Chevron says safe refining boosts throughput

[Forum Administrator]

Posted on behalf of Clyde Young

Thu Mar 26, 2009 3:40am IST
SAN ANTONIO, March 25 (Reuters) - A process safety program has improved the reliability of Chevron Corp's 2.2-million barrel per day (bpd) refining system and yielded a gain in throughput equivalent to a new refinery, Jeet Bindra, Chevron's president of global manufacturing said on Wednesday.

"We found if you operate your refinery reliably and safely, you can produce as much as one new greenfield refinery," Bindra said in a speech to the Hart World Refining & Fuels Conference.

Chevron has reduced unplanned operational interruptions by 40 percent in the past two years through regular planned maintenance and preventive maintenance, he said.

Small increases in operational reliability at each refinery from the safety program have equaled a capacity gain between 100,000 and 150,000 bpd, Bindra said.

"Now is not the time to cut corners," he said. "You need to know the condition of your equipment."

Earlier this year, the U.S. Chemical Safety Board called on refiners and chemical plant operators to not reduce maintenance as they cut costs to counter falling refining margins.

Cost-cutting in equipment maintenance was seen as a factor in the deadly March 23, 2005 blast at BP Plc's Texas City, Texas, refinery, the CSB found.

Chevron is also looking at increased operational automation of operations to improve reliability at its refineries, he said.

"We think the technology is available today to react to upset conditions without human interaction," Bindra said.

Operational control has been centralized in the past two decades in refinery command centers with operators monitoring computer screens and reacting to malfunction alarms with gauges and control valves on refinery units as back-ups. (Reporting by Erwin Seba; Editing by Marguerita Choy)

[Forum Administrator]

Posted on behalf of Gordon Sterling

"We think the technology is available today to react to upset conditions without human interaction," Bindra said.
 
Always get concerned when I see these type of comments. What do you process and controls folks think?
 Thanks
Gordon

[Forum Administrator]

Posted on behalf of Kindra Snow-McGregor

Hi Gordon,

I personally approach that idea with serious caution.   I think we need to understand exactly what it is that they are proposing to have a strong opinion one way or another, but yes I agree with your concern. 

Kind Regards,

Kindra

[Forum Administrator]

Posted on behalf of Perry Lovelace

Good point, Gordon,

But the real point of this article and the one I’ve attached is that careful, planned intervention by Maintenance, Operations and Engineering people will result in fewer upset conditions and greater safety performance.  I agree with you that the attitude of automatic intervention to upset conditions is worthy of concern; most safety incidents have human causes at their root.  Waiting for something to happen to ‘React’ to is the most expensive way to run any kind of operation; imagine if we waited until our cars ran out of gas to find a gas station!  Absurd…

Also note near the end of the article that the largest single source of defects is in the Design phase of a plant.  This is not to point the finger at the Engineers, but to note how vital expert design is.  Any defect that can be foreseen and eliminated in the design phase will not need to be reacted to later and will permanently raise the reliability of the equipment.

It is a sad commentary that our Reliability and OM courses, that teach these kind of plant management skills, do not show higher enrollments.  Any suggestions to improve the visibility of our Reliability and OM courses are welcome.


Perry

[Users must be logged in to view attachments]

[Forum Administrator]

Posted on behalf of Tim Armstrong

Perry,
Your last two paragraphs are interesting.
First, design - yes, this is where a lot of problems develop because it is more difficult to visualize something that is not built yet than to comment on the finished product.  In addition, my experience in the E&C business is that the low cost to the E&C company is the goal which maximizes profit.  Therefore, we end up with less experienced engineers doing work and the more experienced trying to prevent major problems.
Second, Reliability and OM - These are seen as closing the barn door after the horse has run away - in other words, after the fact reaction; oops I ran out of gas once therefore I must fill the car with petrol.  The marketing has to be that this is a predictive approach that does not rely on past failures to improve reliability.  Maybe inspection, maybe more effective and realistic PHA, maybe adherence to standards.
Just some random thoughts.
 
Tim Armstrong
Instructor/Consultant
John M. Campbell & Co.

[Forum Administrator]

Posted on behalf of Ronn Williamson

I want to address Tim's point about seeing Reliability and OM as closing the barn door after the horse has run away.  That perception goes back to the reactive nature assigned to those roles.  Twenty years ago in manufacturing a movement was begun to "design for manufacturability."  I think you are talking here about "design for safety and reliability." 
 
The product design engineers in manufacturing businesses didn't know how manufacturing really made products and didn't have to know - what they had to do was bring the manufacturing folks into the design process.  Same thing here with maintenance and reliability expertise - educate facilities design engineers to avoid all they problems these folks are seeing and reacting to from the last design.  One of my key objectives as a procurement executive was to bring component suppliers into the design process too for their operations knowledge of the components we were designing for them to make.

