HOT|COOL NO. 2/2019 - "Smart Heating System Integration"


heat installations. Their assumption is that by comparing data from these two sources they may, through machine learning, be able to predict what will happen on the secondary side. This is a great example of the fact that because this area is still unchartered territory, asking questions and testing assumptions is the necessary way for a utility to go. It also highlights the need for utilities to use their expertise and data to take more responsibility by providing the framework to better guide end users and plumbers to make the right decisions.


In addition to enabling temperature optimisation, smart meter data, data from additional sensors and advanced analytics also unlock new opportunities for e.g. pressure optimisation.

Figure 2

As with network temperatures, detailed insight into where the network pressure is either higher or lower than necessary will lead to more efficient use of time, resources and energy. As an example, if the digital twin of a utility’s distribution network could show not only temperature levels, but also calculate and show pressure levels, utilities would have full transparency on the two parameters they require to deliver district heating in the most efficient way possible. We are working on models that can assist utilities in making exactly that evaluation, and we have already seen very positive results. It is also worth mentioning, that the optimisation potential mentioned above relates only to the operations side (OPEX). Because of its both cost-intensive and time-consuming nature, the investment and asset management side (CAPEX), which includes infrastructure expansion and lifetime costs, may hold an even bigger savings potential through data-driven optimisation. In other words, not only is digitalisation necessary to create the right conditions for low-temperature district heating and the integration of sustainable energy sources in a green energy future. It also delivers black numbers and increased competitiveness for district heating utilities.

Increasing easiness could be done through data-driven decision support for both end users and professionals in the form of a tool that makes it easier to find both the cause, location and solution to a given problem. This could include using smart meter data to identify e.g. the 10 most common heat installation errors and possible solutions. However, that will require scalability and continuous monitoring. LARGE-SCALE BIG DATA First of all, the way data is used by utilities has to be more scalable. Utility experts should not be the only ones able to decode data and graphs to find and fix the most typical issues. These conclusions must also be made available to home owners and plumbers. Errors could be wrong dimensioning, faulty or worn out components, but often it is a question of making simple adjustments. What is complex is finding the cause and solution – especially for those looking at pipes rather than a computer screen. That takes the right data, delivered at the right time in the right format. Secondly, fixing a specific problem and adjusting a heat installation based on current conditions will only last so long, as those conditions will change over time. Continuous monitoring will help utilities detect changes that will negatively affect the performance of the installation. This calls for digitalisation and machine learning from the large amount of smart meter data available to make this process automated and dynamic. In the project mentioned above, our customer often sees cases where the installation performance varies significantly in both old installations and new ones that should theoretically outperform the former. They are therefore now testing their idea to supplement energy meters on the primary side with additional temperature sensors on the secondary side to get even more information about what actually goes on in the

Steen Schelle Jensen, For further information please contact:


Made with FlippingBook Ebook Creator