Webinar On-Demand: Turning Up the Heat – Data’s Role in Electrifying Thermal Processes

Available On-Demand

Global decarbonization trends are pushing industries to electrification of their process heat. While electric process heaters have been utilized in various sectors for years, many industries still navigate unfamiliar territory with electric heat adoption. This transition sparks understandable concerns about its implications on processes. However, through a comprehensive approach that emphasizes data-driven strategies, we’ll explore how to navigate this transition.

Key Takeaways

Discover the benefits of integrating a complete thermal system solution with data analytics for efficient transition to electric heating.
Identify the key components that should be monitored within the system.
Explore case studies where the implementation of advanced thermal systems equipped with data analytics have resulted in cost savings and improved uptime for customers.

Chelsea Hogard

Industry 4.0 Team Leader

Chelsea Hogard is the Engineering Team Leader for Watlow’s Industry 4.0 Development Team. With a background in teaching high school math, physics, and engineering, she shifted her career focus and joined Watlow as an engineer five years ago. During her time at the company, Chelsea has contributed to research in advanced control methods and has gained practical experience in field service and training. Recently appointed to lead the Industry 4.0 team, she is committed to helping customers by using data analytics for effective problem-solving.

Attendee Questions

Is Data Insights software based on the cloud or On-Premise?

Data Insights is cloud-based. We collect the data and send it to the cloud. It is sent over cellular which helps avoid vulnerabilities on the customer’s network and keeps it secure. The reports are shared on a monthly basis.

Is the maintenance leader able to view this in real time through an app or computer?

The way that we are handling this now is that we are delivering reports to the customer. Real time access to information is not currently available. That feature is on the roadmap for future capability.

Is Data Insights limited to Watconnect control panels or can it be applied to customers processes?

We have done retrofits on systems using a standalone box we call the «sidecar box», that works alongside your system. As long as the components can speak Modbus TCP, Modbus RTU, OPC UA, Ethernet IP we can connect to it and collect that data.

Can the Watlow systems work on GW of input electrical power with the plant operating at 500 deg C?

We are looking at very large megawatt systems. Power does not matter. The system works on all sizes.

What hardware is required to monitor wire terminal heating?

This requires an additional control module and sensors within the panel.

Can you use leakage current to detect element failure?

Current leakage if monitored over time can be a good predictor of impending failure. It does usually require additional hardware in the system. Depending on if there is room in the panel we can add hardware to monitor leakage current.

Can the failed element detection system predict a future element failure, or does it just detect when an element failure has happened?

At this time the product detects failures but is not providing predictive failure. That feature is on the roadmap for a future release and we do have a patent around this. We need to finish that development in the Data Insights package.

Do you deal in field heat treating? My question would be, can contactor ‘on’ time be tracked?

We have done analysis on heat treating applications. Contactor «on» time can be tracked in cases where contactor status data is available from the system.

Will Data Insights have real-time analytics using AI or Fuzzy Logic/Decision tree logic? IE Will Data Insights deliver real-time changes to process parameters for process optimisation?

We do plan to utilize advanced methods for data analysis, such as AI algorithms. In order to maintain system security, parameters are never written back to the control system for real-time process optimization. Rather, these insights and recommendations can be presented to customers in a timely manner so that the changes can be implemented on site.

Which predictive maintenance algorithms are you currently working on to predict equipment failures? Does the data sampling rate have an influence on the performance of these models?

We plan to develop algorithms to predict failures of various components, including the heater elements and power controllers. The right sampling rate is crucial for effective predictive maintenance, as it affects the ability to capture relevant data for early detection of potential failures. In time series data there is typically a high correlation with neighboring data points. So you want a sampling rate that minimizes redundant data and captures the meaningful trends and events. It is application specific depending on process, response of the system, etc.

How are the predictive maintenance algorithms verified/validated… are algorithms run against real systems which are allowed to fail in order to verify the predictions of the algorithms?

Ultimately the best way to validate the algorithms is to see a failure. To some extent, we can test in a lab, but the algorithms will need extensive field testing to understand how these translate to real-world applications.

Will this presentation be able to be shared with others within our organization? Can they register to review and if so for how long?

The recording will be available. Anyone who has missed the webinar is still able to register to view on demand.