Thanks for listening to the EKT Interactive Oil and Gas Podcast Network.
In this episode, we welcome Joe Perino back to the podcast. Join us for this discussion on innovations in the digital oilfield, part of our Digital Oilfield podcast series.
Remember, our listeners get $400 off the registration price (use code EKT400) to the Upstream Intelligence Data Driven Production Conference. It’s happening in Houston on July 6-7.
About Joe Perino
Career experience includes:
+Process engineer in the chemicals, refining and pipeline sectors for Phillips Petroleum, Diamond Shamrock and Northern Natural Gas.
+Business development with process automation and technology suppliers Emerson, Honeywell and i2.
+Positions with KBC Advanced Technologies, IBM Global Business Services; Logica North America and Schlumberger Business Consulting.
In this episode, we discuss the latest innovations in digital oilfield technologies, oilfield cybersecurity, and startup companies in this space.
Digital Oilfield Podcast Series:
We put together this series of podcasts in conjunction with Upstream Intelligence to bring our listeners up to speed with the latest trends influencing the digital oilfield.
[1:15] Joe Perino background in upstream, IoT, digital oilfield
[2:40] IoT vs Process Automation
[4:45] Innovations & Startups – What you’re watching – HTC
[6:30] Predictive Analytics – use grows as sensor prices fall and automation grows
[10:20] The big players are in
[13:00] What upstream production operations can learn from refining and petrochemical segments
[15:55] Cybersecurity – trends and challenges
[18:00] Last thoughts – IT and operations continue to become closer and more integrated
Hi, everyone. Welcome to the Drill Down with Marty Stetzer. This podcast is part of our EKTI Oil and Gas Learning Network and brought to you jointly today with Upstream Intelligence in the UK.
Upstream Intelligence is the foremost provider of business intelligence and analysis for the Upstream Oil and Gas Community. They’re devoted to providing unique industry insight to drive efficiencies, reduce cost, and maximize production.
Today our topic is data-driven production. With an estimated global value of over $31 billion by 2020, the digital oil field is the oil industry’s hotbed of innovation. Now including big data analytics and the industry internet of things (IOT).
I’ll be speaking with Joe Perino, a longtime friend and industry veteran. Other podcasts with Joe are on his Sound Off channel on our EKT Interactive website.
Joe, as I remember, you were working on a digital oil field project for a major offshore operator when we first met at Schlumberger in 2004. We’re happy to have your input on this new and important part of the business. Joe, welcome.
Thanks, Marty. Good to be here.
Can you give our listeners your extensive industry background, especially as it relates to Upstream Digital Oil Field, IOT, and analytics?
Certainly, Marty. I have a long and varied career. I started as a Process Engineer in the refining and chemicals business. I then worked for a pipeline company Northern Natural Gas. Then after that I left and worked for about 20 years in the process automation industry for a couple small firms as well as Emerson and Honeywell.
That led me to a career in consulting, which took me to IBM, I2, and then a seven year stint with Schlumberger’s business consulting group. That’s where we met where I was working on a Shell smart fields project back in 2004 through 2007. Then, I’ve been working more on the IT side with the development and delivery of services to oil companies and oilfield service companies.
Then about 18 months ago I went back with my own management and technology consulting company. As you know, I’ve been associated with you now for about two years through EKTi.
Joe, thanks for that. You mentioned your extensive experience in the process control side of the business. How is IOT different from traditional control system technology? Or is it really different?
Well, IOT and automation are not quite the same thing. I think it’s an important distinction because when we started entering the digital oil field area, probably around the year 2000 with Chevron, Shell, and some of the other leaders, Saudi Aramco included, we were confusing the difference between automation and the intelligent field.
At that point, many fields were not automated and we needed to take that step to automate them in order to produce real time data.
IOT is different in that it’s about the interconnection of all kinds of devices and people at all levels across the business. It’s not strictly the automation architecture that you might find in an ISA 95 model. Now IOT has opened things up.
Everything out there, whether it’s static or dynamic will have a digital personality. In the case of a device that’s a dynamic device, like a valve or a sensor, obviously they’re going to be producing data. It’s going to be connected to other devices and that may be direct connected, that may be wirelessly connected, that may be connected by the Cloud.
Now the important part of this thing for process automation firms is that this opens up additional capability for them to monitor and control what is going on in the facilities, which is where most of the systems are installed. They’re now using IOT and the Cloud and they brought in predictive analytics to be looking at the health of processes and equipment to improve plant operations. That’s the nuance there.
