Given the large number of legacy process plant control systems needing an upgrade in coming years, there is much discussion as to how, when, and why a given company should follow one strategy or another for modernizing or replacing a major process control system or distributed control system (DCS). However, in all that discussion, there is often one overlooked element as planners fixate on finalizing vendor selection and implementation plans.
A major upgrade should not be just about process automation, but should also focus on workforce development. A major migration project offers what is arguably the best opportunity to train the next generation of automation engineers and plant operators in a way that can yield benefits for many years to come.
Think for a moment about how people learn and build job skills. Studies show 70 percent of what an individual has to know to carry out his or her tasks is learned on the job. A major reason is that so much knowledge is specific to a given plant and even to a particular process unit. A student may come out of class understanding how proportional and integral control methods complement each other, but tuning a specific flow loop on a hydrotreater unit is a whole other matter.
Even experienced engineers and operators have much to learn when they move from one plant to another. Although there may be some similarity in the way different plants convert natural gas to ammonia or crude oil into gasoline, no two plants are identical.
Major control system upgrades or migrations are rare occurrences. In fact, a migration project in many cases is a once-in-a-generation opportunity to improve plant operations and to build your future automation workforce.
Many plants are running today with plant control systems that are more than 20 years old. If operators spend their entire career at the same facility, they could be involved in only one or two major upgrade projects. If such a learning and training opportunity is lost, there may never be another.
Control system touches everything
Consider some of the mechanics of what a control system does and the number of critical interfaces it has. From individual field devices to the enterprise level, the main process control system touches almost everything. Consequently, virtually all aspects of manufacturing operations and associated automation systems are addressed with a typical control system migration, and so are most of the stages required to manage a major project. Starting with a front-end loading study and proceeding through all project phases up to and including check out, commissioning, and training—a DCS or plantwide programmable logic controller migration has it all.
The difference between simply getting a migration done and taking full advantage of it as a training tool is asking, "Who can most benefit long term from doing this task?" instead of "Who can complete this step the quickest?" During a migration project, young engineers work side by side with operations, safety, corporate information technology (IT), and plant leadership personnel. When this is considered an opportunity for mentoring, it is also a means to transfer personal and tribal knowledge that is difficult to capture and convey in any other way. As experienced people retire, that knowledge is irretrievable and must be learned again, perhaps the hard way.
There will also be significant exposure to the monetary aspects of project management, starting with financial justification through projected return on investment. From a career growth and succession planning standpoint, what better way is there to develop the next generation of automation leaders?
A control system migration is an ideal time to learn about plant operations, automation, security, IT functions, and project management. Some functions, particularly IT and security-related elements, will be far more deeply involved than in years past, and the need for that kind of knowledge has grown enormously over the past few years. Training in all these areas will go a long way toward transforming a young engineer into a true process automation professional.
- Field wiring and device-level networks
- Safety instrumented systems
- Control valves
- Wireless networks
- Alarm management
- Operator interface
- Controllers, DCS, PLC, RTUs, etc.
- Analog and digital communications networks
- Asset management systems
- Simulation systems
- Process historians
- IT and cybersecurity elements (DMZ, firewalls, etc.)
Understanding automation at every level
Creating an automation system has to begin with a very deep knowledge of the process, or that unit will never operate consistently and profitably. Each element of the process—valves, reactors, instruments, pumps, and so on—has a part and that part must be thoroughly understood for effective control. Inexperienced engineers and operators may look at each element like a dot in a puzzle, but working through a migration lets them draw the connecting lines and see the big picture. Table 1 offers a review listing of the areas that a migration affects, but it is helpful to consider each one in greater detail:
- Instrumentation: Are there enough sensors? Are they the best type for each application? Are they in the best locations? A device-by-device analysis of field instrumentation can help new engineers understand how the process was designed initially and how it has changed over the years. Some devices that are troublesome may need to be changed or relocated for more accurate measurements (figure 2).
- Field wiring and device-level networks: Are connections to control system I/O well thought out and documented? Can newer technologies provide more reliable and maintenance-friendly networks? If getting information to the control room is troublesome, often device-level networks are to blame. Spending some time analyzing what is connected to what can explain a great deal.
- Safety instrumented systems: Do safety systems provide appropriate protection without excessive false alarms? Has the hazard landscape changed over the years? Just because the plant has not blown up does not mean the safety system is up to snuff. Reviewing the hazard picture can reveal much about what is happening inside the pipes.
