Energy is undergoing an important transformation driven by the introduction of renewables, deregulation, and increased attention to climate impact. This shift is assisted by new advances in technology, such as cloud computing, IoT, and big data. In this podcast episode of Tech Lightning Rounds, Beth Kindig speaks to the company Chevron, where technology is being leveraged in transformative ways including with the use of drones and Augmented Reality. She also speaks to innogy Ventures, a company that has been supplying central power for over a century and making strategic investments to take on the next era of decentralized power production with data interoperability. Her third interview is with Schneider Electric on what is involved with distributing power for edge computing.
- Chevron’s Gorgon Project, which uses advanced engineering to maximize production and distribution of natural gas
- How Chevron uses data to move fluids safely and effectively, and to monitor equipment
- How innogy Ventures is investing in the future of decentralized power
- The future of data-driven business models and data interoperability
- Challenges in distributing power to 5G microdata centers
- Schneider Electric’s plans to assist a more open architecture for data centers
00:01 Beth: You have a project called the Gorgon Project. What have sensors and advanced process controls done to contribute to value there?
00:13 Michelle: What’s amazing is sensors now are everywhere. If you look at what’s happened over the last few years, the price and availability of sensors has dropped. We have over a million sensors across Chevron right now. We expect that over the next five years, we’re gonna see that to grow at a rate of at least 100,000 sensors per year. What we can do with this data, especially given the changes in cloud technology, the changes in artificial intelligence, machine learning, is amazing. When I look at advanced process control… And if you’re not familiar with advanced process control, think of it like adaptive cruise control on your car. And we can now put that on our plants. We’ve been doing it for years in downstream, but with the rising of sensors, you can do it even more. For example, in our LNG facility in Australia, we were able to use advanced process control to create $240 million of value in 2018. That’s huge. When you can do that, again, across all of your big plants, it makes a huge value difference.
01:25 Beth: Can you give us a little bit of background about that project? What is it exactly, and why did you initiate that project?
01:33 Michelle: That project… LNG, liquefied natural gas, it’s a way of moving energy across the ocean on boats. We have a very big plant in Australia, and one of the things we found is… Similar to your car, if you can run your car with cruise control, it gives you more control, it allows you to stay within the control limits, the speed limit. With adaptive cruise control, which some of us have on our new cars nowadays, it allows you to sense the car that’s coming in front of you and automatically slow down. When you can put that sort of control on a big facility, it allows you to run it safer, it allows you to run it more reliably, it allows the operator to have more control of the plant, because they’re focused on the one-offs when something… An anomaly happens. It allows the plant to run much, much smoother and safer. And by having a smoother and safer running plant, you can get more throughput out of the plant, which translates into more value.
02:42 Beth: Most companies today are data companies. What kind of data do you use to improve operations? What does that look like?
02:51 Michelle: It’s funny. I’ve been with the company for 19 years, and for 19 years, I’ve seen that Chevron is a data company. When you look at our fields across the world, in Bakersfield alone, in San Joaquin Valley, we have thousands of wells. We have always tried to collect as much data as we could on our wells, but that expands much beyond wells. If you look at our entire value chain, if you start from the rock itself, we have sensors that are down inside the walls telling us how the wells are producing, how the fluid’s producing. We’ve got sensors at the surface telling us about our pressures, our temperatures, acoustics, the sounds. That goes on to the pipelines. It allows us to make sure that we are moving our fluids safely and effectively. That goes all the way into our plants and then our refineries, and then even our point of sale. We are collecting data on absolutely every aspect of the facility.
04:00 Michelle: It allows us just loads and loads and loads of data and information, which we can bring together to do advanced analytics. One amazing example, we have big compressors that help us compress our gas, whether they’d be offshore or onshore. We now have a center in Houston and a few others around the world which take that data in real time, analyze it, and let the fields know if they see an anomaly, so we can actually take the equipment down before it breaks. It’s like getting a warning light in your car, except we’re doing it on big equipment. We saved $40 million last year monitoring our machines.
04:57 Beth: And do you use drones at Chevron?
05:00 Michelle: Yes. Our digital program spans lots of areas of technology, like I mentioned before, artificial intelligence, machine learning, Industrial Internet of Things. Drones and robots are another big piece. We recently showcased one of these really cool drones which you can actually use inside of a confined space. By confined space, I mean inside of a tank or a vessel. And what’s amazing is a process for collecting data, which used to take multiple people going down into a tank to collect the data, I can now send a drone in and have it send data to the people outside of the tank.
05:42 Beth: Do you have any examples on augmented reality that you can talk about or…
05:45 Michelle: Oh, yeah. Probably the coolest one I’ve seen is through our partnership with Microsoft, and it’s called the HoloLens. The HoloLens allows you to be out in the field, and I want you to think of a field location offshore, you’re out in the middle of essentially, nowhere, and you put on this HoloLens, and you’re paired with a person back in Houston who’s also wearing a HoloLens. You, out in the field, can be looking at a piece of equipment, and the person back in the office can see what you see. And when they point their finger in front of their HoloLens, you can see their hand coming out in front of you and they can pull up an instruction manual, a repair manual, and it shows up in front of you. You can have somebody out in the field troubleshooting on a piece of equipment, getting real-time support from one of our subject matter experts that’s sitting back in Houston. No need to fly offshore, no need to worry about visas and travel. You’re able to get that expertise real-time through these, basically, augmented reality goggles.
