We would like to thank the residents of the Vieille Motte district in Neuville-en-Ferrain for their time and patience while completing the questionnaires. We would also like to thank Gautier Despin, master student at ENSAPL at the time of the research and who worked closely with the inhabitants and data collection. We would also like to thank Sabrina Chenafi and Téva Colonneau who actively collaborate on the development of this research. Finally, many thanks to Virginie Thomas (Puca), Philippe Rémignon and Valentin Breton (Vilogia).

Introduction

European and French standards for the energy retrofitting of the residential housing sector

12In addition, the heating regulation RT2012, which is intended to respond to the directives and the french political debate on sustainable development called “Grenelle de l’Environment” (2007), includes the use of conventional calculation methods, certifying a level of energy performance to be achieved. These calculations take data arising directly from Th-BCE regulatory methods of the RT2012, combining the analysis of genuine collected data with pre-determined data arising from hypothetical usage scenarios. This approach presents performance requirements based more on reports emphasising control and regulation systems than on the actual thermal gains and the energy use of the buildings, notably in housing. In actual fact, in the framework of a dynamic energy simulation (DES) study for example, the RT2012 imposed the use of determinist, referenced behavioural models. These are characterised by user scenarios which differ depending on the relative sector of activity (shared housing, individual housing units, offices, etc.) and simulate user activity over time. Nonetheless, such an approach limits the user scenarios, whether through lack of behavioural diversity or their repetitiveness, as they focus on a series of ‘default’ user scenarios. This makes it difficult to interpret energy use and detailed thermal analyses. Indeed, their application limits the calculation of internal gains in the housing units where residents are present, without really considering residents’ genuine behaviour and the energy use observed. In addition, the occupancy variable is simplified and averaged out in the calculation and its value does not always reflect the actual results observed.

Method. A post-occupancy evaluation: based on detailed knowledge of resident behaviour and energy awareness

16Our methodological process rests on a continuum which spans the digital maquette, the construction and materiality, as well as the potential gains and historical, architectural and societal issues. It has the advantage of being able to manage numerous factors in retrofitting projects by exploiting, optimising and testing them. This general method is based on an interactive, collaborative approach that brings together the digital universe and the real world.

17As we see in fig. 2, the method is developed over several stages: after a phase of detailed analysis of the building based on archival research and data collected on site, a 3D model was created in a BIM environment. This allowed us to integrate several important data to construct energy simulations (about the site, the spatial layout, structure, building materials and stratigraphy of the envelope). This stage centred principally on the treatment and representation of analytical, constructive data on the architectural, urban characteristics of the estate. We chose to think of it like an informative, adaptable ‘processing tool’ so that the units modelled could reflect a faithful interpretation of the genuine conditions of the layout of built-up areas and (housing) non-built-up areas (landscaped spaces, roads and pathways, green spaces). In the continuation of this article, after explaining the process by which we constructed the DES, we shall focus on the last two stages of the methodological schema: strategic interventions and awareness campaigns.

The survey protocol for the characterisation of resident behaviour

19To determine typical behavioural models, as precisely as possible, the questionnaire was designed to be scalable and adaptable. The survey phase marked a key stage in the methodological process of the study, by facilitating the communication of the aims of our research to the residents. Based on the number of answers, the phase of analysing data obtained was undertaken concurrently with the following stage, relating to estimations of hourly occupancy modes and the management of heating, activities and daily energy consumption. This questionnaire was the first factor in building a personalised database (dataset) for each resident, which would objectify the energy audit based on an in-depth analysis of the variable data which are neglected by regulatory calculations.

Proposal of a behavioural model calibrated on the detailed profiles

20Behavioural modelling of residents proved to be a delicate stage of our methodology. To better manage the complexity of the notion of comfort and its multifaceted definitions (thermal, visual, acoustic, etc.) and the subjective character of the perception of same, this phase aimed to outline an in-depth model of residents’ behaviour calibrated on genuine profiles established in situ through the surveys. These questions were answered implicitly by the very modelling of the residents’ activities: the hypothesis being that, even if comfort phenomena are not modelled, the residents’ actions reflect, amongst other things, the quest for comfort.

