Statistical analysis of process data to improve the operation of chemical plants.

Modern refineries and petrochemical processes produce massive data with sensors measuring the plant process variables. These data help monitor the correct functioning of the plant, but operators rarely exploit the full informative potential that they contain.

We apply chemometrics and artificial intelligence techniques to transform data into helpful information that improves plant operation.

For example, data can be used to develop inferential sensors predicting the value of not directly measured variables in real time or to highlight operating conditions potentially triggering upsets.

Depending on the specific requirements, we analyse historical data to highlight hidden patterns and prevent undesired operative performance