Sensitivity Analysis of NTA8800 for a Dutch Building Renovation Tendering System

To achieve the national 2030 climate targets, we will have to rapidly increase the pace of renovation projects to over 50,000 existing homes per year by 2021. One of the barriers to achieving higher renovation rates is that most renovation activities are still at the craftsman stage [1], in which they are planned and implemented in separate approaches per dwelling. Given the mentioned energy efficiency targets and deadlines set by EU countries, including the Netherlands, the housing renovation process needs a substantial upscaling. To this end, renovations’ requirements need to be elicited in a structured format such that the building industry can offer standardized solutions that allow large scale, high speed construction processes. These requirements should be incorporated in a publicly accessible digital platform enabling required data for heat transmission, CO2 emission, and construction cost calculations to prepare the foundations for tendering renovation clusters allowing the construction companies to adapt their production process and achieve economies of scale.
During this process, building energy simulation systems, especially those designed based on national-level regulations (e.g., NTA8800 in the Netherlands), play an essential role since they can calculate current building energy performance if the required information had been provided. Performing a holistic energy calculation requires two types of energy usage information. First, building-related energy consumption, which is related to the building's ability to avoid heat loss in winter and heat gain in summer and maximize natural light use. Second, user-related energy consumption, which is related to occupants' gas, electricity, and water consumption for cooking, domestic hot water, and electric equipment and appliances [2]. Although many of these inputs can be correlated based on their calculation methods, the required information must be aggregated from various data sources and seeking all of them simultaneously would be a burdensome and time-consuming task. When it comes to mass-scale renovation, the situation becomes even worse since data collection and information production must be handled not for a single house but thousands of dwellings. Thus, prioritizing input information is essential to facilitate and illuminate the orderly way of acquiring preliminary information.
This report's conducted study aims to ascertain the prioritization of the most and least influential parameters for determining the total heating and cooling energy demand (BENG1) derived from NTA8800 regulations. For that purpose, two well-known global sensitivity analysis methods, commonly used in the building energy performance simulations (Morris’s and Sobol’s methods), were selected and applied to the WoonConnect energy calculation software, which has been developed based on the NTA8800 regulations. Based on the literature, fifteen physical characteristics of a typical three-floor row dwelling were selected as the input factors for the sensitivity analyses. The analyses were performed with an adequate number of simulations to ensure results’ robustness, which was further confirmed by attaining the same results from both applied methods. From the obtained results, the previously chosen input parameters were categorized into three priority levels. The top priority level gathers the parameters with a sensitivity measure greater than 0.05. The parameters in this category dramatically influence the heating and cooling energy demands, so providing this information has the top priority, and information must be as exact as possible. This priority level includes mostly dwellings’ geometrical properties. The next priority was assigned to the parameters with sensitivity measures between 0.01 and 0.05, which moderately impacted the BENG1 value. However, since they have mostly non-linear behavior, the uncertainty in information provision should be minimized as much as possible. All the parameters that reside in the second priority level directly relate to buildings’ thermal properties. The last category encompasses all the other parameters with negligible effects on determining the BENG1 value. These parameters are mostly associated with the sun radiation part of energy demand calculations.
Within the following phase of the project, we will focus on providing information for required input parameters based on their priority level. Thus, in the next step, we will attempt to determine dwellings’ geometrical properties from available data sources. Publicly available point clouds derived from aerial LiDAR surveys are valuable sources to extract dwellings required geometrical information. Because they are available countrywide, it is possible to use them in large-scale information extraction to accelerate dwellings’ energy demand calculations and subsequently facilitate upscaling of the renovation process.