Reduce emissions in district heating plants with SUPERSENSE forecasts

Harri Lääveri
reporting emissions forecast

Live dashboards and near real-time emission reports play a critical role in determining the emission efficiency of district heating plants. Utilising SUPERSENSE's data analysis and reporting tool has helped Kemi Energy and Water Oy to reach their own emission targets - an essential step, given their ambitious long-term targets for reductions in emissions. 

In the future, it will be even easier for SUPERSENSE users to make the right decisions, as we have developed a machine learning-based forecasting tool that can predict production needs up to 66 hours into the future. 

Niels Bohr, a Nobel laureate in physics, famously stated that "prediction is very difficult, especially if it's about the future!" But at the time, Bohr did not yet have access to machine learning. Fortunately, we do now, and machine learning has allowed for a great leap in the emissions efficiency of district heating plants. 


How does SUPERSENSE'S machine learning work?

The machine learning works by combining historical production data and weather data from the Finnish Meteorological Institute, and comparing these with weather forecasts. Based on past production needs, SUPERSENSE can provide accurate estimates of future demand. As more data accumulates, the forecasts become even more accurate. 


How do production demand forecasts affect emissions efficiency?  

Emission efficiency, production efficiency and cost efficiency are all closely intertwined in the energy industry. When production can be kept at an optimal level, wasted power from overproduction is avoided, and oil does not have to be burned due to underproduction. Ultimately, this is beneficial both financially and environmentally.  


Optimising the use of combustible materials will soon be possible too 

Another important aspect in improving emission efficiency is optimising the production process itself and optimising the use of combustible material. By monitoring the overall situation and production demand forecasts, one or more district heating plants may be optimally managed, but that does not necessarily influence what materials are used in production. 

We are now developing our own tool for optimising the production process that can determine to what extent, for example, wet peat and dry biomass should be burned for heat production. This will be linked to production demand forecasts, in order to make recommendations on which production methods would be efficient in terms of both emissions and cost. 


How is SUPERSENSE used? 

The SUPERSENSE technology is largely based on Microsoft solutions. In principle, taking advantage of SUPERSENSE only requires the customer organisation to use Office 365. The data comes directly from the district heating plant’s control room system and IoT-solutions, and all data is owned by the energy company itself. The data is securely transferred to the Microsoft Azure cloud service, where SUPERSENSE analyses and aggregates the data into a useful format. Using Power BI, different visualisations of the production data can be created and displayed, and rights to these can be securely provided to anyone requiring them for their work.  

The forecasts also utilise the built-in machine learning resources of the Microsoft Azure cloud service. Deployment is quick, and results are achieved almost as quickly. It is clear – improving emission efficiency should begin today.  

Harri Lääveri

Save electricity with SUPERSENSE Eco

real-time data eco electricity
Antti Lehikoinen

Monitoring decentralised district heating production with Supersense Analytics

supersense real-time dashboard district heat

Show all