The OPA (Open Process Automation) initiative

 A little background history

If you are already familiar with the OPA initiative, please exercise your patience for a couple of minutes. For those who do not, here goes a summary of the story so far:

In 2016, Exxon Mobil realized that the company faced the upcoming, and in many cases unavoidable, need to either revamp (with the mandatory control system’s update) or rebuild a significant percentage of its global production facilities in the next ten years. 

This was due to the usual period of planned life expectance of a typical process plant, which is roughly between 25 to 30 years, with a usually mandatory revamp/upgrade by the middle of its lifecycle to remain competitive. 

When Exxon made the calculations for the expected costs in control system’s upgrades and actualizations over a ten year period, the final number obtained was so high the company’s engineering team started to consider seriously the fact that, it might be possible to invest that amount of money in the development of an open standards compliant control system specification, designed purposely for the Process Industry, publish the specification, collaborate with the adequate suppliers and still have some change in their pockets after all the dust had settled .

A blood-tinged market

It is an unspoken fact that DCS suppliers have held both the Upstream and Downstream industries in their hands since the development and wide adoption of DCS systems due to several factors. This is a market where the customer ends up being dependent of the supplier through lifecycle contracts with expensive annual services fees, the unavoidable need to perform massive system updates every time a new release of the control system is released, the need to apply actualizations in a regular basis, and a black box approach that DCS suppliers prefer when interoperability requirements become a necessity.

The clean white sheet of paper method

ExxonMobil thought that to turn the new open standards-based concept into reality, they needed to use an “outside of the box” kind of approach to avoid conceptual constrains. 

A clean white sheet of paper kind of start would avoid any limitation from legacy systems. The engineering teams involved should not have any kind of conceptual constrains in the development.

The decision taken was to contract an external consulting team with a proven background on the integration of complex systems using diverse component suppliers.

The consultant services of Lockheed Martin, the company with a long story of aerospace related developments. And that makes sense, since most aerospace projects are basically the result of the integration of dozens of subsystems supplied by diverse manufacturers. After all, this company had been involved in groundbreaking projects like the SR-71 reconnaissance ultrasonic jet plane, the F-117 Stealth Fighter and in most of NASA projects. They are also involved in the Orion Crew Exploration Vehicle and in the F-35 Lightning II projects, but well, nobody is perfect.

The “upcoming” IIoT and Industry 4.0 revolutions are already here

An additional factor, that Exxon Mobil wanted to address, was to design an approach to handle the upcoming massive increment of data volume that will be received from its plants. 

Exxon Mobil handles 38 plants distributed all over the world. Those plants generated (at the time of the start of the OPA project) about 13 billion data records from 5 million device tags. With the advent of IIOT and Industry 4.0 based technologies, those numbers are likely to increase dramatically over time in the medium term. The amount of data being generated is becoming too large to be managed in traditional ways, so the use of novel approaches such as cyber-physical systems, machine learning techniques and artificial intelligence needed to be considered and taken in consideration.

The open standard approach

The ExxonMobil/Lockheed Martin partnership was the basis for the creation of the Open Group’s Open Automation Forum, an independent entity in charge of managing the standards that will define the Open Systems Architecture, which is the proposed solution to the initial request made from ExxonMobil.

The open standards method has proved to be the best way to create, develop and supply new technologies to the industry. The success obtained by similar organizations such as PI International, the Hart Foundation, and the OPC Foundation is the actual proof that a well-managed, open standard based, customer needs driven technology approach is the right way to introduce radical advances in the industry.

The DCS world

To understand the idea behind the Open Process Automation Forum, we should comprehend how do the current DCS based solutions work and what kind of business model is the one adopted by the DCS suppliers.

When a DCS supplier wins a project, this prize usually implies hefty discount levels in the supplied hardware (you may be surprised by the % involved, but they are high). In most of the cases the hardware is sold at a price that barely covers material costs plus transportation, it is a blood-tinged market. 

Then why would anybody venture into it? Because the real prize comes afterwards, in the form of expensive maintenance fees, onerous per hour service fees, costly system updates and revamping. All of them guaranteed for the life expectancy of the plant, which as we mentioned before, has a life span from 25 to 30 years. And you may remember that middle life cycle, mandatory revamping and update of the plant required for the sake of keep on being competitive. This business model is maintained by the DCS suppliers using proprietary internal communication protocols and a black box approach in both the hardware and software packages supplied.

