InTech APR23 How DCS Migration Improves Operator Experience

Sharing your cookies helps us improve site functionality and optimize your experience. Click here to read our cookie policy.

A migration provides many benefits that will help improve the lives of those involved in several areas of the plant floor. The maintenance technicians won’t be chasing rare spare parts on the gray market. The plant manager will have remote access capabilities to monitor the plant operations from the office (or beach condo) on the new system.

With a project of this magnitude, the biggest impact is on the folks that use the DCS to run the plant—the operators. They need to understand the benefits specific to their area. Operators are often resistant to change and might be detractors from the outset, so convincing them might be challenging. If they still don’t see the benefits after the migration is complete, the company has, unfortunately, failed in its business objectives. Success depends largely on upfront planning and getting buy-in from the operators and other key stakeholders.

Consider how a DCS migration can improve the operator’s experience running the plant in three primary ways. To illustrate, think of the improvement process in terms of sight, sound, and feel.

Feel is associated with process control— the way the DCS controls the system. How much of the control is automatic versus Sight is associated with the graphics—what operators see on the screen and how they get to the information they need.

Sound is associated with the alarms—improving operator alarms and their situational awareness, as well as what operators react to during abnormal conditions and the action they are required to take.

Feel is associated with process control— the way the DCS controls the system. How much of the control is automatic versus?

Consider improving the sight—the graphics. By now, most people have heard of the ANSI/ ISA-101.01-2015, Human Machine Interfaces for Process Automation Systems standard and high-performance human-machine interface (HP-HMI) best practices, which equate good modern graphics to greyscale. This is a big part of improving what operators can easily see and respond to. Changing the color palette to a grey background with sparse use of color draws attention to abnormal situations and is part of the solution (Figure 1).

HP-HMI is more than just color changes, however. It’s about planning. It’s an evaluation of what needs to go on the Level 2 (L2) overview screens that an operator uses to run the plant.

When migrating legacy DCS graphics, replicating the screens that operators use is usually inefficient and doesn’t improve their situation. Even though the operator has been used to the system’s look and feel, it doesn’t mean the user experience can’t be improved. This is where the new HP-HMI best practices come into play: L1, L2, L3, and L4 screens should expose the right data and controls to reduce the operator’s cognitive load and deliver the information needed, when it is needed.

Here are descriptions of what type of information is shown at each level:

L1: Always up, wall-mounted trends or displays that are always visible. L1 gives everyone a quick status of the unit or critical process data and how it is operating.

L2: The money makers. These are the screens an operator uses to monitor the process and take routine action to make product.

L3: Details. L3 provides more information on a particular process unit, which is required to diagnose, troubleshoot, or perform non-routine operations.

L4: Pop ups. These are controls and faceplates that operators interact with until they go away.

Before moving from a legacy DCS system, best practice is to do an HP-HMI storyboard workshop. Get the engineers and operators together with an experienced facilitator to lay out the story of how the plant should be operated—not how it’s operated today, but how it should be run. The storyboard should include what information the operator needs to see at a glance and what can be hidden until needed.

The benefits of a storyboard workshop are twofold: The first is associated with the cost of the migration and the return on investment (ROI). A storyboard session can reduce the number of graphics to migrate by as much as 40%. The cost of the session is quickly offset by the smaller number of graphics to migrate.

The second benefit is in the operation of the plant after the migration. With better access to information, the operators do a better job running the facility. The improved situational awareness allows them to recognize ways to improve the operation and head off potentially costly problems before they start.

Even with effective HP-HMI graphics, operators can’t see all things all the time. That’s what alarms are for, so improving the sound of the control system alarms is also important. ANSI/ISA-18.2-2016, Management of Alarm Systems for the Process Industries, provides guidance from a standards perspective. A modern DCS has the capability to configure at least six or seven alarms on every input—highs, lows, bad signal, etc. The options are endless, but the operator’s bandwidth to filter and respond to these alarms is not.

