Over a million developers have joined DZone.
{{announcement.body}}
{{announcement.title}}

Oil Refineries Using Digital Eyes and Ears to Keep Gas Flowing

DZone's Guide to

Oil Refineries Using Digital Eyes and Ears to Keep Gas Flowing

See how IoT has impacted the oil industry in the past few decades.

· IoT Zone
Free Resource

Address your IoT software testing needs – improve quality, security, safety, and compliance across the development lifecycle.

The job was quite a bit different for engineer Mike Roe when he started working at an oil refinery 37 years ago.

Back then, turning petroleum into gasoline and other products demanded that technicians constantly keep a vigilant eye out for the slightest clue that something was wrong with their machinery. An unusual sound, a strange gauge reading could be the first step to a catastrophic failure that could shut the facility down. To stay on top of things, they'd regularly send samples to the lab for analysis, record data on paper charts and manually manipulate control buttons and dials that governed the process.

Things started changing a bit in the 1970s for Roe, a process engineer, and his colleagues. His refinery installed "newfangled" instruments called gas chromatographs to the plant's control boards. Automatically separating and analyzing liquids during the production process, the chromatographs helped computers constantly monitor petroleum inputs and products, adjusting controls better than human operators could to keep things running smoothly.

"It was hard to accept those new machines at first," Roe says. "They worked well; you just had to get used to them."

Even with this technological upgrade, though, the work was still often tremendously taxing while being tedious to the point of boring at the same time. There was no crystal ball to tell when a machine was about to break down. Samples still needed to be collected and sent to the lab. Then there was the part of the job that they called "The Rounds," constantly keeping an eye on the same streams of data coming from the refining process in the same order. It involved monitoring huge amounts of numbers that rarely deviated significantly or unexpectedly.

"Most of the time, when things are running well from the process side, the job is pretty boring because you do not have to make changes," he says. "You're just walking around or you're looking at the data."

But there's a danger to staring at the same information for too long, especially when machinery has been behaving badly for a long time—you can become desensitized to when you really need to take action.

"If you accept something that's not normal—when you become conditioned to a faulted state, it could be dangerous," says Roe. "A lot of times if you're in the field and you get a faulted condition, an engineer will say, 'It always runs like that.' But that could be bad if you don't know when it's going to fail."

 The digital brain can send out an early warning to maintenance crews or automatic controls to alter the flow of steam through a pipe or tighten a bolt somewhere, for instance, that effectively fixes a machine before it needs to be taken out of service. 

And then there are the other deficiencies present when keeping petroleum processing on track the old-fashioned way. While constant watchfulness may let an engineer catch a maintenance issue when it is still relatively small, it doesn't give any insight into problems that have yet to occur. In other words, current refining conditions can be understood but seeing early warning signs and being able to better predict when a machine will start breaking down is always just out of reach.

Now, though, things are once again changing for process engineers like Roe, who eventually moved from working in refineries to producing high-tech software at GE Digital that helps these integral facilities run better.

Just like when chromatographs appeared in the industry, technology is coming online that is offering a step forward. This one is shaping up to be a giant one. Refinery machinery is being fitted with new sensors that constantly monitor conditions like vibration, chemistry changes and temperature. This information is being fed into smart software running on GE's Predix platform, which can analyze it to tell when a machine or process isn't working as well as it could. The digital brain can then send out an early warning to maintenance crews or automatic controls to alter the flow of steam through a pipe or tighten a bolt somewhere, for instance, that effectively fixes a machine before it needs to be taken out of service.  

Roe says the Industrial Internet and predictive analytics software like SmartSignal mean that refinery operators won’t have to look at every chart, every day, every minute. The instruments and software will keep a vigilant eye, instead, telling technicians if they've got problems. 

"Plant operators will be able to avoid some of the random failures by picking up on more precursors—some kind of sign that there's a problem, which nobody has picked up on because they didn't hear a change in sound or they didn't see anything different," he says. "Now, the software will pick it up and catch stuff early so they don't have major failures."

Accelerate the delivery of high-quality software in the connected IoT era through an integrated analysis, testing, security, and analytics platform

Topics:
digital ,ge ,iiot

Published at DZone with permission of David Ovadia, DZone MVB. See the original article here.

Opinions expressed by DZone contributors are their own.

THE DZONE NEWSLETTER

Dev Resources & Solutions Straight to Your Inbox

Thanks for subscribing!

Awesome! Check your inbox to verify your email so you can start receiving the latest in tech news and resources.

X

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}