Dynamic energy simulation of the detailed behavioural model: a key step for the scenarisation of intervention strategies

22With the aim of preparing an in-depth analysis of current residents, we constructed a framework which facilitated an investigation of the building and its overall function that was made possible via a digital maquette using BIM methods (Building Information Modelling - capable of translating the modelling of the information or data on the building).

23This stage of the research aimed to create a faithful representation of the existing buildings and on-site conditions. This phase dealt with the analytical and constructive data, architectural and urban characteristics of the estate being studied. Integrating the use of such an all-encompassing, scalable ‘processing tool’ allowed every element to be modelled and provided us with a faithful interpretation of the genuine conditions in the built-up areas (housing) and those that were not built-up (landscaped areas, roads and green spaces). We examined the existing building in several stages (structures, spaces, envelopes, site parameters…) a necessary process for the creation of the DES; we used the digital model to support our exploration and analysis of the site which helped us to highlight the building logistics and the actual typo-morphological features of the building under investigation (fig. 4).

24From that point on, we undertook the first analyses of sunlight and its overall energy impact on the housing units. This enabled us to complete the preliminary phase of developing our knowledge of the existing building. In addition, thanks to the software tools used during the DES phase, the digital model facilitated the analysis of the different typologies of housing observed in relation to their function and interaction with geographical factors on site.

25So, we construct behavioural models based on data collected directly from the residents. We will establish in-depth profiles, which include the occupancy variables, then subsequently organize them via the implementation of the elements collated through the DES software present in the same BIM environment as the model. This allows us to define the genuine occupancy variables and energy consumption in the case studies under investigation (fig. 5).

26By defining the input methods specific to the chosen calculation engines to carry out a DES, the quantitative translation of genuinely observed occupancy data which subsequently permitted the hierarchical, tree-like construction of the in-depth profiles. They responded to variables depending on the individual data entry system of each tool we used and are generally categorised in the five variables, previously mentioned. For each category, the design of the time and/or seasonal timetables allowed us to guarantee the division of activities and occupancy of each room, while simultaneously integrating the usage patterns and needs of every profile, and also facilitated the use of all the information gathered by the BIM digital model.

27At the end of this implementation phase, it was possible to carry out the energy behaviour study of the estate being assessed and took both the “regulatory profile,” standardised via the RT2012, and the in-depth profile into account, thus including the actual data of residents. The comparison of the DES results according to these two methods clearly demonstrates the impact of genuine behavioural models on the functionality of the estate being assessed. These observations were only made possible through the following analysis criteria: taking stock of the residents’ behaviour in their building environment.

28The energy balance reveals the necessary energy balance between energy expenditure and energy gains in the building studied. It takes account of both the behaviour of the residents in their own environment and the impact of the residents’ interactions with the building.

29Therefore, highlighting occupancy variables as a genuine issue in the improvement of energy efficiency and thermal comfort makes it possible to consider the construction of third-party scenarios integrating the characterisation of “resident energy profiles”. These hypothetical scenarios, verified via their application in the DES and established by adjusting the gains and gaps in each in-depth profile, are therefore potentially viable in the construction of an energy retrofit strategy. Consequently, to ensure adequate consideration of the identified energy scenarios, it makes sense to include the residents actively in the realisation of an information phase and raise their awareness about the Renovation 2.0 method.

The self-assessment survey on residents’ behaviour and satisfaction levels

30Raising awareness about the effects of energy use and behaviour aims to encourage residents to put the Renovation 2.0 approach into action in their daily lives and allows them to:

31- Implement future interventions with non-standardised, minimal logic.
- Prioritise the interventions based on experiments undertaken on the variety of residents’ energy behaviours.
- Conserve the original architectural features of the buildings by avoiding full-scale overhauls of the ETI – external thermal insulation.

32Renovation 2.0 is based on a case-by-case approach, which begins by measuring factors which have even the slightest impact on the building:

33a. Socio-behavioural measures are first deployed from a sociological point of view relative to the issue. This means putting measures in place related to the occupancy of the spaces and the residents’ behaviour. They encompass a change in the way the building is occupied and inhabited, the improvement of the management of systems through the promotion of more “eco responsible behaviour”.
b. Socio-technical measures, where both social and technical issues have little economic impact. It deals with the management and behaviour of the residents and appliances in their homes that are not inbuilt. These measures include a move towards greener household appliances and minor changes to the heating, lighting and ventilation systems, or even modifying the joinery.
c. Technical measures and targets concern larger scale interventions on the distribution of internal spaces and the building envelopes, internal thermal insulation (ITI) or ETI. These types of intervention carry certain risks and greater cost which should take each building’s characteristics into consideration.

