Digitalisation is a hot topic these days, especially in the refining and petrochemicals industry around the world. The discussion in downstream today has moved quickly from why digitalisation to how quickly can organisations transform.
Digital technology has all the buzzwords — big data, cloud, Industrial IoT, machine learning, augmented reality, virtual reality, etc. — which are all horizontal technologies. All these technologies are being deployed at scale in both midstream and downstream industries. For example, using data that is collected across hundred thousand, or may be even million data points in an operational facility, and providing visualisation in an operational way. Another great example of disruption is how machine learning is being applied to predict unplanned downtime. This is really where the current state-of-the-art is in terms of the digital transformation in the worldwide downstream industry.
Digital transformation of the industrial world
AVEVA has committed itself to the role of digital transformation in the industrial world. The company has doubled in size in the last twelve months by combining the heritage of AVEVA business, which has been around for more than 50 years, focusing mainly on EPCs and oil operators, and the legacy of Schneider Electric software business, which focused on the operational aspects of different industries, including the oil and gas industry, into one entity. Currently, the combined enterprise is worth $1bn in revenues and operates in 80 countries around the world. Craig Hayman joined AVEVA in February 2018 as chief executive officer.
“We have almost 5,000 employees and almost 100 of these employees are PhDs. We spend $100mn in a year in research and development on committing ourselves to the digitalisation of the industrial world. We have about 16,000 customers around the world. We are learning from them every day, how to further optimise midstream and downstream sectors, using Industry 4.0 in very compelling ways. What we are learning is fantastic,” Hayman reveals.
“Forty-five percent of our business is from the oil and gas industry, and much of that is from the midstream and downstream sectors. Our business has two aspects — OPEX and CAPEX. OPEX is the operational cost of running the downstream facilities, which are there for, may be, 20 to 50 years. And, CAPEX is the new investments in the greenfield, or brownfield environment. We are focused on optimising both these aspects — delivering better insights to the customers like reduced downtime, increased safety, optimised RoI, reduced project handover time, etc.,” observes Hayman.
In the last three years, the downstream industry’s focus is driving innovation and not just building up new facilities. Innovation in downstream industry means operating facilities at higher operational efficiency, less downtime and better safety. Digital technologies are used in this transformation of downstream production facilities.
The digital tools
Downstream plant operators have a very difficult job — they have to drive increased performance in their facilities and also try to reduce cost, at the same time. To do that, they are looking for the best practices that can be applied within their current budgets.
Most downstream operators want to drive digitalisation. They already have various projects to drive the digital transformation. The only questionthey have is how can they move more quickly enough in their digital aspirations? How can they do it with low risk? How can they do it with a small incremental investment?
“AVEVA helps our customers scale quickly in their digital journey, by running a 90-days pilot, for example, for these plant operators. We start collecting data of the operational facility. It could be from a couple of thousand data points. It could be about a compressor, or a pump, or about vibration, viscosity, pressure, temperature, etc. We collect all of that data. Then, we start visualising it in a reliable, and appealing way. This visualisation gives transformational insight into their business,” Hayman explains.
“Then, we start applying machine learning. We look at the operational data of today versus every other day up till now, to detect anomalies. For example, we may detect a pump is vibrating little bit differently today than it did every other day. We compare that pump with every other pump from that manufacturer. By this process, we can predict when it will fail, before it will fail. The result is, when the plant operator is doing planned maintenance in the facility, the pump could be taken offline and the maintenance on it could be done so that there is no unplanned downtime. As a result, the savings are just incredible for the customers.”
“Machine learning is the fastest growing part of our business. It accounts for 10% of our revenues. Another area is around virtual reality and augmented reality. This is around operator training — being able to take a worker through an operational facility virtually and being able to simulate the maintenance procedure — for example, in which direction a valve should be rotated, in which sequence a valve should be rotated, how that should be kept in concept with the original design of the facility, etc. — this is an amazing boon to increasing operator safety,” adds Hayman.
“A third area, which has been around for a while but really starting to happen now, is the idea of point clouds. Most of the operators do not have their operational facilities digitally modelled. We go through the facility and collect a point cloud of that facility.”
“For example, if you are used to drive a Tesla down the street, it builds a 3D point cloud of the street. We do the same in an operational downstream facility and capture everything in 3D. This allows us to overlay the operational data coming out of that facility and visualise it in a way that is useful for the operator. For example, if pump 335 is about to fail, the operator can know this from an operational dashboard and can look at it in a 3D layout and see that pump in 3D simulation. This is a remarkable innovation,” Hayman asserts.
