[Paper Reading] A literature review on driving factors and contextual events influencing occupants' behaviours in buildings

Publication year: 2017
Authors: Francesca Stazi, Federica Naspi, Marco D’Orazio

Introduction

• the consumptions difference among buildings with same physical features is mainly related to occupancy patterns, users’ lifestyle, comfort preferences and interaction with building systems
• Researchers usually found a gap comparing simulated and real energy consumptions and they assessed that occupants’ behaviours is one of the main reason that causes this difference
• including the stochastic aspect of occupants’ behaviours into building energy performance simulation software, would lead to better predictions of the buildings’ final energy consumptions, especially for low energy ones
• Beside the environmental component, many other factors have been identified as influencing (contextual, physiological and psychological), but they are often reported as fragmented and rough information, because of the difficulty in investigating, quantifying and correlating them
• Moreover, researchers studying the same topic reach different and sometimes incomparable results mainly because each building is a peculiar case study, characterized by typical features (e.g. orientation, construction type and materials) that make it unique. Each survey is referred to different parameters, sample dimension and monitoring duration
• The review includes over one hundred studies (both original papers and previous reviews) from 1978 to 2017, which concern the human energy-behaviours inside buildings. The surveys are mainly related to offices and residential buildings, while other building uses (e.g. schools) are less investigated

  Occupancy Models

• Occupants have a great impact on building energy consumptions, since people presence affects the use of building systems and plug loads and influence heating, ventilation and lighting requirements. Moreover, occupancy is an essential precondition to assess users’ interaction with building devices.
• One of the most used are passive infrared sensors (PIR); However, such technology cannot detect immobile occupants (e.g. people working at desk in offices) and it is not able to decode the number of people

  Window Use

• Windows and environmental drivers
  • Temperature and indoor air quality requirements are the main environmental stimuli that trigger window adjustments
  • Outdoor temperature
  • Indoor temperature
  • CO2 concentration
• Windows and time-related drivers
  • The interaction between users and windows in buildings is largerly driven by time-related events, which depends from the building usage typology
• Windows and models
  • Several behavioural models have been developed and integrated in simulation programs with the aim to describe windows use
  • Stochastic models seem to perform better than rule-based ones during free-running season, both in terms of window-user interaction and thermal comfort evaluations
  • the authors underlined that behavioural models application should be supported by empirical observations and rigorous investigations
  • In relation to residential buildings, very few models are present in the literature. For this building use the most used parameters are CO2 concentration, indoor and outdoor temperature

Light Switching

• Light and environmental drivers
  • The users’ “laziness” regarding the interaction with some systems appears for turn off activity. In fact, once the lights were on people tend to leave their status unchanged until they leave the room. Moreover, sometimes occupants do not notice that the lights are turned on, so they do not switch them off when indoor daylight illuminance levels are high.
  • Work plane illuminance
  • Illuminance
• Light and time-related drivers
  • Researchers evaluated that switch-on events at arrival occurred between 77% and 95% of times
  • Turn-on probability at arrival is strongly related to the minimum work plane illuminance, in particular lower than 100 lux
  • The majority of departures were marked by turn-off events
  • Many studies assessed that arrivals and departures are the main drivers for turning lights on and off respectively, nevertheless their frequency is influenced by illuminance levels
• Light and models
  • The model worked simulating also the blinds’ adjustments, which depended on the type of blind (i.e. manually or automated), environmental parameters (i.e. direct sunlight) and contextual parameter (i.e. first arrival)

Shading and Blind Use

• Their use is not only linked to visual and thermal comfort, but also with necessity of view and privacy. Moreover, recent studies underlined that blind use is one of the most influencing factors on building energy consumptions
• Other great differences between the case studies concerned blinds’ position (i.e. inside, outside or within the envelope), types (e.g. roller shades, louvers) and shapes (e.g. venetian and vertical blinds).
• Shading, blind and environmental drivers
  • the frequency of adjustments: blinds position remains usually unchanged for weeks and months. The interaction with blinds is less frequent compared with windows
  • Also the position is generally standard, keeping them fully open or fully lowered. Once the blind position is set, users wait until a critical situation occurs to change them
  • A general result on trigger parameters for the blind usage is still missing because of the wide variation of the recorded and correlated factors among the studies
  • blind use is not only dependent on climate but also on contingent features, as building exposure and orientation, shadings’ typology and desks’ position.
  • Illuminance
  • Solar radiation
  • Glare
  • Indoor and outdoor temperature
• Shading, blind and time-related drivers
  • The interaction with blinds is little time and seasonal dependent
  • These opposite results suggest that actions on shadings and blinds are partially related to time drivers but they are highly influenced by occupants’ habits
• Shading, blind and models
  • indoor horizontal illuminance, outdoor illuminance and shaded fraction before action
  • different daylight glare index (DGI), view angles, vertical solar radiation

Air Conditioning Use

• AC and environmental drivers
  • Indoor temperature
  • Outdoor temperature
• AC and time-related drivers
  • The main differences both for switch on and off patterns were between weekday and weekend
  • during weekdays switch-on peaks occur around 6 p.m. and the off-events usually happen from 6 to 12 p.m., while adjustments in weekends are operated along the all day
• AC and models
  • both with environmental (i.e. outdoor temperature) and individual factors (i.e. users’ preferences and background)