Ronn Williamson

[Forum Administrator]

Posted on behalf Darrell Carver

Actually, the technology has been implemented for the last 15 years or so. With newer DCS upgrades the increased computing power gives us the possibility of automating reactions to events and allows it to be more easily implemented. If the project also includes alarm optimization, as it should, then those process variables and associated alarms identified as critical and needing immediate operator response are candidates for automation or expanded logic in the DCS, safety system, or unit controls.
 
Multivariable controllers are another example that has been around for 15 years or so. The interactions of all control loops are evaluated such that as one control loop changes, the other dependent loops are adjusted also to limit their setpoint variances allowing the facility to operate nearer its maximum capacity.
 
Think of stability controls on your auto. It uses motion sensors, wheel and brake sensors, engine parameters, etc. to determine if your auto is sliding and automatically applies the necessary braking to the proper wheel to bring it under control.
 
The key to any automation of responses to the process is having the proper measurement inputs and outputs. If you don't include them in the initial design, then it is difficult to justify after startup unless economics can be shown.
 
Darrell Carver
Instructor/Consultant
John M. Campbell & Co.

[Forum Administrator]

Posted on behalf of Andrea Mangiavacchi

To extend the analogy to the last point, just as product designer did not have to become experts (although a basic understanding did help) of the manufacturing process, but had to bring manufacturing people into the design process, the same can be said about operations and reliability. While it does help if the designers have a better  understanding of what happens in operations, ultimately the real improvement occurs when the people that will have to operate a plant have a chance to influence and shape its design.
 
Andrea Mangiavacchi
Offshore Instructor

[Forum Administrator]

Posted on behalf of Clyde Young

In order to insure that those who will have to operate and maintain the plant have a chance to influence the design a Hazard and Operability (HAZOP) study should be performed early in the design phase and then again later.  The HAZOP should be performed with an interdisciplinary team and should always include operations and maintenance personnel. 
 
I've participated in HAZOPs where the design engineer didn't include enough valves to allow operators to operate and maintenance to maintain.
 
All part of a management system that yields production results.
 
Clyde

[Forum Administrator]

Posted on behalf of Perry Lovelace

Another situation that influences the optimization, hence safety, of plants and personnel is the fundamentals of budgeting.  Many, if not most, designs are performed under a capital (CAPEX) budget where the cost ceiling was set in the very early stages of the project and engineering projects are rewarded for coming in under budget and on time.  This often leads to short cuts and pushing design to less experienced (cheaper) engineers.  On the other side of start up, the operation of the plant comes under the Operating budget (OPEX) where cost minimization pressure is always felt.  That’s why, in an era of lower sales prices, maintenance programs (Reliability, CMMS, training, Condition Monitoring, etc.) and personnel are the first to get cut.  Of course, with the Laws of Physics always in force, this reduced attention to maintenance starts ‘time bombs’ that frequently result in unplanned outages or failures later on.  Apply this philosophy over a number of years and someone will wind up holding the bag in a neglected plant.  Hence the need to invest new capital to ‘modernize’ or sell off the asset to a lower cost (or National) company and the downward spiral continues.

 

All of us have experienced plants that are well operated and maintained and those that have been neglected.  I’ve been nervous going into some plants for fear that today might be the day the ‘time bombs’ go off…

 

Great discussion,

Perry

[Forum Administrator]

Posted on behalf of John Berry

Yes, I've participated in a HAZOP study on a 750 mile liquids pipeline coming out of Wyoming, over the Divide, and terminating in Kansas.  I'm a pipeline engineer and not a plant engineer and this was the first formal study that I had been involved with.  Very impressive and creates excellent communications.

[Forum Administrator]

Posted on behalf of Dennis Perry

Interesting thread. I was a manufacturing engineering manager for a few years and the same concepts reared their ugly heads. One major lesson was that you can not inspect quality into a product It must be there from the beginning. Another lesson was to get the maintenance team on board early in design to keep from building a product that could not be easily fixed.

Dennis Perry
Instructor/Consultant

[Forum Administrator]

Posted on behalf of John Kanengieter

This is indeed an interesting thread and speaks a loud lesson to understanding the concepts of organizations as systems and system dynamics. When applied to efficiencies it means depts working collaboratively on organizational goals.  When applied to safety it looks beyond the point of failure (i.e. a guy who turns a valve the wrong way), and looks for the system dynamics that allowed the failure to even take place. It's why safety will always be more than a check list....it's also an attitude.

Interesting stuff-
john k

[Forum Administrator]

Posted on behalf of Peter Williams

This is the same message that keeps coming up in Lean, Six Sigma, or any other performance improvement program - correcting problems is expensive, inefficient, and sometimes ineffective, so we need to get the key stakeholders involved early in order to ensure that we minimize late changes. Whether the context is process safety, engineering management, or quality improvement, the theme is the same.
 
Peter Williams

[Forum Administrator]

Posted on behalf of Clyde Young

It's important to always remember that latent failures are ALWAYS there.  AND we generally build them into the system or process. 
If we can learn to identify these latent conditions early on, the swiss cheese model doesn't rotate around to where all the conditions are lined up for a failure. 
 
Good stuff,
 
Clyde