HMI-Design Essentials For Today’s SCADA Systems

Data efficiency is a must in empowering operators to make quick, action-related decisions.

The entire purpose behind SCADA is easy data visualization for rapid decision making. For some reason, system integrators and owner/operators continue to make HMIs more complex than is necessary for efficient use. The result is overcrowded HMIs that are jammed to the brim with data points, making it difficult for operators to do their jobs.

According to Allan Evora, founder and president of system integrator Affinity Energy (affinityenergy.com, Charlotte, NC), for most applications, the best way to design an HMI is to integrate the bare minimum of data needed to make intelligent (action-related) decisions. “The format should be easy to interpret, and the design should be intuitive,” he explained. “As operators interact with their HMIs, they shouldn’t feel stressed, but feel empowered to know exactly what’s going on in the facility.”

Evora offered the following HMI-design tips for redesigning controls systems for maximum visibility.

Only the most crucial, high-level information should be on the main HMI screen.
HMI images courtesy of AutomationDirect.

Design overview screens around KPIs and layers.

If operators can’t tell what’s going on within seconds of looking at an HMI SCADA screen, it’s too cluttered. Displaying data points just because they exist is pure distraction to an operator. After all, data for the sake of data is not the point of SCADA.

Industry best practice is to simplify graphical user interfaces and bring attention to process deviations, equipment malfunctions, or abnormal conditions that require operator intervention. Only the most crucial, high-level information should be on the main HMI screen. How, though, do you decide what data deserves that coveted spot?

Focus on a few key metrics, rather than a mass of data points. One of the easiest ways to keep a clear interface is choosing only to display KPIs (key performance indicators). Those KPIs should be directly related to business goals, i.e., process output, reduced downtime, and safety, among others.

Selected KPIs should be very clear metrics (specific to an owner/operator’s industry) that inspire operator action. Based on unique features and devices at various facilities, there’s no one-size-fits-all HMI. They should look different (be customized) according to a site’s particular needs. For example, a landfill waste-to-energy plant might choose to display KPIs such as:

• active alarms
• kW
• kWh
• methane %
• O₂ %
• CO₂ %
• gas pressure
• gas flow.

The HMI of a utility-scale solar plant, on the other hand, would feature much different KPIs, including:

• interconnect voltage
• interconnect current
• interconnect kW
• interconnect kVAR
• inverter status
• irradiance
• alarm summary
• breaker status of switchgear.

To avoid tracking too many KPIs and creating an over-crowded display, use fewer than 10 KPIs on a screen. If additional information is needed to support a more-detailed analysis, it’s best to use layers. With this approach, information is accessed by drilling down to lower-level screens from the main overview. In most cases, to reduce the time required to access important, detailed information, an HMI should have no more than three layers.

Show analog-value ranges along with upper and/or lower alarm-limit indicators.

Make analog data more meaningful.

It’s not an optimized process if operators are expected to memorize each device’s preferred levels and numbers, and cross-reference to the data displayed on the HMI. This situation also complicates training of new operators. There’s an easier way.

With regard to analog values, e.g., temperature, pressure, kW, a best practice is to show the value range, along with indicators of upper and/or lower alarm limits.

Another way to make an operator’s life easier is to indicate the direction in which a variable has been moving over a given timeframe, including the minimum and maximum values. This, of course, requires balancing the specific details with the parameters and understanding that not all information is treated equally.

An additional issue with systems in which comparison of analog information may be required involves the ability of operators to truly understand what the values should be. Designing HMI graphs with percentages, rather than raw values, makes it easier for personnel to identify problems with just a quick scan of the data. This is data normalization, which allows users to look at dozens of data points and quickly identify outliers.

By avoiding alarming on events, an HMI’s screen should clear up considerably.

Declutter the alarming functionality.

Operators may love SCADA, but most of them are challenged when it comes to managing alarms. The amount and/or frequency of alarms can create stress and ultimately result in critical conditions being missed due to excessive alarm noise.

An effective way to minimize the alarms that an HMI displays is to differentiate between them and events:

• Alarms indicate an abnormal condition, e.g., leak detected, fuel-level low. They signal a call to action for the operator. Alarms should never confirm a normally running process.

• Events are situations that need to be logged for future reference, but most likely may not require operator action, e.g., chiller started, motor running. Events should still be logged for future analysis, but never alarmed on.

By avoiding alarming on events, an HMI should clear up considerably. This approach also helps avoid operator apathy (a common situation among those who have experienced too much noise in the system for alarms to be meaningful.)

In larger systems, another effective way to reduce alarm clutter is to use conditional alarming. Most SCADA systems allow logic to be applied to the alarm configuration. Alarm logic can be used to suppress children alarms when a parent alarm occurs. For example, if putting a piece of equipment into bypass/maintenance mode triggers certain related conditions, a decision can be made to only configure the bypass condition as an alarm, and configure the other resulting conditions as events.

The International Society of Automation (ISA), Research Triangle Park, NC, has published Standard 18.2 that discusses best practices as it relates to understanding and applying alarm-management systems.

Crafting HMIs to be intelligently obvious

Displaying data from a SCADA system on an HMI is light-years away from past methods of monitoring critical plant systems. Instead of frantically running around recording values on a clipboard, personnel can calmly view an entire facility’s performance on a single screen.

In Evora’s view, the effective operation and maintenance of a site and/or process will only be as good as the tools provided to its operators. “A SCADA system can be a highly valuable, productivity-enhancing tool,” he said, “but it must be designed properly.” Clearly, the less time operators spend analyzing data, the more time they have to fix the actual problems that an HMI identifies.

Evora noted that it’s up to the owner/operator to stress the importance of data efficiency to the system integrators. “Owners should place great emphasis on turning a screen full of data into actionable information that’s easily readable at a glance,” he concluded. “Otherwise, data efficiency may not be a design priority.”

A Certified Measurement & Verification Professional (CMVP), Allan Evora has more than 20 years of industry experience working in every capacity of the power-automation-project lifecycle. Charlotte, NC-based Affinity Energy (affinityenergy.com) is a partner of AutomationDirect (automationdirect.com, Cumming, GA).