Customer case: Production efficiency and cost savings with an accurate prediction model

Prediction accuracy is critical for a district heating provider  

Lappeenrannan Energia is a company that produces and provides district heating services in the city of Lappeenranta in Finland. The company’s production facilities utilise a mix of bioenergy, electricity, natural gas, and oil in district heating production.

For Lappeenrannan Energia, accurately predicting the required heat load on a day-to-day basis is essential for the company’s financial success. With accurate heat load forecasting, they can plan how to run operations as cost-efficiently as possible and determine which production assets to operate.

The Combined Heat and Power (CHP) plant produces both heat and power. If the forecast is accurate, Lappeenrannan Energia will know how much energy they will be able to sell on the energy market and reduce the risk associated with having to compensate for the difference between the prediction and the actual amount of energy produced.

For heating plants that operate on natural gas/oil or electricity, it’s crucial to have an accurate heat load prediction to know how much natural gas and electricity should be purchased. With accurate heat load forecasting, the prediction for the amount of natural gas and electricity will also be more accurate. Hence, the risk associated with purchasing too much or too little will be reduced.

Lappeenrannan Energia’s prediction system needed updating, and the Beyond Spot team was chosen as their partner in finding a new efficient solution.

Building a customised prediction service that helps Lappeenrannan Energia to strive

The Beyond Spot team set off to create a tailored and efficient prediction system for Lappeenrannan Energia. The Beyond Spot team analysed past district heating demand data and studied the dependency on the time of year, weekdays, and weather.

The Beyond Spot service forecasts the future heat demand. The heat demand forecast is created based on a temperature forecast and past heat load data. The operating order of the heating plants is optimised with the Beyond Spot control panel. Combining this information and the heat load forecast allows the final natural gas usage forecast to be formed. Lappeenrannan Energia's asset dashboard, which enables them to plan the future utilisation of different heat-generating assets, was also customised further. 

Throughout the project, the Beyond Spot team followed agile methodologies and continued to constantly adjust to Lappeenrannan Energia’s needs and situation. During the project implementation, Lappeenrannan Energia’s business environment encountered a sudden and crucial change – natural gas became scarce and expensive. Together, the project team considered different alternatives and was able to find a solution that allows Lappeenrannan Energia to more flexibly utilise different heat production resources.

An effortless, future-proof solution – the impact is tangible on many levels 

The Beyond Spot team created a prediction system that allows Lappeenrannan Energia to effortlessly and more accurately predict the amount of district heating Lappeenrannan Energia needs to produce. This allows Lappeenrannan Energia to streamline its operations and determine the most cost-efficient way to produce the required amount of heat. Furthermore, accurate heat demand predictions will result in more accurate predictions for the amount of electricity Lappeenrannan Energia will be able to sell on the energy market.

Lappeenrannan Energia is now able to benefit financially from more accurate predictions through a decrease in costs for prediction errors. Additionally, switching its prediction mechanism from software to a modern cloud service allows Lappeenrannan Energia to be more time and cost-efficient with a higher level of ease and trust for the system. Lappeenrannan Energia doesn’t need to worry about maintenance costs, the new tool is more scalable and flexible, and it can be accessed from anywhere. 

The prediction tool has been customised so that Lappeenrannan Energia doesn’t need to alter its existing processes. Moreover, the tool updates data and predictions automatically on an hourly basis, whereas in the past, all updates needed to be done manually. This also reduces the risk of human errors.

With time, the predictions have kept becoming increasingly accurate as the tool gets more data and is adjusted. When Lappeenrannan Energia conducted a district heating network operation optimisation project, the difference in data was instantly mirrored in the Beyond Spot tool. The formula learned and adjusted according to the new operation profile – with new data, predictions also changed.

On the other hand, there have been big changes in Lappeenrannan Energia’s operational environment recently. These changes in the operational environment are also continuously taken into account with the Beyond Spot tool.

The way the tool has been implemented allows further improvements and system expansions, e.g. more data parameters and further automation of processes. If there are changes in Lappeenrannan Energia’s operations or organisation, those changes can also be easily taken into account. Necessary adjustments to the tool will be made by the Beyond Spot team.

Benefit highlights: 

• More accurate predictions for the district heating need, use of natural gas, and produced electricity
• Financial benefits that are the result of more accurate predictions
• The prediction system is more time and cost-efficient, more scalable, more flexible, and it can be accessed from anywhere
• Automatically updating, more frequent predictions
• The predictions keep becoming more accurate with more data, while also adjusting to changes in the environment
• Future expansions and changes to the tool will be effortless for Lappeenrannan Energia, as the Beyond Spot team continues to support them