OSHA Process Safety Management compliance

[Ken Moore]

Any of you folks have to work with a process that's covered by OSHA PSM? How do you handle management of change when it comes to plc or dcs logic?

My plant has a general management of change procedure that's used for everything from changing a valve to changing plc logic, and it doesn't work that great for the plc logic.
If I need to change logic, be it one line or 100 lines. I have to document why, when, and the expected results. Then three other people have to sign off on the changes before they are made. Of these three people, one is the safety guy and knows nothing about PLC's, one is the department supervisor, knows nothing about plc's, and one is the engineering supervisor, once knew plc's, but hasn't done any programming in years. He always approves the changes.

So I have a piece of paper that okay's the changes, and it's been approved by three other people. My butt is covered. But, it's just a piece of paper, I could easily blow the place up, and the three guys reviewing my changes wouldn't have a clue.
I'm good, but I'm not perfect, someday I fear that I will overlook something important.

Does anyone have a better method?
Whenever we get audited we always pass, because we have a management of change program in place, and use it. It doesn't matter that it's ineffective. In my opinion software changes should be reviewed by people that understand the changes. Not the same people that review changing out a valve or pump.

Just ranting, but any and all suggestions would be appreciated.

Ken

 

[mordred]

Could be woirse the company I'm working with don't seem to know the meaning of documentation.

 

[shoeman]

Yet another topic near and dear to my little ole heart. I was in the CL2 industry for 10+ years. I am very familiar with PSM regulations and how they do not fit with PLC/DCS changes.

I made an unsuccessful attempt to change the policy I was working under. As you stated, all of the checks and balances were of no meaning for controls systems. What I tried to do was include a rider form that was a DCS/PLC check off sheet. If you checked certain things, you triggered a full blown MOC review. If it didn't rise to the level of a MOC, then at least you could show the change was considered, and reviewed by appropriate persons.

I was met with a great deal of resistance, because our PSM guru, was a lazy ******* who didn't want anyone in his arena. We had a line in the standard review that said, "software changes". If something were to go wrong, OSHA could come in and have field day with that. It could be argued 10 ways till Sunday.

I argued that minor changes to graphics were software changes, but not one requiring the signature of 4 to 5 people and several reviews. There was also an argument about tuning. Standard PID tuning was fine, but there was too much wiggle room when you started to feed forward, do adaptive tuning, or tinker with override conditions to get finer control.

My suggestion is come up with a form that fits your needs and see if you can get it incorporated. Most people don't know the right questions to ask if they are not a controls minded person. If you are changing a PSM covered process, *** cover is a good thing.

I wish you better luck my friend and hope you are more successful than I was. Please feel free to e-mail directly as I have more thoughts on this, but this post is already long enough.

[email protected]

 

[Mickey]

It has been said many times on this site PLC's are not safety devices and should not be used as such.
I don't know what OSHA (PSM) says about safety and PLC's but if you are relying on it for safety DON'T.
IMHO

 

[Ken Moore]

OSHA PSM's purpose is to prevent a catastrophic event, ie make a crater. We have independant safety shut down systems to handle this.
However, you could change plc code and cause an event that is less than catastrophic, such as adding 100% nitric acid to a carbon steel vessel, or charging two chemicals together that don't like each other, etc....

The safety shut down systems key off of high pressure, high temp, run away reactions, high LEL etc....

 

[mordred]

From what I understand of the OSHA. If you modify a piece of machinery whether its software or hardware. If those changes affect the operation of the machine. You then need to gain a certificate of compliance. This can be done through various companies such as CSA or
Underwriters are two examples.

 

[RSL]

I have worked with PSM in the past and have recently come under it's scope again. I have seen very restrictive and lengthy MOC's and very simple non-restrictive policies.

The one common denominator is that the folks that sign off on a change form usually have no clue about PLC or control theory.

I think the trick is in the testing and validation of the change. Meaning a change is requested stating we need X change in the process. We are going to make Y change in the control parameters. Finally we ended up with Z change in the process. Does X=Z or is it close enough? Is it safe?

I worked in the pharmaceutical industry, and when the chemical engineers were working on a process they rarely knew exactly how to get there in one step. They made educated changes and carefully validated their results for safety and desired effect. Many times this involved water batching and or dry runs. PLC changes, wiring, and piping should all be treated the same way.

