Bottlenecks and challenges in processing energy efficiency data

There is currently much talk about machine learning and artificial intelligence but unfortunately, at industrial plant floor level, things seem to have remained in another decade. The further we move from the core process, the less data is available for different inspections. When checking energy efficiency, for example, these challenges come strongly to the fore.

When checking energy efficiency, these challenges come strongly to the fore.

When trying to collect data for intelligent solutions, the first bottleneck is often the amount of available measurements, and possibly the location. When investigating the performance of heat exchangers, for example, flow measurements and temperature measurements are often unavailable. In some cases, missing information can be replaced with data from other parts of the process. This data may provide indications regarding phenomena, even though data regarding e.g. absolute output or efficiency are unavailable..

Even if measurements are available, the next bottleneck may be the storing of measurement data. Typically, data required for energy efficiency optimization is stored in process control and building management systems. Process systems are commonly supported by data collection from which measurement data is easily extracted.

The situation in building management systems, however, is often the opposite. Such systems, for one reason or another, are less like to be supported by databases, which means that data is not necessarily stored at all. Sometimes, it may have been possible to save data on a computer hard disk rather easily, but the feature was not taken into use. In many building management systems, it is also possible to set up selected positions to capture data, but the amount of measurements captured may have been limited by licenses. It should also be kept in mind that capturing always needs to be activated separately.

A third possible bottleneck is the efficient utilization of information. Challenges may include, for example, a lack of experts or the disjointedness of data in different locations. In recent years, energy managers have become more common in industry, but in many places, such matters are handled by other managers for whom this is not a priority. Even if appropriate energy experts are in place, the disjointedness of the data in different places can be a challenge. In such cases, gaining an overall picture can easily become difficult.

What solutions are available to meet the challenges and reduce bottlenecks? Firstly, field measurements have been given an unnecessarily bad rap by both industry and service providers. In many cases, targeted field measurements and data logging can generate a very comprehensive picture of the situation, if an experienced process expert conducts the measurements. If there is an ongoing need for measurement data, fixed measurements should naturally be increased.

Targeted field measurements and data logging can generate a very comprehensive picture of the situation.

Needs-based data collection is a solution for measurement data storage. When the entity is developed according to needs, costs do not necessarily increase markedly. Special care should be taken to ensure that important features are included in the tendering phase also when procuring new systems. A needs-based approach can also counter data disjointedness. By planning data transfer overall, one can often find ways to collect data in one place from where compiled reports can be generated. A solution to the lack of resources can be the use of a service provider and a “virtual team” for the plant, where both parties have defined roles and the operational model is light and clear, in which case help request and responses thereto can be easily handled.

Challenges in processing energy efficiency data are not insurmountable. Solutions can be found when the matter is approached comprehensively and systematically. Traditional field measurement work should also not be discarded, even though continuous measuring of essential factors is nowadays the norm.

Read more about Elomatic’s data analysis services here.