- Control valves: Because valves are often among the most maintenance-intensive elements of a plant’s hardware, how are they performing? Are they sized accurately and capable of controlling adequately? Companies improving valve performance and maintenance find a fast way to save money. Like instrumentation, it has a huge impact on stable operation.
- Wireless networks: Were wireless networks added following a strategy or bolted on without a comprehensive plan? Analyzing such a deployment provides insights into the mechanics of network design.
- Alarm management: Do control room operators receive relevant alarms that they can deal with, or do they have to sort through nuisances daily? A study of alarming strategy in connection with safety systems can show how those systems work together to keep a plant safe.
- Operator interface: Can operators see the information they need quickly and easily (figure 3)? Do you have high-performance graphics standards? Understanding how operators do their jobs can lead to more effective interfaces. There may be better ways of showing how a plant is running than depicting animated piping and instrumentation drawing loops.
- Controllers, programmable logic controllers (PLCs), remote terminal units (RTUs), etc.: What secondary control functions are sprinkled around the unit, and why are they there? A study of small-scale and embedded controllers helps indicate how the original designers saw the best approach and increases understanding of the entire automation system. Some of those functions may benefit from being made more autonomous, while others may be better tied to central control.
- I/O: Can each of your devices communicate all its information in a timely and efficient manner? The communication capabilities of field devices and I/O should reflect each other for the greatest effectiveness. Smart devices connected to dumb I/O represent a wasted resource.
- Analog and digital communication networks: Everything needs to communicate, and each level needs the best method. From 4–20 mA to Ethernet, understanding how communication capabilities need to match information is the most basic design building block of networks.
- Analyzers: Product quality depends on having product attributes matched with the right type of analyzer at the right point in the process. Improving these critical measurements can have a major impact on plant productivity and profitability.
- Asset management systems: Is maintenance driven by actual equipment condition, a calendar, or simply responding when things break? Studying maintenance history under the illumination of equipment diagnostic information can help identify and eliminate the bad actors causing the most outages.
- Simulation systems: While a simulator may be necessary during a test of the new control platform, it can also serve as a training tool when the new system is fully deployed.
- Process historians: Often historians are the primary connection between a plant and the enterprise level. Working with these systems can help identify what company management considers important.
- IT and cybersecurity elements (demilitarized zone [DMZ], firewalls, etc.): Plant-level systems used to be isolated, but now the IT department will invariably be involved in any new deployments. Training plant networking engineers on how to work with IT systems and security measures is easier and cheaper than moving IT people into the plant.
Involved from start to finish
A major upgrade project can take years from conception to daily operation on the new platform. It is not something that should be forced into a compressed timeline, and frequently the transition period may actually be extended to spread out the cost. Most follow a fairly predictable series of steps depending on the overall scope, but just as no two plants are identical, neither are any two upgrade projects.
Young engineers, and even experienced ones, can realize huge benefits by participating in all the necessary steps to lay out an overall timeline. Table 2 summarizes the steps, which are listed below with explanations.
- Brainstorming the pie-in-the-sky wish list
- Initial budgetary estimates
- Hands-on tracking down and interpreting legacy drawings, and revising them to reflect current as-built reality
- Interpreting legacy code to develop current state process narratives, then optimizing those narratives for a new and better way of operating
- Troubleshooting control loops
- Compiling a cyberasset inventory and characterizing the security aspects of each asset
- Determining patterns of normal network communication to help identify abnormal traffic involving security problems
- Brainstorming the pie-in-the-sky wish list: Imagine all the things a new system can do. The list may need to be pared back, but this is a critical analytical step.
- Initial budgetary estimates: The sooner actual numbers begin to emerge, the better. Understanding the true value as well as the cost of each item is an important step.
- Hands-on tracking down and interpreting legacy drawings and revising them to reflect current as-built reality: Few things can help a new engineer understand a process like going back and seeing how the plant or unit was designed originally, and how it has changed over the years. Analyzing the “hows” and “whys” of those changes can be very instructional.
- Interpreting legacy code to develop current state process narratives, then optimizing those narratives for a new and better way of operating: Along with studying legacy drawings, looking at legacy process narratives and how they have changed gets down into the nuts and bolts. Figuring out how the process can be improved is key to developing ongoing process design skills.
