Summary of positions and project

We are looking for four researchers who together with a researcher from the field of machine learning will form a team responsible for developing an advanced parametric modelling approach based on computational building performance simulation to support the decision-making of the sustainable renovation of dwellings. This approach aims to integrate: 1) the broad range of housing characteristics present in the existing housing stock; 2) the wide variety of renovation measures ranging from insulation, to HVAC systems to renewable energy technologies; 3) the various user profiles and behaviours. For this full design space of alternatives, a fit-for-purpose simulation approach will need to be developed to assess the performance of a broad range of KPI's (costs, comfort, health, CO2, heating demand, peakload of the grid, etc), in order to be able to find 'optimal' solutions that support decision-making. For these positions we are looking for four researchers with extensive experience in the field of building performance simulation, either in the form of a PhD or otherwise. The four positions will focus on: 1) housing characteristics and variations; 2) renovation measures (insulation measures, HVAC systems and renewable energy technologies); 3) user profiles and behavior); 4) key performance indicators and the associated modelling choices. By applying to this position, you will automatically be considered for all four positions. If you have a preference for a specific position, you can outline this in your motivation letter for the application.  

Long Project Description

Frequently asked questions when it comes to the energy transition of the existing housing stock are:

• Does my home have to be a Zero Energy Building? Is cavity insulation enough or not?
• With a heating network of 50 degrees, to what extent should the homes in that district be insulated?
• What are cost-optimal solutions?
• And what about the winter- and summer comfort and health of the indoor climate?
• And what about the load on the electricity grid?
• How will construction companies be able to standardize and industrialize?

In the Netherlands, and more broadly in the EU, these questions cannot be answered with the energy performance calculation models that are typically used as a basis to give sustainability advice to home-owners. This project aims to answer these questions. There are four main features that distinguish this project from these existing models:

• Monthly & Static Climate vs. (sub) hourly & dynamic climate; Firstly, many of the (advice) tools used are based on models that use monthly average calculations. This means that questions about summer comfort and winter comfort can be answered very poorly, because these are dynamic aspects that cannot be assessed on the basis of a monthly average. The same applies to aspects such as a healthy indoor climate and peak load on the electricity grid. This is why dynamic simulation models are needed (such as EnergyPlus or TRNSYS).
• Average user vs. real user; Secondly, existing models calculate with a 'standard' user, while it is known that user behavior can have a great influence on energy savings, and it is therefore impossible to determine this properly with a 'standard' user. That is why this project uses a multi-zone simulation model with a wide variety of user profiles and behaviour, and a link is made with measurement data of current use, so that more realistic user behavior can be calculated. As a result, the potential energy savings can be determined much more accurately, and it is also possible to search for 'robust' solutions that are less sensitive to user behaviour.
• Average home vs. real house; Thirdly, existing models calculate with an average home per housing type (e.g. terraced house 1945-1965). This while a recent study by Eindhoven University of Technology has shown that the insulation measures required to switch to low-temperature heating also differ greatly within a type of house. That is why this project calculates a wide variety of housing characteristics, to make it possible to determine the optimal solution for individual homes, which means that considerable costs can be saved. In addition, work can be done on a more intricate housing classification with which the balance between optimal solution at housing level and broad rollout capability across the housing stock can be sought.
• Few standard renovation solutions vs. determining optimal solutions based on all possible combinations of standard measures; Fourthly, existing models often calculate only a few solutions / scenarios, or the user has to compile a package of measures by hand, while in reality the amount of solutions is very large, and a good advice tool should help the user to find the right combination of measures to achieve a certain performance. That is why this project calculates a large number of renovation solutions, including various heat sources, heat distribution systems, ventilation systems, etc.

This project therefore calculates all possible renovation solutions with (sub) hourly, dynamic models for a wide range of users and homes in the Netherlands. For each combination of renovation measures, this provides insight into the performance in terms of summer comfort, winter comfort, healthy indoor climate, costs (both investment, energy savings and total costs), CO2 savings, and more. Various 'optimizations' can then be made on the basis of these results. In this way, it can be determined for each house what the cheapest way is to prepare the house for switching from natural gas to the future heat source. But it can also be determined how much more CO2 can be saved with little extra investment, and how summer and winter comfort can be optimised. This makes it possible to arrive at much better advice, which is of better quality and leads to more cost-effective solutions.