07:06 Beth: Globally, what are the energy needs in the regions where you invest? Can you tell us, globally speaking, what do you see going on with energy, whether that’s Europe, Asia?
07:16 Florian: Yep. The reason why, after 110 years, we need innovation, is… What’s happening right now across the globe, in the US, in Europe, in Asia, is a move from a very central energy system, very centralised energy system, where you have huge factories that produce a lot of electrons that are connected with thick cables that then ship the electrons via cables that are getting smaller and smaller, as when you come to the end customer, and then sell them to end customers. That’s the old model. This is shifting rapidly to a de-central model where you have renewable generation on rooftops or community solar or industrial-scale solar or wind farms that are just everywhere. Now this old model doesn’t work anymore. And I think you can, on the distribution sides or electricity grids, this is like you have a one-way street, and everybody knows it’s a one-way street, and all of a sudden you have two-way traffic. So, because in the electricity case, the electrons, they don’t go only from A to B, but now they also go from B to A and from B to D, and so that has changed the landscape dramatically, and it’s a big, big challenge.
08:37 Florian: Now, the next mega-trend and also challenge for this industry will be immobility, electric cars. Driving an electric car is great, but you need to charge it. Where are the electrons coming from? Well, the electrons come from the energy system, and the amount of additional load that is put on the local infrastructure via electric cars is very, very significant. Now the question is, how can we enable immobility, that people can charge the cars when they want to charge them and when they need to charge them without producing massive problems, blackouts, hardware blowing up? Which, by the way, happens already. Also, in the Bay Area, there are some nice videos on YouTube where you see the stuff that’s on the pole blowing up because you had five guys showing up at six o’clock, plugging in their Teslas at the same time. Now that’s a very visible example of the problem. And you have immobility push in China, you have it in Europe, also in the US, customers will demand these cars, and the system has to somehow deal with that.
10:33 Beth: How is data informing energy infrastructure? Can you give us a few examples?
10:38 Florian: Sure. That’s why we came up with the idea of data-driven business models, because if you look to energy systems, they are very data-rich. There’s a lot of data coming from the hardware in the field. Just picture a wind farm. Each of these wind turbines has hundreds of sensors that produce massive amounts of data. Then you have grids. You have a lot of sensors in the grids these days, they produce a lot of data. Then you have consumers that have connected devices in their homes or a car that is connected. If you have a Tesla, it produces data all the time, what you’re doing, where you’re going, when you’re charging. There is a massive amount of data that is produced by the system. Today, there’s really no very easy accessibility of this data, end-to-end management of the data, because you have different value chains with sometimes different owners of different segments of the value chain. It’s a very fragmented data space. It’s not like Facebook, where Facebook has all the data, yeah, and they can do a lot of things with it because it’s in one hand. Here it’s completely fragmented.
11:57 Florian: The interoperability technology from Intertrust is very helpful to break up the silos in a way that people feel comfortable with the silos being broken up, because you need to share data with others, but you need to ensure that you still keep the ownership. You don’t want other people to walk away with your data. That needs technology to secure that, and you wanna know who gets access to your data. You need to have digital rights management in place. Now, the reason why we invested is because they are very knowledgeable and very capable of enabling the monetisation of digital assets. They have proven that in other verticals, like entertainment media. And that’s where typical in the past, you found monetisation models around digital assets. And what we found out is that we can use the same technology that works for an app store, like an Apple App Store or a Google app store, to monetise digital assets, also for data assets that we own and operate.
13:01 Florian: Things like data interoperability, trusted data management, secure data management are things that are very important for us, because we operate critical infrastructure or we own customer-related data, assets which needs to be protected. But in order to unlock the value that is in the data, you need to have these technologies in place. That has developed very, very nicely. I think it’s the biggest success story that our company has. Why? Because we took the technology, we incubated the technology around the energy vertical, which is where we come in, where we have certain expertise. And then out of that, we started, we developed a new product, which we tested with customers, of course, and turned that, in some cases, into a new business.
13:54 Beth: What will IoT connections do for the energy space? What are you most excited about here?
14:00 Florian: Yeah, I’m most excited about that we get more data and more connectivity. Let’s say, in the old world, you had a TV, you had a fridge, you had an oven, you had appliances that were not connected. In the future, they will be connected, because the device manufacturers simply connect everything. I was at CES, there is a connected toilet, there’s a connected bathtub. Everything is connected. And it will take time before these things will penetrate the homes, but it will come, because it will just be a built-in feature. This means you get more and more information about usage of energy. And that’s what we are interested in. We are… We don’t try to socially profile people, for example, what they watch on a television that’s irrelevant for other people, not for us. We would like to know when do people use electricity, because with that information, we can help stabilise the energy system, which is under this enormous amount of pressure, because of this shift from a central to a de-central system.
