The Energy Performance Certificate predictive model
The NEEM modelling approach enhances existing energy signature models using detailed data on weather, heating use, and energy use. This report outlines the challenges encountered during the implementation and testing of our approach on buildings in pilot studies.
The main issue revolves around the need for non-linear optimisation methods. Even though the model specification is relatively simple, the complexity of real-world data required us to take extra caution to ensure numerical stability. Furthermore, non-linear optimisation methods often suffer from poor scalability. Consequently, we also implemented our model in C++, a programming language known for its speed. However, this presented a new set of issues which needed to be addressed.
Our efforts culminated in the creation of scalable modular software which can be used to simulate energy use. Over the long term especially, the model exhibits a high level of accuracy. Additionally, it can provide an inexpensive overview of building characteristics related to energy efficiency. This, in turn, can be used to estimate a building’s EPC label.
Even though the model was mainly used to assess buildings using district heating, its applicability extends to other areas without concerns about its validity.