Thermostat Use

• both indoor and outdoor temperature but also solar radiation, ownership conditions and perception of indoor environmental variables
• Outdoor temperature
• Indoor temperature
• some authors assessed that the major differences in control patterns were correlated habits, routines and time of the day

Fans and Doors

• these controls have a little direct influence on building energy consumptions, but their use is significant for users’ thermal perception
• Very few studies (all performed in office buildings) analysed fans and doors usage in relation to environmental parameters
• Fans, doors and environmental drivers
  • Indoor temperature
  • Outdoor temperature

Discussion

• This gap is mainly caused by occupants’ interaction with building systems: the way users behave is very different from the modelled one
• Summary on actions drivers
  • Windows use is clearly driven by indoor and outdoor temperature, while CO2 concentration seems to have influence only in residential buildings.
  • Light-switching behaviours are linked to illuminace and they are often in tight connection with blinds
  • blinds and shadings evidence that illuminance, solar radiation and glare are all identified as important factors
  • air-conditioning units is found to be consequence of both indoor and outdoor temperature.
  • thermostat use is influenced by temperature, especially by the outdoor one
  • fans and doors: the use of fan is related both to indoor and outdoor temperature, while, as expected, doors are operated mainly in relation to indoor temperature
  • windows in offices is greater at arrival, decreases substantially in intermediate periods and is related to closings before departure
  • Light switching activity is the pattern most driven by time-related events: in offices the greatest part of the turn-on events occurs at arrival or entering the room after a period of absence, while almost all the switch-off events are related to departures. Blind and shading use is more seasonal than day dependent: the daily adjustments are very rare and the position is kept constant until a discomfort situation occurs.
  • air-conditioning units, the interaction with this control is also a consequence of the different life style between singles and families, of the day of the week and of particular events (i.e. sleeping and cooking)
  • Window adjustment is a more complex and variable pattern because, in addition to environmental variables, also properties and geometry of the windows and personal preferences of the users (i.e. privacy) play a key role in windows’ adjustments
  • fans and doors were less or not at all investigated
  • The discussion suggests that time of the day, occupancy patterns, daily routine and users’ habits drive occupants’ behaviours as well as environmental variables and that their prediction can be more uncertain, since some actions are more variable
• Impact of users’ actions on energy consumptions
  • The actions were ranked from the most to the lowest incisive: windows, metabolic rate, clothing level, blinds and fan/heaters
  • Fans and clothing levels do not affect the building performance directly
• Behavioural model development: recent trends and open issues
  • Window adjustments, artificial lighting behaviours, solar shade controls, AC settings, etc. are not deterministic patterns. The people stochastic behaviour was recognized and many models tried to express it in analytical form, with the aim of software implementation
• Modelling human behaviours
  • One of the most difficult aspects in modelling users’ behaviours is to reproduce many actions and catch the mutual influences between different users
  • The ABM approach represents occupants as “agents” that interact with each other and with the environment, following pre-determined rules
  • However, achieving a unique result is still an open question since the comfort levels can be evaluated according to different theories: the PMV comfort model [138], the Belief-Desire-Intention model (BDI) and the Perceptual Control Theory
  • Empirical ABMs (i.e. based on real data) provide better performances compared to simple abstract ones, however they often are more complex and they need big data and very precise functions
  • Another type of approach, to consider the occupants dynamic behaviour, is to propose a generic model that can include many different patterns.
• Implementing behavioural models in simulators
  • An additional step is the possibility to modify or overwrite the source code and to implement a specific behavioural model
  • According to some authors, co-simulation seems to be one of the most promising approaches. Even if the implementation is not very easy, co-simulation allows the simultaneously run of behavioural model and environmental simulation
• Lacks and future challenges
  • The program IEA EBC Annex 66 (Definition and Simulation of Occupant Behaviour in Buildings) [20], focusing on this issue, aims at standardizing descriptions and classifications of occupants’ behaviours
  • Following this framework, it has been also provided a questionnaire survey focused on collective and social aspects among different climates, cultures and norms
  • some authors [16,151] directed their attention at assessing how much people’s lifestyle (e.g. energy conscious, habit related and high quality) affects the building performance and the energy use
  • In fact, simulation studies [154] found energy savings between 22.9% and 41% when applying occupant behaviour measures. Switching off lights, plug loads and HVAC systems in unoccupied spaces are identified as the main behaviours to avoid energy waste

Conclusions

• This article reviewed the occupants’ behaviours inside buildings assessing the main driving factors for their actions, divided into environmental and time-related stimuli. The studied devices are: windows, lights, blinds and shadings, AC units, thermostat, fans and doors
  • Window use is driven by indoor and outdoor temperature both in offices and residential buildings, while CO2 concentration is a stimulus only for the latter
  • Light switching behaviour in offices is mainly linked to timerelated events: arrivals and departures
  • Shadings and blinds are rarely adjusted and almost always in order to reach visual comfort
  • Air-conditionings, thermostats, fans and doors are triggered by indoor and outdoor temperatures
  • Behavioural models are usually focused on a single action, triggered by one or more environmental variables. However, recent studies are supporting approaches to consider many behaviours, the interaction between users and different lifestyles
  • The connection between behavioural models and simulation software is shifting towards a co-simulation approach, in order to obtain more realistic results