- Troubleshooting control loops: If a plant runs with many of the operating loops in manual, figuring out what it takes to get them running in automatic makes life in the control room much easier, while significantly improving process operations.
- Compiling a cyberasset inventory and characterizing the security aspects of each asset: The first step of effective cybersecurity programming is understanding what cyberassets are in the plant and what software they are running. Without this, protective measures can never be adequate.
- Determining patterns of normal network communication to help identify abnormal traffic involving security problems: One of the characteristics of industrial networks is that traffic is somewhat predictable. In normal operation, the kind of communication that should be going on does not change much. Understanding this communication helps identify abnormal communication, which signals a problem or an intrusion.
Why training is so critical
The demographic time bomb facing industries around the world today has been discussed for some time. The bi-modal distribution of working engineers has its peaks at baby boomers reaching retirement and millennials entering the workforce. The
individuals nearing retirement are not being replaced fast enough, and those joining our companies now are unseasoned. Getting young people to join the ranks of working engineers will require effort from many directions. MAVERICK Technologies has joined with ISA and others to develop training programs reaching as deeply as elementary school to encourage math and science.
Companies not paying attention to this changing landscape could soon find themselves unable to recruit and or retain new engineers. Millennials do not think or act like the baby boomer generation. Young engineers need to be trained using different methods than their predecessors, and they have new ideas of what they expect technology to look like at home and at work. Any migration or system upgrade project that does not take these facts into consideration could be missing a critical opportunity, which could severely handicap a company going forward.
How does your company participate and contribute?
Unless your company is one of those very rare organizations with plenty of in-house engineering capacity, you will need to turn to your control system supplier or an automation system integrator to help facilitate your DCS upgrade project. Quality system integrators know it is important to make training an integral part of the control system upgrade as part of workforce development.
Many automation suppliers offer schools or other training programs for users of their systems, and these can be hugely beneficial, particularly when they are integrated into a larger training effort. However, caution is in order. Some end users see such vendor training programs as something of an easy out rather than working on a more individualized approach.
Migrations and other large projects are opportunities to work with the major supplier and integrator to tie project plans and requirements to formal career development plans for key employees working on the projects. This typically extends beyond engineers to include technicians at various grades and project management personnel.
Many system integrators can deliver in-house training at your plant or their facility using various classroom formats. Because your new system will likely contain major elements from a variety of suppliers integrated together, training can also be integrated and may include participation from your suppliers.
On the end-user side, internal automation personnel can work with and lead their systems integration partners and automation vendors though DCS migration and other projects. In the process, each of the three groups will learn from each other, and all will improve. As plant automation systems depend more on traditional IT networking techniques, a migration project is an excellent opportunity to build bridges between control systems and IT professionals, and to cross train individuals in both areas.
For most process industry firms and service providers, the best time to train is now, while key senior people are still on staff and have yet to retire. This is an ideal time, and may be the only time, to extract knowledge from those seasoned engineers, operators, and technicians and then transfer this knowledge to a younger generation. This knowledge transfer window is closing at many plants, so the opportunity will never be better than today.
A control system migration is often the most exciting and transformational time a process plant will ever see, and as such it is a rare moment to develop a vision for the future, and to then convert that vision into reality through intelligent use of automation. The new generation of engineers loves a challenge, and a migration program is just that as it combines highly technical tasks with management of a major project.
To succeed, engineers must learn and apply both technical and people skills. Young engineers having strong organizational and people skills will emerge quickly as leaders in this kind of situation. Some will embrace the opportunity fully and seek to learn everything they can. These people are critical to the future of a company, and they must be guided and nurtured.
Like the medical profession, where young doctors work with experienced practitioners so they can see a patient through all the phases of treatment, a DCS migration is one of the few projects that touches upon virtually every aspect in an automation professional's toolbox. In addition to learning by doing, young engineers will gain further experience at the conclusion of the project when they train plant personnel, particularly technicians and operators.
In summary, seize the moment to not only transform plant operations through a well planned and executed migration, but to develop and transform your future workforce at the same time. Done properly, this type of investment will pay dividends for decades.
Paul J. Galeski, the chief executive officer and founder of MAVERICK Technologies, specializes in high-level operational consulting, as well as the development of automation strategy and implementation for automation technology. He is also involved in expert witness testimony, and is a contributing author to Aspatore Books’ Inside the Minds, a series of publications that examine C-level business intelligence.
A version of this article also was published at InTech magazine.