You’ve seen developments in probably your 13, 14 years of watching this, Joe. What are some of the innovations that you’re now seeing out there? Especially in some of the startup organizations?
Yeah, so I’m glad you mentioned that. One of the things that I’ve been involved with for the last year or so is the Houston Technology Center, which is an incubator for startups. They help energy firms, IT firms, nanotechnology, life sciences, and so forth.
I got introduced to HTC when I was at IBM, and I started following IOT a lot more closely. In particular, all of these startup analytics firms that have popped up and there are literally hundreds of them. I keep a list of around 12 to 15 companies that have surfaced that are addressing the oil and gas space. This is in addition to the larger names like IBM and EMC, Amazon, Microsoft.
It’s interesting to see how these firms are coming to market, where they’re focused, what they do, what value they’re bringing, and how they’re getting traction in the market and, frankly, whether they’ll survive.
We’re still very early in the IOT space and this use of predictive … Pardon me, I should say advanced analytics and big data for improving the oil and gas business.
He said the real challenge seemed to be in the analytics side. Not in the sensor side.
That we’ve had information available, like you’re suggesting, in process control for decades. Now that the new tools are enabling us to analyze that data is the real innovation. Are you seeing the same thing in your side — especially looking at some of these new startup organizations?
Well, yes, I am but let me comment on Tony’s observation. One of the things that has changed is that 10 years ago, when I was working with Chevron and Shell, they would have many fields that could not justify automation because of the cost.
For example, a full transmitter then might be $1500 per well. Now sensor technology has gotten very inexpensive. If you’re just going to be monitoring a production battery you can now buy sensors that may be $100 to $150. They give you a pretty reasonable reading in terms of accuracy and you’re able to actually collect that data wirelessly now. There’s almost no excuse for not automating most onshore fields. Most of the offshore fields are all automated because of their size and the investment.
Now Tony is right about the analytics side. We’ve been doing predictive analytics, and I use the word predictive in this case, for quite some time. About 15 years because that space, particularly around rotating equipment, is where things got started. Companies like Sass and a company called Smart Signal, which was acquired by GE and it’s now being migrated to the Predix platform. Those people have been putting neural nets and machine learning algorithms on rotating equipment like compressors.
Now it’s spreading to all kinds of things that it’s being applied to.
· It’s being applied to artificial lift mechanisms.
· It’s being applied into enhanced oil recovery.
· It’s being applied to frack stage optimization so with completion design.
· It’s being applied to safety and security and cyber security. All kinds of areas where this is going.
The ability now of data to be collected, moved to the Cloud, analyzed, and moved back. Or in fact with today’s computing power, we can put a server on the size of a quarter and put it on a machine with memory and actually run analytics there. That’s what we call edge computing.
When you combine that with what’s going on in the Cloud you have this hybrid computing going on. If you have a network of all of these things going on, this creates this term called fog computing. All these new buzzwords that we need to learn to describe what it is that we are doing.
This thing has just exploded. It is a challenge because we’re not yet at a place where we have what I’d call a mature set of platforms and solutions that users can buy that are pre-configured for the application they want.
For example, a pull down menu with all the artificial lift mechanisms on it. We’re still in the point where we have platforms and solutions which require customization if not outright programming. Usually in Python or R which is the language that people use for machine learning.
We’re still very early in this phase. I think there’s still a fair amount of hype but lots of people are testing lots of things and so this is a very rich area that we expect to grow significantly in the next few years.
Joe, you mentioned earlier GE. I noticed that Jeff Immelt, the president of GE, said that the industrial IOT was the key to the future for GE. Now they’ve been doing remote diagnostics on in their power systems divisions for years. Do you see some of this being more accepted now that even the big players like GE and probably the Emersons are making more of an inroad to help the operator understand how to use all this information?
By all means, GE probably started that in their industrial area with trains and aircraft which are monitored continuously. Now they’ve got this platform where they are migrating a lot of their independent applications. Smart Signal, I mentioned, but they’re also looking at some of their software that they use for artificial lift management. You and I both know that because of the training we did for GE.
It is moving over. It’s being applied. We know that BP is using this. I think they’re going to be speaking or, at this point by the time the listener hears this, they will have made a presentation hopefully at our Data Driven Production Conference that’s coming up in early June of 2017. Yes, I think we’re at the point where this is about to take off.
But I would say this, overall, I don’t think we’ve crossed the chasm yet on the maturity of these technologies. There’s still a lot of testing. Aside from GE there’s only a couple other platform vendors with a focus on oil and gas. When I mention that there were 100 or 300 —there’s so much of this that’s so much farther down the road in other industries. As usual, the oil and gas industry is behind in it’s uptake.