The convergence of the operational technology with the information technology in the downstream industry facilities is something which is moving very rapidly. These facilities are meting out large amounts of data every day. But this data is not reaching the hands of the very people who need that data in a usable context to operate that facility in an efficient and safe manner.
“What we provide is the way to capture that data and visualise it in a way to make it available in context for the real operator — who is working in manufacturing, or feedstock quality, or maintenance, or in shift handover. We can apply the best practices and algorithms to deliver some insight for the people who are operating the facility,” points out Hayman.
Delivering capital projects on time and at cost
The total CAPEX spend across the world in the downstream industry — lot of that in the Middle East — was averaging around $450bn a year five to ten years back. This is not the case today. The total CAPEX is around $250bn a year now.
There was a big shift in the industry from the belief that building out capacity in downstream facilities would be how it can scale in the market. When the CAPEX spend plummeted from $450bn to $250bn, the operators wanted to make money at any price. This drove innovation in the downstream industry.
Getting capital projects delivered on time and at cost with low risk and low contingencies is a key element in the organisational success in the downstream industry. The design tools pulling in supply chain data connected to procurement of standard materials for capital projects are used by everyone in cloud. This is an area where AVEVA is investing in very heavily.
“Let us make a step there from the design of the facility into the handover. How can we reduce handover time from nine months to about nine weeks? How we achieve this is by delivering not just the facility but the structured data behind the facility into the hands of the operator. Perhaps, it might also specify into the facility some sensors — ability to flow data out of that facility — as it is being made operational. The operator can look at that data and understand how it is performing back to the design,” Hayman clarifies.
“The next step is enforcing a simulation at the same time when you are doing front-end engineering and design. Every other industry does simulation as part of design. In the oil and gas industry, we bring process simulation into the design sequence. So, as we are designing various piping layouts, etc. in the plant, we can see exactly how that facility is going to operate at different levels of efficiency, different levels of production, and different levels of energy. I am convinced, that in the next five years, these approaches are going to be the de-facto method of how we work in capital projects. Of course, today that is not how we work.”
The Middle East connection
“In the downstream industry in the Middle East, we have many customers. We are working on different aspects of their facilities. For example, materials management in the construction of a downstream facility, ensuring the procurement of the standard materials into a facility is optimised,” comments Hayman.
“In the earlier world, where the flow of goods were relatively easy with very few trade barriers, procurement of materials for CAPEX projects were relatively easy. But now, as we get the trade barriers going up between different economies of the world, there is much more focus on ensuring access to consistent flow of standard materials for the construction of a downstream facility. We are involved in several projects around this aspect in the Middle East.”
“Another aspect we are actively involved in the Middle East is big data. In many projects, we are collecting large amounts of data from the facilities and presenting it to the operating companies in a way that they have visibility to their production footprint. Economics makes total sense here. If you can drive a dollar a barrel in efficiency out of these facilities — purely through technology and not through more downstream capacity — you can drive enormous profit savings into your business and the economies which you support. This is what we see extensively happening in the Middle East,” Hayman explains.
Machine learning will be the key tool
“Between now and 2025, there will be phenomenal change in the digital downstream landscape. AVEVA will be right in the middle of this change with our customers. We connect our customers together. When an organisation works with AVEVA, they are not just working with us; they are working with all our customers all over the world, with all the best practices we have in the digital transformation journey,” opines Hayman.
“What will happen between now and 2025 is that these best practices, or patents will be codified into software as standard digital packages and tools. We are so excited about the road ahead for the oil and gas industry, specifically around downstream, because we see so much innovation and digital transformation coming into the industry. It will be an astounding time for the industry.”
The digital downstream facilities in 2025 will look much the same as those do today. But how those operate will be in highly efficient manner, how those deliver will be optimal, and how those turn over RoI will be exponential. The level of risk will be minimal and the level of safety will be significantly higher than it is today.
The way that is going to happen is by data in the design — simulation of those facilities before those are constructed, or data in the sourcing materials on how they are constructed, or simulation on how those operate at start-up, how those operate at low efficiency, high efficiency, or how those absorb unplanned downtime, or how those use large amounts of data in machine learning to predict unplanned downtime. With modern digital tools, within few seconds, an operator will be able to understand deeply how a facility is operating.
“Our customers want to design and operate not just an average facility, they want to operate the very best, the most efficient, the safest, and the cleanest facility that they can, which is going to operate between 20 and 50 years. And, to make this possible, machine learning is going to dramatically change how those facilities operate by 2025,” Hayman concludes.
For the latest refining and petrochemical industry related videos, subscribe to our YouTube page.