Historian use improves process optimization

The benefits of Advanced Process Control and advanced analytics have become increasingly well established in the process industry. Through continually evolving technologies, the benefits of APC are now accessible to a much broader range of industries and users by providing mainstream tools and leveraging diverse industry experience.

By John Leppiaho, GE Fanuc Intelligent Platforms and Derick Moolman, CSense Systems August 1, 2009

The benefits of Advanced Process Control and advanced analytics have become increasingly well established in the process industry. Through continually evolving technologies, the benefits of APC are now accessible to a much broader range of industries and users by providing mainstream tools and leveraging diverse industry experience.

Recently, industry requirements of plant historian and HMI/SCADA products have grown, driven by pressure to reduce production costs, meet regulatory requirements and survive as competition increases. The challenge now is to meet the industry’s expectations to use these products to improve process optimization.

Transforming data into information

Many manufacturing and processing facilities have invested in historian technology, allowing them to capture large volumes of process information. These valuable repositories provide trends, reports and analyses from which companies make decisions about process performance. It is now necessary to have improved client tools that transform the captured data into actionable information.

When selecting Advanced Analytical tools for use with a historian and HMI/SCADA, the challenges to consider include:

  • Few intended users have an advanced statistics and modeling background

  • Analysis of historical data is not sufficient; knowledge extracted from data must be made actionable, preferably in the HMI/SCADA

  • To be really beneficial in an industrial context, a wide range of users should benefit from the Advanced Analytical tool, including operators, maintenance personnel, process engineers and other key operational employees

  • The ease at which Advanced Analytical tools connect and access data from a plant historian

  • How effective these tools are in enabling data preparation such as allowing for process lags and filtering out bad data and connecting to multiple data sources.

    • Analytical tool design

      Given these challenges, the most important consideration to be included in the design of analytical tools is that the value of information decays over the increasing time taken to become actionable, and is a function of the number of collaborating people and systems. A number of key requirements can be inferred from this.

      Tight integration — Analytical tools must either be able to interface to plant historians, or be tightly integrated. They must be capable of the seemingly basic but very real challenges of interfacing to multiple data sources, and automating data preparation such as cleaning up data or filtering it. Failing this, as much as 95% of the time can be wasted on the acquisition and preparation of data, instead of drawing conclusions and actions from it.

      Address complex, real processes — Analytical tools must be sufficiently powerful to address the complexities and scope of a real process, implying robust multivariate modeling techniques, and extracting non-linear relationships. At the same time, they must be easily usable by the typical process or automation engineer.

      Bridge offline, real-time environments — The next key requirement is bridging the chasm between the offline and real-time environments.

      In practical terms, this means that the knowledge discovered from historical data in the plant historian must be consolidated with current knowledge and best practices, and made available 24/7 in the HMI/SCADA in the form of real-time models with advisory actions to the operator.

      APC tools — For process optimization, the eventual requirement is seldom simply extracting knowledge from historical data and providing real-time causes and advisory actions. Advanced Process Control tools and methodologies are needed, with market-leading end user companies who apply this proven APC technology adding 3.5%-7% to their total annual gross profits, according the ARC Advisory Group.

      This approach can be described as moving from data to information to knowledge and finally, to deep understanding. The benefits to the end user increase, but so does the complexity. If this growth path to knowledge and understanding is mapped to automation products, it can be depicted as in Figure 2.

      In a typical plant historian, historical data is collected and enables reporting on the question, “What happened?” Properly integrating multivariate analytics with a standard plant historian enables an intelligent historian that helps to answer the question, “Why did it happen?”

      Similarly, in real-time, a standard HMI/SCADA answers the question, “What is happening?” Adding real-time model-based analytics helps to answer the question, “Why is it happening?” “What might happen?” and ideally, “What should be done?”

      The ability to do Advanced Process Control can automate this process by calculating the optimal set points that are then written automatically back to the plant.

      Fig. 1. Practical considerations in applying advanced analytics in process optimization indicate that the value of information is inversely proportional to the time it takes to become actionable. However, the value of information is directly proportional to the number of collaborating people and systems.

      Fig. 2. The growth path to intelligence using automation products leads to enabling the user to predict outcomes.

      Author Information
      John Leppiaho is product general manager for Production Management Software at GE Fanuc Intelligent Platforms. Derick Moolman is president of CSense Systems. GE Fanuc Intelligent Platforms and CSense Systems have collaborated to address current market requirements.