In defense of that, we have a tougher set of problems. More difficult data with data conditioning, data quality, time lags, and all kinds of other things. Someone asked the question the other day, “What would you rather have? A better machine learning algorithm or better data?” The answer is better data.
Yes, this whole analytics thing is taking off but it’s also refocusing our efforts on getting good data from all the right sources so that we can actually use the analytics for better decision-making.
It’s interesting you mentioned other industries, Joe. When we spoke with Tony he said that they had seen some applications in mining and I know coalbed methane is part of our oil and gas industry but it’s a totally different animal than the normal oil and gas production.
I want to go back to the part of the industry that has always had good solid process control. Is there something that the production operators can learn from the way that refineries and petrochemical plants are managed?
Well, two or three things. Refining went through a modernization period from the late ’80s into the early 2000s. Here they did a lot of standardization of their technology stack, if you will, which is not only the architecture around the facility but the software tool set that goes with that. Until recently, they seemed to be pretty satisfied with that. It’s pretty mature.
Upstream oil & gas has come from where organizations basically let the asset or the divisions choose what they want. So you had all kinds of different installations, different technology stacks. We then went through and are still going through a phase where IT wanted to basically standardize this; and in some cases centralize it because these organizations are so federated. They’ve had some success in that.
Now we have IOT and data analytics and this is making us look back again at the business and operating models of these firms and ask the question, “Should we be thinking about what tech stack we’re going to use — even before we buy an asset — because our ability to manage it is now going to become more dependent on this use of advanced technology?”
Instead of just buying the asset and plugging it into the same old architecture with the same old tool set, we now have the ability to let’s just say customize that within some degree and scale what we need to the business benefit we expect to get from the asset. This requires more upfront planning and thinking about how to apply this from the beginning It has potentially more complexity for IT because they have to deal with the IOT situation where data is coming from all kinds of directions and flowing back to all kinds of directions.
In particular, cyber security now becomes paramount; because you’ve opened up the access to control systems and other data sources that heretofore were not inter-connected like they can be now with IOT.
What are some of the innovations you’re seeing in the cyber security area, Joe? Or is it being developed in parallel with some of these new innovations?
Well, in a cyber area most of the activity when you think about it is what we call end-point detection — which is how to detect the bad guy once they get inside the firewall or inside the network? That’s of course important because you want to limit the damage. A lot more focus is now being placed on being able to detect the bad guy before they get into the network.
That is where they’re using artificial intelligence, machine learning, and other techniques to try and look at patterns of how the door is being knocked upon and try to say, “Hey, we may see some activity here that looks suspicious and we’re going to block it before it even has a chance of getting through.” That’s one thing.
The other thing is that it’s forcing people to look at redesigning their hardware and making it cyber secure from the ground up rather than relying on a network only to block the bad guys. We saw this with Stuxnet a few years ago. Once that virus got into the programmable logic controllers in the nuclear program it shut down their centrifuges. The answer is that the PLCs themselves, the control system, has to have the cyber proof, if you will, from the beginnings – so that it’s not just relying upon the network.
This is forcing everybody to rethink how they’re designing and engineering things and how they’re manufacturing too. It’s a challenge for automation companies who heretofore had proprietary systems that once you put a wall on it – you limited the connectivity were relatively isolated from the outside. Well, now with IOT and especially with Exxon Mobile’s push toward open systems, we’re going to see more openness and when you have more openness you need more cyber security.
Joe, this has been an amazing amount of information in a short space of time,. I think our listeners will really enjoy the importance of understanding the analytics, the many new companies that are out there leading the charge, and especially the importance of cyber security.
What I also heard was now IT is taking an important role right from the design engineering, through the equipment design; and the way the process control systems like we had in the refinery are being operated. Would you like to add anything else for our listeners before we wrap up?
Yes, Marty. You made a good point there. It’s now almost impossible to separate IT functions from the rest of the business —in particular, from OT or operational technology meaning the control systems people. It’ll be interesting to see how organizations adjust themselves and build their capabilities to address this.
I know of at least one major organization that merged IT and OT. That’s one thing I think we want to keep an eye on. I think in my opinion is that that makes sense. I think that most organizations are lagging on that in terms of merging their organizations. I’ll leave it at that.
Joe, thanks again.These insights will surely be valuable to the EKTI and Upstream Intelligence listeners.
Marty, thanks so much for having me.
Enjoyed listening to you, Joe. For more information about the important oil and gas industry, be sure to check out our free oil 101 series at www.EKTinteractive.com.
Again, thanks for listening.