15:09 Florian: And I think the the biggest lever is with cars, with electric cars, because they demand the most loads. In this transformation, this explosion of IoTs, sensors, and also IoT data, I think will at the end help us to have a hopefully, a more reliable, a cheaper, a more efficient, a better energy system in the future.
15:33 Beth: Can you put into context how the demand for power is increasing? How is our current electrical system up to par or not up to par for demand?
15:42 Steve: I think IT is driving a lot more of the power needs going into [15:48] ____, it’ll drive a lot more going forward. Currently, it’s about 2% of the worldwide demand for power and we’ve seen forecasts that it’s going up to eight or nine percent. And a lot of that is being driven from Edge applications. We have applications from data collection and content delivery of the Edge with IoT emerging, and millions of IoT devices proliferating the need to move power from centralized out to the Edge is a big challenge.
16:19 Beth: Yeah, exactly. From a power distribution point of view, what will micro-data centers do? Will they increase power consumption?
16:28 Steve: In the past, you’ve had a lot of factories and a lot of big industry that’s located closer to sub-stations. They’re strategically located where the power is. As things move out, an example would be like a farm that’s implementing IoT management system for agriculture. They’re gonna be deploying thousands of systems, they’re gonna have a data center there. They may need 10 megawatts of power at that site where they may not have needed that in the past. There’s a need to beef up the power distributions towards the edge of the power grids. We’re seeing a lot of that today. And with applications like 5G coming, where we’re gonna have to deploy hundreds of thousands and even millions of these micro data centers globally, it’s gonna be a tremendous challenge to bring enough power to those Edge sites.
17:24 Beth: Just dialing more into making 5G micro data centers more efficient, can you give me a comparison with the level of power consumption we have now, versus how that will look with 5G in the future?
17:36 Steve: Sure. With 5G, like I said, we’re gonna have to deploy micro data centers at scale, and we’re involved with projects for 10,000 pieces of micro data center. If you take, say, an application or a deployment that’s 2,000 pieces and say each one of those is 10 kilowatts, that’s 20 megawatts of power. We’ve had a lot of experience in larger data centers, making those as efficient as possible and refining them over the years. With micro data centers, it’s a little bit different. It’s more of a design challenge than a deployment, and then tweak it, and then optimise it. You wanna do in integration of the power, the cooling equipment, you wanna use the enclosures for the environment. You wanna use the latest battery technology, and you wanna use the most innovative cooling technologies in these designs. The result would be a design that’s optimised to be able to deliver the application as efficiently as possible.
18:52 Steve: And we also have advanced software tools now that we’re using to gauge the performance of those data centers. Now, when you have that many data centers deployed, thousands of data centers deployed, it’s very impractical to have the data center management systems on site, so we have cloud-based systems now. We’re taking the data from these micro data centers and we’re creating a simple and easy access point to the cloud, where you could look at all of your micro data centers together, and see if they’re operating at the optimal efficiency and the efficiency that they were designed to do, and if there’s any problems that arise. It’s a whole new type of approach to data center management, going from one data center with hundreds of thousands of access points, to the access points being distributed to the many different locations.
19:52 Beth: What does the global rollout look like for Schneider? Will you primarily be putting the micro data centers in Europe, or the United States, or what’s the roll-out plan for data center architecture globally?
20:04 Steve: Well, 5G is actually interesting, ’cause it’s a new architecture that needs, relies heavily on a data center architecture. And the old telco was kind of a country club, it was very closed. There was the carriers and then the telco equipment providers. With 5G, it’s more of an open system. And with an open system, there’s a lot of deployments of these MEC data centers, the IT part of it is actually standard servers, or white-box type servers, or whatever you wanna call ’em. It opens up this 5G architecture for more players, like the traditional IT vendors are now looking for a way in to participate. Schneider’s dealing with… We’re dealing with the traditional Telco, we’re dealing with the traditional equipment providers, we’re also dealing with the IT providers, we’re also dealing with the cable TV companies that are trying to also get involved with the 5G rollout.
21:03 Steve: And then you have the internet giants. Then the giants are all looking at, “Hey, the telco providers are not really that experienced with deploying these data centers at scale, so maybe it’s something we could do to break into this market.” Because the market… The projected rollout’s is estimated to be over $300 billion of equipment. Everyone wants a piece of it. And Schneider’s kind of perfectly positioned, ’cause we’re here to support across the board, anyone that wants to have a stake in the game. What I didn’t really mention was that the initial rollouts could actually be LTE advanced, or they could be four-and-a-half, some people are calling it four-and-a-half G. And the reason that will work with 5G is because the software is disaggregated from the hardware. You could deploy a four-and-a-half G application now, and that’s what… Like AT&T is calling 5GE, and some people are calling it pre-5G. The speeds are gonna get a lot faster as we roll out these new micro data centers. And then when the 5G software is more readily available, they’ll flip it over to the new 5G software.