Probabilistic approach to the assessment of uncertain input parameters when energy renovating existing buildings

Since the building stock stands for significant part of the EU’s total energy use, energy renovations are needed in order to lower the negative environmental impacts. In Sweden, only 1 % of the total building stock is newly built i.e. the focus should be directed upon existing buildings. Moreover, most of the buildings in the “Million Program” from 1970´s are in need of renovations in order to keep them operational. This makes for excellent opportunity to energy renovate these buildings at the same time, and follow up the work that can later be implemented to all other types of existing buildings. This project is about trying to determine an existing building total energy use over a year, which is the main question that needs to be answered when it comes to determining the viability of energy renovations on existing buildings. For this, computational simulations were performed as well as statistical methods such as sensitivity analysis and Monte Carlo simulations. The statistical methods are used instead of performing conventional expensive and time-consuming measurements, when obtaining the values of the input parameters used for the simulation tool. In addition, installing the heat recovery on the current ventilation system is also looked upon in order to see what additional energy savings could be made. The results show that the building total yearly energy use should lie within the interval of about 160-190 kWh/(m2·year) and that the average use should be around 177 kWh/(m2·year). After installing a heat recovery system, the building total energy use should lie between 115-135 kWh/(m2·year), which is a decrease of around 24-35 % in total energy savings.When trying to predict a building’s yearly energy use with the help of a computational tool, many of the input values often prove to be uncertain. Par example, the actual U-value of a wall that is several decades old could never be assumed nor calculated correctly, due to old age and worsened condition of the materials. Therefore, this paper investigates a strategy that uses statistical methods, by which these uncertainties in the input values are accounted for. In order to know if it is viable to energy renovate an existing building, and to what degree, it is important to know its current yearly energy use. At present, computational simulations offer the easiest way of finding this out. However, many different input parameters need to be known before making any simulations. This information can be obtained by measuring them in real-life, or as this paper presents, using statistical methods instead. Thirteen input parameters were processed in this paper. Different guidelines and regulations were used to find out what is the highest probability that a value of a certain input parameter should adopt for the particular type of building. Then, this value was varied into four additional steps in order to compensate for the fact that it is uncertain. Moreover, a random number generator was used to pick one value per input parameter at a time, and run a simulation with the tool. This was done 100 times which lead to an interval being obtained, within which the actual building yearly energy use should fall. In addition, a test was made in order to know what the energy savings would be by installing a heat recovery onto an existing exhaust air only ventilation. The results of this test showed that the energy savings should be around 30% only by making this single change to the building. It is important to keep in mind that this project is purely theoretical, and that the results are based on uncertainties. However, statistical theory provides a solid ground for the purpose, which backs up the results obtained in this paper to a certain degree. Nevertheless, further studies are needed in order to implement the methodology presented in this paper to other types of buildings