BP management lack of attention to safety issues. This plant has had lots of fatalities over the years. Let's hope Marathon turns things around now that they own the plant. BP continues to talk one thing and do another. I am retired now, but I was involved in lots of plant builds. It seems that those most inherently dangerous have the best safety record, with the exception of Ethylene Oxide facilities. My personal observation over the years is that 99% of incidents that could be catastrophic cause only minimal damage. Why? Pure luck, but sometimes good design. Once it gets built it usually won't improve. Budgets won't allow it. Until something goes wrong. As for operator response, I assume that the operator will always take the wrong action. But a good HAZOP session should uncover problem areas that require attention. My experience as an engineering contractor executing a client project is that the client team is mostly low-level with little experience. Or they want to focus on design detail. It's a hard business, and my hat's off to those companies that specialize in facilitating HAZOP sessions.

 

Bobby

Edited by Bobby Strain, 09 May 2013 - 07:42 PM.

[Dacs]

We're on the same page as far as safety is concerned and I think plants being built recently are required to do safety/risk assessment anyway (whether for legal and/or liability reasons).

 

And I don't think no one here implies that the OP has to cut corners to conveniently avoid expensive solution to his problem. I bet that regulations will dictate that whatever revamping his plant is doing has to be subjected to HAZOP and (God forbid if otherwise) the decision to do safety modification doesn't rest to a single person.

 

And I certainly don't imply that you can rely on operator credit at all times because the decision was only made after the Owner made their point that they have a certain level of confidence in how they select and train their operators.

 

Fortunately for that particular study, we didn't rely too much on operator action as redundancy are in place to supplement operator action.

[CSNK]

Just for curiosity, isn't failure of instrumentations along with power failure a kind of double jeopardy.?

 

chiks,

 

No, because while you might not having instrumentation failure you cannot take credit for favorable instument response.

 

Thanks for the reply.

Dear Fallah,

 

If say my two (out of 3) Pump around pumps (which help in heat sink) are running and due to partial power failure one of the pump trips.

 

In such case the second pump will also trip (due to overload) or will not trip?

 

If say it trips, then if i require to run the second pump can i take credit that the controller at pump discharge will reset to some desired value (so that pump is not overloaded)?

 

Dear Dacs and bobby - thanks for your inputs. ( now operator credit is ruled out and we wont be taking any operator credit)