The most attractive target for utilizing digitalization in energy and process industry is improving process efficiency – especially in the early phases of “digi-journey”. Process data contains massive amounts of specific process information that can be used, utilizing methods of digitalization, to improve production and operations. This is true as long as process data quality if fit for the purpose.

 

What is a sufficient quality?

Process data consists mostly on measurement values, and uncertainty (i.e. smaller or bigger measurement error) is always there. Measurement errors vary in time, and this makes analysis of process data extremely difficult.

Process efficiency KPIs, such as operational efficiencies, material efficiencies and emission levels are calculated from many measurement values. Minor error or change in these factors results as major change process efficiency, and through that also on company’s profitability.

Accuracy of these KPIs should be at least on a level of 1 – 2 %, so that they can be used to improve process efficiency. Good news is that digitalization has enabled continuous validation of process data quality.

 

Ensuring sufficient accuracy of process data

Continuous balance analysis of process (or its part) is a central method for process data quality validation. Measurement errors are recognized when the balance functions do not add up. When there is a plenty of data and inter-correlated measurements, balance analysis provide error estimations for measurement values. The most significant estimated errors are validated via field calibrations before corrections are made for measurement values.

Data accuracy is sufficient when balance analysis doesn’t recognize errors that would have significant effects of conclusions drawn from process data.

 

Our experience

Today Indmeas has experience of validating process data quality for 20 customers. The working procedure is typically of the following form:

In the kick-off phase a remote connection is implemented for sharing process data, balance functions are defined and necessary field calibrations are conducted. Typically, after few balance analysis rounds and validation calibrations measurement quality is sufficient for producing information for improved processes and operations.  

In the continuous phase the process data quality is secured via monthly balance analysis. The observations of these typically lead to few validation calibrations annually.

Summary

Continuous validation of process data quality is “a must” when digitalization is used for creating valuable information out of process data for improved processes and operations.

 

Risto Kuoppamäki

Indmeas Oy

CEO

 

We help the energy and process industry to utilize process data in daily operations and business development.