The two biggest scientific challenges

The development of this project involves the following two biggest challenges:

• Complexity and uncertainty management. Developing this project is not simply choosing a calculation/simulation model and then simply modelling all possible variants of homes, users and renovation measures and all possible KPIs. This would quickly become impossible due to the exponential nature of such modelling approaches. It is therefore very important to conduct extensive research into how to develop this project 'fit-for-purpose'. Not more complex than necessary, but complex enough to generate the answers and insights that matter. For each KPI, it must be determined which model complexity is required to accurately determine the KPI, given the available input data and the uncertainty thereof. These choices can be made by gaining insight into the trade-off between the approximation error of models and the uncertainty of the required input. For the most influential parameters, greater model complexity and greater input certainty will be important, while this will be less important for less influential parameters.
• Data management and big data analysis. Even when careful choices are made in the development of the input variation of the off-the-fly parametric models, there will still be an extremely large amount of simulations and outcomes. Processing, storing and analysing the many combinations of input and output requires expertise and use of High Power Computing, Big Data Management, and Data Science. In addition, it is important that the models are not only developed open source, but that the code is constructed in such a way that a large team of programmers can work on it, and that the structure and input are transparent for peer review by experts.

The team you would be part of consists of the four researcher positions below, and an additional researcher position will be opened on the topic of data analysis and machine learning. These five positions are the 'core' of the team, but we envision this team to grow in the coming year.

The team will be embedded in the Building Performance research group - https://www.tue.nl/en/research/research-groups/building-physics-and-services/building-performance/

Four researcher positions:

1. Housing Characteristics: identify common geometrical variations; parametric modelling of housing variations; sensitivity analysis of housing variations; clustering analysis of housing variations based on optimizations of renovations;
2. Renovation Measures: modelling insulation measures, HVAC systems, renewable energy technologies; modelling implications of specific technologies; parametric modelling of renovation measures; sensitivity analysis of renovation measures
3. User Profiles & Behavior: modelling user profiles and behavior; sensitivity analysis of user behavior; decision-making under uncertainty; machine learning / surrogate modelling / neural network approach;
4. Key Performance Indicators: evaluating winter comfort, summer comfort, indoor health, grid-dependence and load, costs, energy demand; modelling implications of key performance indicators; sensitivity analysis; decision-making under uncertainty; fit-for-purpose-modelling

By applying to this position, you will automatically be considered for all four positions. If you have a preference for a specific position, you can outline this in your motivation letter for the application.  

We are looking for excellent and highly motivated candidates with either a PhD in Architectural Engineering, Building Science, Building, Civil, Mechanical or Environmental Engineering, or a MSc in these fields and extensive subsequent experience in computational building performance simulations. Interest in physical processes (heat, air and moisture transfer) in the built environment, sustainable energy technologies and extensive experience in computational building performance simulation is essential. The team will need to work in a very collaborative way to ensure a good outcome, so in addition to the expertise listed below under job requirements for each position, you would need to have personal characteristics of a team player, such as: flexibility, active listening, problem-solving, effective communication, positive attitude.

We offer a stimulating and ambitious research environment. To complement this environment and for the specific projects mentioned above, we are looking for outstanding candidates that meet the following requirements:

• You are highly motivated, talented, enthusiastic as indicated above.
• You have a relevant MSc degree as indicated above.
• You have extensive experience in computational building performance simulation as indicated above (preferably a PhD, but not necessarily). Preferably extensive experience with EnergyPlus and Python. Affinity and preferably experience with data analysis.
• Ability to conduct high quality academic research in the field of building performance simulation, demonstrated for instance by a relevant PhD thesis and/or publication(s).
• Excellent mastering of the English language, good communication and leadership skills.
• Be a team player and able to work in a dynamic, interdisciplinary context.

Specific for positions:

• Particularly for the researcher focusing on housing characteristics an interest and experience with AI / machine learning is preferable.
• Particularly for the researcher focusing on renovation measures, experience with HVAC systems modelling is essential.
• Particularly for the researcher focusing on user profiles and behavior, an interest and experience with AI / machine learning is preferable, and experience with modelling user behavior is essential.

• A meaningful job in a dynamic and ambitious university with the possibility to present your work at international conferences.
• A full-time employment for 2-4 years (dependent on the position).
• You will have free access to high-quality training programs for research and valorization,  professional development courses for PhD students, and didactical courses of the TEACH training program.
• A gross monthly salary and benefits in accordance with scale 10 (min. €2.846 - max €4.490) of the Collective Labor Agreement for Dutch Universities.
• A broad package of fringe benefits (including an excellent technical infrastructure, moving expenses, and savings schemes).
• Family-friendly initiatives are in place, such as an international spouse program, and excellent on-campus children day care and sports facilities.