Cristal transforms time-series data into measurable improvement

Cristal is one of the world’s largest producers of titanium dioxide (TiO2) and a leading producer of titanium chemicals.

The company builds innovative products that harness the unique functionalities TiO2 to create a cleaner and brighter future.

Nukon recently worked with Cristal and introduced Seeq, a process data analytics specialist and OSIsoft solution partner, to:

• Enable fast time-to-insight on time-series data
• Reduce the time spent in spreadsheets, and
• Improve processes to deliver better operational efficiencies.

Cristal wanted to have the ability to continuously improve their operations and achieve a better return on investment for their production facilities.

However, in order to meet this objective, the company realised that they needed to overcome two main challenges:

1. Improve the tools available to engineers and production team

Cristal needed better tools to analyse time-series data – that is, tools that delivered beyond what traditional tools could provide.

The company recognised the clear limitations of excel in analysing data and the need for a modern set of tools that enabled better process data analysis for incident investigations.

They also wanted to reduce the time spent on spreadsheets. This would greatly improve the speed of root-cause analysis – decreasing the time needed to understand production issues including start-ups and asset performance.

2. Gain valuable insight across processes and products

Cristal lacked production visibility across processes and products. For instance, they didn’t have enough information on the impact of different ingredients or different volumes of ingredients on the process.

Cristal needed to have a better understanding on why and how things were processed. They wanted to have the ability to reconstruct the conditions that led to a particular incident – and then be able to review historical data, validate assumptions and record the findings.

Cristal knew that all the data was there in the process-driven plants. However, they didn’t have the ability to use the data to gain valuable insights and confirm their hypotheses on root cause issues.

Cristal was already aware of Seeq even before the Nukon team proposed it, as the preferred time-series and data-interrogation toolset for process analytics.

Nukon and Cristal were therefore agreed that Seeq would be able to address Cristal’s main challenges and have a direct benefit on the company’s operations.

After a few meetings, we received the approval to start working on two Cristal sites. Nukon consultants spent the first few days training process engineers, then followed by doing deep-dives on actual use cases.

This not only helped the engineers set things up correctly on Seeq, but also enabled them to look at the data in a different way, ensuring they get valuable and actionable business insights.

The initial work involved:

• Analysis of new project performance
• Incident investigation, and
• Providing a dashboard for monitoring start-up production and performance against KPIs

With the use of Seeq, the Nukon team helped Cristal engineers:

• Properly analyse and contextualise the data from use cases – and gain real and actionable insights.
• Do more of the insight analysis on their own and have a better grasp of the new toolset.
• Analyse data across multiple points from a single location and do a comparison between plants.
• Compare operational conditions by a number of variables – product time of day shift, or crew running the plant, or start-up and reaction performance under different conditions.
• Reduce time on data analysis, improve processes and deliver cost savings on capital projects.

“The way Nukon implemented Seeq was both agile and successful. Our process engineers are now able to analyse time-series data and validate hypotheses, plus easily document and share insights with the rest of the business.” - Dave Worley, Process Control Superintendent, Cristal.

Nukon and Cristal are now discussing further analytics and workflow services using the Seeq tool.