34Defining a hypothetical occupancy scenario with a socio-behavioural character is carried out at the end of the first phases of the DES, the assessment method suggests adapting the actual household energy consumption based on the findings and results of the in-depth analysis actual household consumption, to determine whether substantial energy savings can be achieved because of the awareness raising effort.

35The “Self-assessment” part of the awareness raising questionnaire was presented to all residents. It was in the form of a multiple-choice questionnaire, made up of 11 sections organised in three parts: “interaction” (the ability of residents to adapt their behaviours for improving energy performances), “observation” (the ability of residents to understand the energy effects of changes in their behaviour) and “control” (the ability of residents to manage adaptive thermal comfort).

36The “Assessment of Levels of Satisfaction” section of the awareness raising questionnaire had four parts, designed for residents to fill out on the phases of the previous study. The first presented a summary of the results, the factual and quantitative factual data, a brief overview of the occupancy rates and the thermal gains measured. The second, presented a hypothetical ideal scenario and potential savings to heating (thermal gains). The information in these first two sections was simplified and condensed to facilitate the residents’ understanding of the data, and research methods. The third section proposed energy saving suggestions and recommendations specific to each household. These took the form of simple changes, identified in the DES in the hypothetical scenarios. The fourth and final section of the questionnaire invited the residents to express themselves with regards to the energy-saving suggestions in the hypothetical scenarios compared to what they do at home. They were also asked to comment on the progression of the study with regards to a potential renovation 2.0 on their homes.

The case study: a social housing project in the Haut-de-France region

The Vieille Motte district in Neuville-en-Ferrain: Composition of households and housing types

37The social housing estate situated in the Vielle Motte neighbourhood in the town of Neuville-en-Ferrain in the North of France, displays architectural characteristics like other estates with individual housing units in the greater Lille area (fig. 4 and 6).

38The “Gémeaux type houses” were erected in 1975 with a concrete slab and brick wall façades for two-room flat (T2), a four-room flat (T4) and six-room flat (T6). The T2 and T4 flats are on the same level, while the T6 flats have a partial storey.

39These typologies display interesting and revealing organisational and spatial qualities in relation to the urban morphology of the site and the estate. They bear witness to a time which was marked by problems linked to the large-scale modernisation of social housing, as well as the economic and functional issues that led to the construction of housing adapted to the changing needs of working families. At the present time, this estate is the subject of research into policies and strategies on full-scale energy retrofitting interventions based on standard exterior insulation solutions, which risk transforming the intrinsic architectural qualities of the existing buildings which define the current aesthetic of the neighbourhood. This is why, for this and many similar concerns observed on other sites, we considered the Vielle Motte at Neuville-en-Ferrain as a prototype allowing us to roll out the approach of in-depth analysis and awareness raising proposed by the research more extensively.

40Following a preliminary study on the composition of households and their energy consumption, our detailed analysis centred on households predominantly made up of retired people living alone or as a couple, in all three types of households. In the range offered by the estate management, we chose to study the following housing types: a T2 facing due South 49 m² (1), a T4 facing Northwest 98 m² (2) and a T6 facing Northeast 143 m² (3) (fig. 7).

41To assess the pertinence of the research from the point of view of raising awareness about energy and to validate the usefulness of the research, the initial simulations which resulted in the detailed knowledge phase allowed us to clarify the strategic challenges of the characterisation method employed for the construction and study of residents’ behavioural models. These energy simulations went as far as to expose the major disparity, in terms of a passive attitude towards energy, in the various situations described above, which had to take three comparisons of the initial results into consideration. In fact, many of the housing units (four & six room flats) are north facing, most of them are under-used and many of them have garages with very low levels of natural light, a factor which could have a major impact on the analysis of energy performance balances.

Categorisation of typical resident profiles and their impact on energy performance

42Our aim was to identify the possible crossover of resident profiles examined for this study and interpret the results primarily assessing occupancy balances in the range of household types on the Neuville-en-Ferrain site.