There is rarely a day that goes by that most of us don't make a PLC change that could harm someone if we make an error or don't validate our work. Some processes are a lot more dangerous than others though. Working in a PSM environment always made me feel safer. You are constantly reminded to document your work and validate it.

Change in any process should not be a trivial everyday event. Only through taking the time to evaluate a change can we learn from it and guarantee process improvement. Safety is a pre-requisite. Many times change on the fly just is not smart. I always feared the temporary change request most of all. They were rarely thought through as well as a full blown change. Sometimes "No" is the only answer I could give if pressed for something too quickly.

I agree that it would be better to have these PSM reviews, PHA's etc. gone over by knowledgable folks. How many programmers are chemical engineers, or mechanical engineers? Some are, many are not. I think the team concept really has to apply for a PSM program to be effective. Everyone has to take it seriously and work together.

If your background is PLC's and controls then stick to that. Someone else needs to specify a change in the behavior of a system based on their knowledge of the process. Our goal is to safely give them the desired change through code and controls.

It is a less than perfect system, but I think it can work. In many ways I feel like you do. That the buck stops at the programmers desk.

I don't think I have offerd any solution to your situation, just my experiences with it.Everything above is just my opinion nothing more.

RSL

 

[leitmotif]

QUOTE FROM INITIAL POSTING
If I need to change logic, be it one line or 100 lines. I have to document why, when, and the expected results. Then three other people have to sign off on the changes before they are made. Of these three people, one is the safety guy and knows nothing about PLC's, one is the department supervisor, knows nothing about plc's, and one is the engineering supervisor, once knew plc's, but hasn't done any programming in years. He always approves the changes.

So I have a piece of paper that okay's the changes, and it's been approved by three other people. My butt is covered. But, it's just a piece of paper, I could easily blow the place up, and the three guys reviewing my changes wouldn't have a clue.
I'm good, but I'm not perfect, someday I fear that I will overlook something important.

Does anyone have a better method?
Whenever we get audited we always pass, because we have a management of change program in place, and use it. It doesn't matter that it's ineffective. In my opinion software changes should be reviewed by people that understand the changes. Not the same people that review changing out a valve or pump.

Just ranting, but any and all suggestions would be appreciated.
END QUOTE

I am sure glad someone else if raising this issue.
After 20 years in the Safety and Health business I bailed out because of having to honor the paper god. I was the safety guy who signed off on PLC designs and did not know a thing about them. It was a good thing I worked with good engineers who knew what they were doing. I did not sign off for the bad engineers though. I ALWAYS was there when we checked out the operation though.




Like he said he has the piece of paper so he is OK officially but he can still royally screw up.

This is somewhat like preventiative maintenance - computer says the PM got done 20 times - but why is the broken drive belt laying in the dust?
OSHA in my opinion is relying way too much on paper. The emphasis is on having the plan in the file drawer - it does not matter if you ever went to look at the equipment operation or process.

In house audits are a joke they are just more useless paper piled on more paper. OSHA inspections especially where these paper programs are required are a joke. Both rely on the paper being in place and do not go out in the shop where the work is being done.

WHAT TO DO IN YOUR CASE
1. Get the safety guy sent to PLC school. He does not need to know it all but needs to know enough to ask the right questions and understand some of the difficulties.
2. Get that safety guy out of his office and out in the field where he belongs.
3. Always have your work independently double checked - unless you have stopped being human and nver mak mistaks - IF YOU HAVE SUCCEEDED please teach me.
4. Check the revised operation out. I have made bunches of money from electricians who do not take the time to do this.


Dan

 

[GregPLC]

We incorporate PSM in our ammonia refrigeration system. The one thing I REALLY enjoy about it is that it opens up a line of communication with many people that typically isn't there.


Greg

 

[Ken Moore]

Well...
Sounds like I'm not alone in this concern. Many of you have the same issues that I do. I'll just keep on doing what I always do,
the best that I can and document everything I do.

If anyone has a software change procedure that they believes works, could you share it?

I will be working for a new owner soon, and perhaps they will have a better system. I'll update the post if I find out anything useful.

Ken