Increasing data historian efficiency

Modern process control systems are continuously increasing the number of process data points monitored in a process control system. These data points, used for process control or energy/environmental management systems, can be brought into a process control system through one of many of possible data interfaces (i.e. 4-20mA, HART, Modbus, OPC, Fieldbus, Profibus, ProfiNet, etc.). The majority of these data points are not only used for process control, but are also logged into a historian database to be analyzed and used in calculations for reporting. Corporate historian databases provide process and plant management teams' access to process and environmental status reports and dashboards, allowing the ability to monitor process, environmental performance, and Overall Equipment Effectiveness (OEE). However, to create useful reporting values, the data from the real-time data historian must be compiled, statistically combined under different aspects such as time or production state, and stored in data warehouses prior to viewing. Depending on the complexity of these calculations or the amount of data to be analyzed, the data compilation process can be time consuming and inefficient to the operation of the data historian system. Overall, this database architecture has several points of inefficiencies that are compounded the higher up the ladder the data is transferred. In many instances, these inefficiencies can be directly translated into increased application costs. These increased application costs can be seen in initial historian engineering, engineering effort to create reports, and reoccurring historian engineering costs whenever there is a change on the plant level. What these engineering costs do not reflect are the hidden costs that exist, such as overloading the business/IT network with process data and reporting delays while waiting for queries to populate production data calculations. Understanding what type of data is useful at the different levels of a process control system is important as well. Straight real-time data logging for trend purposes is useful for plant operations but does not serve as much of a purpose for corporate process management teams. Compiled, statistically combined, and calculated process values such as Key Performance Indicators (KPIs) and process efficiency data is useful to corporate process management teams, but this information in real-time can be just as useful to control operators, allowing them to adjust and react faster to process changes and environmental conditions. The information contained in this paper describes how moving select data calculations from the Historian level to the Process level can improve the overall efficiency of the data historian operation. Calculating process data where it is collected improves efficiency and simplicity of the historian system. In many cases, this results in the reduction of requirements on hard drive space and database processing power requirements. Besides a standard process control system, this paper will identify systems such as a CEMS (Carbon Emission Monitoring System) and other energy/environmental management systems where historian database management is important. The examples included in this paper describe how this method of pre-compiling data into an array of data prior to historian database storage increases overall data historian data acquisition efficiency.