43The final categorisation of the typical household profiles was carried out based on the answers to the questionnaire which was widely distributed amongst the residents. The data collected were combined with the data which arose from the observation of residents’ behaviour. Our objective was to establish several levels of interpretation of the usage and energy consumption. Consequently, we could also take the occupancy variables and different usage habits of household appliances into consideration, as well as the various management strategies of thermal comfort and energy consumption data in our calculations of energy needs and gains. In the following part of our study, to illustrate our approach and compare our DES analysis results, we chose to categorise three variable, heterogeneous family situations which present similar characteristics in terms of the composition of the household to those homes investigated in the preliminary study. We selected one of each inhabitant profile as follows: profile A (one person/T2), profile B (couple/T4) and profile C (couple with one adult child/T6), the three households were used as case studies to construct the typical resident profiles and for the results in the awareness raising phase (fig. 8, 9, 10).

44The DES results demonstrated that by considering the activity and lifestyles in these households we could obtain a more precise interpretation of the variety of behaviour and the shift in energy consumption. The findings also revealed the varying disparity between genuine energy data obtained from residents. Although the annual energy consumption balances calculated using the simulation methods during the study are approximative, they nevertheless allowed us to discern certain differences in the results on a case-by-case basis, in comparison to conventional ways of calculating estimated costs. In fact, most of the occupancy profiles selected were above or below the basic averages of the standard energy balances carried out in similar households (fig. 11).

45We also observed in these three typologies occupancy rates that were on average higher in the ground floor spaces, which had larger rooms with better natural light and access to the outdoors. In these different case studies, the bigger the housing unit, the greater the gap between occupied and unoccupied living spaces and this was palpable via the interpretation of the energy gains and consumption balance. The nature of the households, as well as their occupancy particularities and human activity, showed that variations in the typology models are certainly present and they directly affect the sanitary condition of the buildings and the performance of the household energy systems, as well as the typologies reserved for living spaces. We also noticed that the typological distribution patterns depend on these numerous parameters that are specifically linked to the effects on energy of residents’ behaviour and the varying intensity of occupancy rhythms.

46The collected and simulated data served as the foundation for the characterisation and definition of socio-behavioural strategies, based on energy use recommendation scenarios suggested to the residents with the aim of assessing the impact of a change in their daily energy use behaviour, changes to their heating and lifestyle. In terms of the awareness raising phase, the scenarios suggested had a series of implications and variables impacting on behaviour, personalised on a case-by-case basis, which could result in a uniform energy intervention strategy. Their recommendations varied essentially according to the daily routines of each resident, the layout of the rooms in their homes, their closeness to one another and their exposure in relation to the plans (fig.12).

Energy Recommendation A. The temporary/ permanent increase of the number of residents living in the housing unit

47Destined to be applied to profiles B and C, this energy use recommendation suggested residents vary their occupancy balance and how frequently they used the rooms depending on the number of people present. For them, this was about a new timetable of occupancy which suggested more regular presence in the heart of the home, either through permanent occupancy of the spaces occupied or unoccupied, or by occasionally welcoming other people into the household (house-share or rentals which modified the use of the home, or the make-up of the household in a more appropriate typology) (fig. 13, 14).

Energy Recommendation B. The strategic planning of energy usage and the consumption relative to occupancy and activity

48Destined to be applied to profiles A & C, this energy use recommendation suggested residents plan their usage to suit their need for energy consumption, in relation to their behaviour. For them, it meant modifying their daily habits, reassessing the usage most likely to influence internal gains, or better ways to distribute the household appliances around the home (fig. 15).

49This recommendation, just as the first, falls into the category of socio-behavioural measures which have a minimal impact on the building.

Energy Recommendation C. Spatial clustering of increases in internal gains

50Destined to be applied to profile B, this recommendation suggested adapting the residents’ energy use to get the most out of the passive energy particularities of the existing building. Depending on the typological layout, the exposure of the windows and doors and their proximity to the most occupied rooms, it meant allowing the residents to make the most of their energy behaviour by evaluating the energy-efficiency of spatial clustering of the increases in internal gains due to occupancy and household appliance use (fig. 16). This recommendation, that could be defined as “socio-technical”, allows architectural qualities of the building to be conserved by avoiding heavy duty renovations (ETI or ITI).