The control system should only alarm the operator when an action is required. This takes discipline and work. The remedy for this situation is well known. Alarm rationalization is needed to evaluate and document all the alarms and decide what is necessary and what can be eliminated, consolidated, or made smart.

One method for smart alarming is alarm shelving or disabling based on other plant conditions. For example, a low-flow alarm may be redundant if the pump providing that flow is not running. Alarm shelving is the capability to hide or remove multiple alarms based on a process condition. For example, if a compressor is taken out of service, the operator does not need to respond to any of its alarms. Shelving those alarms until the compressor is placed back in service can ease the load on the operator.

The payback for an alarm system that only rings when needed is twofold: First, it gives operators the ability to recognize abnormal situations quickly and act to mitigate them. After all, this is the purpose of an alarm system. Second, without nuisance alarms occurring constantly, operators have time to do their most important job—running the plant. It gives them the time and focus to improve yield and reduce errors.

One of the most common improvements in a control system migration is loop tuning and mitigation. Since every loop must be reprogrammed, it is an excellent time to improve their tuning and control conditions. Undiagnosed problems or problems that have been ignored for years can be evaluated and corrected.

Another improvement to the feel of a control system can be the difference between operating an almost self-driving car and a backhoe. In a modern car with cruise control, lane detection, and safety features, we can guide the vehicle where we want it to go, but we don’t have to control every action all the time. A backhoe, for example, requires the operator to directly manipulate every move with a separate joystick or knob. This method works, but it takes a lot of skill and manual control. Many legacy plants run that way.

Even with effective HP-HMI graphics, operators can’t see all things all the time; that’s what alarms are for.

One of the most promising methods that is increasingly becoming accepted is statebased control (SBC). In SBC, each unit or piece of equipment has a set of operating states (Figure 2).

For example, a distillation column may have idle, filling, heating, running, and total reflux as its set of states. Rather than manipulate valves and proportional-integral derivative (PID) controls directly, operators can step the column through the states, or the control system can automatically move the column to a new state based on previously defined conditions. This allows the operator to spend more time setting the course, rather than driving the bus. Less time spent on easily automated tasks gives the operator more time to run the plant.

States also can be used to enable or disable smart alarms when conditions make that alarm meaningless. If the column hasn’t filled with liquid yet, the low-level alarm is meaningless.

Another method of increasing the control of the plant involves a step change in the control strategy. The expert implementation of advanced regulatory control techniques such as feed forward, ratio control, and inferred properties can have an impact on a plant’s ability to operate closer to constraints and thereby increase throughput. Again, a DCS migration brings the right expertise together and touches every control strategy in the controller. There is no better time to at least consider the execution of these strategies.

A control system migration is a big endeavor and requires buy-in from key stakeholders across the enterprise. The re-creation of graphics and control strategies on a new system is a challenging task, but it is also a great opportunity to improve the operator experience and overall production processes. Careful planning and some upfront work can improve the operator’s control of a plant in three key ways: sight—improving the graphic layout for the operator, sound—improving the alarm system, and feel—improving the operator’s control of the process.

All these improvements eliminate distractions and allow the operator to focus on thereal job—to keep production processes and the plant up and running. The ROI that these improvements provide can pay out big and help companies maintain their competitive edge.

We want to hear from you! Please send us your comments and questions about this topic to [email protected].

Scott Hayes ([email protected]) is the DCSNext Portfolio Manager at Rockwell Automation. Hayes is a licensed control system engineer with more than 20 years of experience leading automation projects and programs, as well as hands-on configuring and networking of DCS, PLC, HMI, process historian, and visualization solutions.

Fekri Abdullah ([email protected]) is a senior engineer at MAVERICK Technologies, a Rockwell Automation Company. Abdullah has a bachelor’s degree in electrical engineering from the University of Michigan Dearborn. He has been working with MAVERICK in the automation field for more than 14 years, supporting numerous projects in different processes and platforms.