The OPAS initiative

This is the expensive business model that ExxonMobil decided to avoid by the Open Process Automation initiative. By specifying the use of open standards and a “clean white sheet” mentality, ExxonMobil wanted Lockheed to create a specification that enabled suppliers to provide solutions that did not require backwards compatibility with previous systems and enable the incorporation of modern methods for data handling and processing, an enhanced level of interoperability between suppliers and the incorporation from the initial design of cyber security and enterprise-wide integration. Those functionalities were going to be natively built in the systems and not added as retrofit features afterwards. 

As time went by, several major automation suppliers joined the OPAF, as well as several major automation end users. And developments started to appear. This is obviously a consequence of the huge profits that may be collected by first time OPAF solutions suppliers.

 A new approach to an old problem

The change in the control system’s architecture is radical. The OPAF is not designing a system, but instead is creating a set of specifications and standards that component suppliers must comply. Instead of a centralized control system architecture based on servers using proprietary software to manage proprietary controllers distributed in a plant by proprietary communication protocols, the OPAF concept specifies a standard connectivity framework that would enable all the system components to interact between them as needed. It features native OPAS components as the DCF or Distributed Control Framework, DCP or Distributed Control Platform and the DCN or Distributed Control Node.

The DCN is the combination of a DCP and a DCF, the DCP is the hardware base controller, which is somewhat reminiscent of the PAC concept, but featuring additional edge computing device functionalities. DCPs use a DCF as a framework employed to run applications. A modular approach would use DCNs to control diverse plant production units. The OPAF even provides the specifications required to create a DCF that would enable non OPAS compliant devices to be integrated in an OCS.

But what does all this new tech mean in straight words?

The answer is the following: the OPAS architecture would enable any automation supplier to play in the same level as any DCS supplier, no hidden magic cards accepted.

This may be the explanation of why Emerson Process Management acquired GE, not by a sudden interest in the PLC market, which is mostly a mature commodity one, but an interest in the eventual DCN market based on PACs that can be foreseen in the future.

But I need an example

As an example of how this concept may be put to work, we can analyze a pilot installation currently being run by Exxon Mobil. It employs an OPC UA based communication framework. The controller of choice for this test plant is one of the models available in Phoenix-Contact’s PLCnext line of hybrid controllers.

When you look at Phoenix Contact’s PLCnext technology, you can see the embodiment of the DCN concept, a controller using a real time OS based in Linux with a Real Time add on, capable of performing control functions as a traditional PLC, but also capable to escalate and work in combination with other DCNs via the use of OPC UA as the backbone communications protocol employed by all the system’s devices. Systems applications could run in whatever computing device is connected to the communications framework, but a PLCnext controller has the hardware necessary to perform PLC tasks and Edge computing device tasks. And it can do all those tasks using standard IT derived programming languages while maintaining the IEC 61131 compatibility for PLC related applications.

The PLCnext platform even provides an app market, where third party suppliers can offer modular solutions for specific markets and applications. The backbone of the system is provided by the communications framework, so application programmers do not have to deal with interoperability and connectivity issues. The currently available real time industrial grade CPUs are powerful enough to run any kind of conceivable application. In the near future, the amount of data will certainly increase in orders of magnitude, as concepts such as pervasive sensing and ubiquitous condition monitoring sensors enable systems to perform diagnostics of anything connected to them

Conclusions

Finally, the OPAF concept is based on a modular basis, so any plant will be capable to increase or reduce its production by adding or subtracting production modules. This approach is where the OPAF and the NAMUR Organization encounter themselves in the same space.

I do not know what the final outcome of these developments will be, but I have the feeling that the usually conservative and innovations wary world of Process Automation is about to change in ways we may be unable to imagine today.

Mirko Torrez Contreras is a Process Automation consultant and trainer with an intense curiosity about what the future keeps on hold for us. The opinions exposed in this article are strictly personal. No affiliation exists between the author and the companies mentioned. Al the information required and employed in this article are of public knowledge.