Awareness of the energy effects of modifying residents’ behaviour

51Taking the Renovation 2.0 approach into account targeted in the study and the observations raised by the DES, it seemed like an interesting idea to develop a tool capable of assessing the implementation of different scenarios for energy recommendation guidelines. These guidelines have a simple objective, to raise the residents’ awareness of their occupancy balance and energy consumption and to allow them to assess themselves and evaluate their habits and adapt their behaviour in line with the recommended energy parameters for their household.

52The design and distribution of a questionnaire on raising awareness, based on the results of the scenario phase, constitutes the final accomplishment of the methodology of the research. It aims to define the strategic guidance to assist architects and estate managers in their selection of profiles for renovation 2.0. which in turn provides a better follow-up over long-term scenarios for energy recommendation guidelines developed during the study phase of this research. Based on the principle that the energy renovation 2.0 of an estate also depends on levels of observation, control and interaction with residents, we attempted to identify, amongst the selected profiles for the study, behaviours that were receptive to these three ideas which could influence the increase of internal gains (due to occupancy and the use of appliances) and consequently a reduction in energy consumption. Based on the findings, the answers were recorded in the form of tables, then scored and divided into thematic sections, in line with the research methodology

53The “Assessment of satisfaction levels” section allowed us to obtain ⅔ of the answers and opinions of the residents involved in the case studies. The ‘self-assessment’ section, on the other hand, generated 20 answers out of all 94 housing units on the Vieille-Motte site. Amongst the households surveyed were notably six T2 flats, seven T4 flats and seven T6 flats. Two T2 flats presented situations which were similar to profile A (exposure of the housing unit, number of occupants and types of activity). One of the T4 flats is also very close to the situation studied for profile B. For profile C, we identified three types of T6 flats which were similar in terms of exposure (North/ Northwest) and the number of occupants.

Results and discussion

54Answers in the “Self-assessment” section were analysed according to a range inspired by the Lickert scale. We were convinced that this psychometric scale was a good fit for our interpretative approach to our results. To implement it we determined six scores, numbered 1-6, which seemed an efficient way to facilitate the quantitative treatment of the statistical data collected.

55Profile A reacted to the suggested recommendations, based on the in-depth analysis and the preferred hypothetical scenario. It scored 5/6 relative to the assessment of recommendation B. In addition, it scored 4/6 on the feasibility of the implementation of measures outlined in the scenario. However, when it was necessary to express their desire to continue discussions and experiences related to the scenario, profile A scored their motivation 2/6. They indicated that in the raising awareness questionnaire, the observation, control and interaction levels were lower compared to averages in similar typologies (fig. 16). The difference between the observation levels is most striking in profile A (2,9/6) compared to the average levels in similar typologies (3,47/6). This demonstrates a lack of concern regarding the impact of energy variations due to behaviour. The interaction score of profile A (3,2/6), indicated, however, an adequate level of control of their thermal comfort (heating).

56Although it was not stated clearly in the assessment of recommendation C, together with the feasibility of the implementation of measures suggested by this scenario, profile B scored interaction and control levels far higher than the averages recorded by the other T4 flats (fig. 17).

57Only the observation level scores were lower in comparison to other typologies — 2,81/6 for profile B and 3,31/6 for similar typologies. This score represents the residents’ lack of concern with regards to the impact of behaviour in the household on energy consumption. The levels of interaction (4,4/6) and control (3,97/6) do, however, confirm the sensitivity and attention paid by profile B to the energy variation parameters and their influence on adaptable thermal comfort.

58The observation, interaction and control levels of profile C could not be established, as so few residents participated in the questionnaire phase. This prevented us from assessing recommendations A and B. However, other residents of similar typologies did participate willingly and engaged more with the study.

59The results of the awareness raising questionnaire show clear disparities between the results of residents of T6 flats and those living in T2 and T4 flats. 39,5% of the occupied T6 flats, 71% of those interviewed for the awareness-raising questionnaire live in a couple. The low occupancy rates of the rooms in the T6 flats can be explained by the findings observed (3,02/6) and underline the difficulties in grasping how to impact energy benefits by changing your behaviour. Similarly, the interaction levels measured (2,6/6) reveal the increasingly complex management of the comfort control measures. Indeed, in the less occupied T6 flats, the increase in internal energy gains was limited. The management of the atmosphere and the level of comfort rely moreover on the technical systems at the residents’ disposal. However, it turns out that according to the results of the DES and the technical systems at residents’ disposal and the residents’ answers to the questionnaires, they distinguished effective ways of reducing energy and restoring the losses and maintaining a level of comfort and lower energy consumption. The overall scores (3) confirm challenges shared by most of the residents surveyed, while recognising the passive energy of their homes and the impact on energy due to their behaviour. On the other hand, what is crucial is the comparison of the residents’ behaviour to each other -interaction - where the average scores for self-assessment were much more varied and raised further questions on the levels of acceptability, adaptability and application of the various methodological approaches of the research. In fact, the exploitation of the outcomes of the questionnaire presents a major asset to the self-determination of the targeted profiles of Renovation 2.0, based on the established averages. The case-by-case synthesis of the various analysis criteria (observations, interactions, controls) give us a greater understanding of the parameters that influence the covariation of the behavioural impact on energy (analysed in the DES) together with residents’ awareness levels. In the case of the site at Neuville en Ferrain, we identified several answer categories which were dependent not only on typo-morphological factors of the housing units, but also on the relation between residents and the socio-morphological composition of their households (fig. 18, 19).

60When we combine the results of the in-depth analysis of the three case studies of typical profiles in the experimentational phase, we noticed a correlation between residents’ energy behaviour and their ability to influence the feasibility, and even the cost-efficiency of the Renovation 2.0 approach. Except for profile C which did not want to develop this phase of the study, it seemed that the awareness level of profile B corresponded to a category of typical resident profiles which were attentive to the effects of energy variations due to behaviour. This was not the case with profile A who showed a level of awareness closer to the averages of all the typologies, which could correspond to the categories of typical profiles who were insensitive to the effects of energy variations due to behaviour.

61This observation justifies the methodological response of the research in terms of the need to interpret the DES results based on the analysis of energy awareness levels of certain residents. Those who were the most responsive to the study’s recommendations and/ or occupied the largest types of housing units, were representative of the typical profiles which had the greatest facility to adhere to the Renovation 2.0 energy approach. We also noted that the scenarios suggested for optimising residents’ energy behaviour in the heart of existing housing types (Recommendation B), could serve as a first step to becoming involved, according to the nature of the profiles identified. Secondly, the scenarios which proposed the densification of energy behaviours in existing typologies (Recommendations A and C), will only be envisaged on the condition that these profiles already have a home capable of accommodating the recommended changes.

Conclusions

62As a result of these experiments, we have been able to demonstrate that the use of occupancy and activity planning in households can have significant benefits in terms of reducing energy consumption needs and improving adaptive thermal comfort. We also demonstrated that “Renovation 2.0”, with its participatory, collaborative approach, represents a major factor in facilitating decision making and offers several small-scale, non-standard potentially highly adaptable solutions, depending on the typologies and/or targeted resident profiles. In the framework of our case study, the quantitative results of the households investigated via the DES, brought us to the conclusion that the household had huge potential for energy improvements with minimal interventions on the building, notably the T6 flats. Due to their spatial configuration and low levels of occupancy per metre squared, these household could carry out low risk, low-cost intervention measures and taking on socio-technical changes, all of which brought about considerable energy performance improvements (partitioning of less occupied rooms, changes to the typological allocation, increases in the number of occupants, repositioning of energy use and appliances, etc.). The categorisation of resident profiles through questionnaires’ proposed methodology also allows us to better target households with a higher potential for the deployment of socio-behavioural measures. Indeed, the profiles which attained self-assessment scores of an average higher than 4, showing greater sensitivity to energy issues, could be adapted more easily to the implementation of “eco-responsible” behavioural changes.

63Finally, the assessment of the possibilities for changes in behaviour, based on digital simulations carried out with the BIM maquette, allowed us to go back and forth to implement energy improvements qualified as ‘step-by-step’. Feedback on the steps to be carried out, allowed us to test solutions under a range of different conditions and to check every minute detail to clarify their inherent risks and advantages.