Occupant satisfaction in mixed-mode buildings

‘Mixed-mode’ refers to a hybrid approach to space conditioning that uses a combination of natural ventilation and some form of mechanical ventilation and/or cooling. This study focuses on mixed-mode in buildings with operable windows (as opposed to natural ventilation through vents). By utilizing active cooling only when and where it is necessary, a well-designed mixed-mode building offers the potential to improve the indoor environmental quality while minimizing the significant energy and operating costs of air-conditioning. However, limited information exists on the performance of mixed-mode buildings, particularly with regard to occupant satisfaction, and evidence can be a powerful part of the argument to minimize the use of air-conditioning. The results of web-based surveys conducted in twelve mixed-mode buildings are described and compared with a benchmarking database of 370 buildings, with over 43 000 individual responses. The survey focuses on seven areas of indoor environmental performance: thermal comfort, air quality, acoustics, lighting, cleanliness, spatial layout, and office furnishings. The mixed-mode buildings are performing exceptionally well compared with the overall building stock, especially with regard to thermal comfort and air quality. The best performers were those that were newer, in more moderate climates, had radiant cooling or mechanical ventilation only, and allowed high degrees of direct user control without changeover window interlock systems. Le ‘mode mixte’ fait référence à une approche hybride de la climatisation des espaces qui utilise une combinaison de ventilation naturelle et d'une certaine forme de ventilation et/ou de refroidissement mécanique. Cette étude est centrée sur le mode mixte dans des immeubles équipés de fenêtres manoeuvrables (par opposition à une ventilation naturelle par des orifices de ventilation). En utilisant un refroidissement actif uniquement quand et là où c'est nécessaire, un bâtiment équipé d'un mode mixte bien conçu offre la possibilité d'améliorer la qualité environnementale intérieure tout en réduisant au minimum les coûts énergétiques et opérationnels importants de la climatisation. Néanmoins, les informations existantes sur le rendement des bâtiments à mode mixte sont limitées, s'agissant en particulier de la satisfaction des occupants, et des données probantes peuvent peser fortement en faveur d'une réduction au minimum de l'utilisation de la climatisation. Les résultats des sondages en ligne réalisés dans douze bâtiments à mode mixte sont décrits et comparés à une base de données de référence de 370 bâtiments, comportant plus de 43000 réponses individuelles. L'enquête se concentre sur sept aspects de la performance environnementale intérieure: confort thermique, qualité de l'air, acoustique, éclairage, propreté, agencement des espaces et ameublement des bureaux. Les bâtiments à mode mixte se comportent remarquablement bien par comparaison avec l'ensemble du parc bâti, tout particulièrement en ce qui concerne le confort thermique et la qualité de l'air. Les plus performants étaient ceux qui étaient les plus récents, avaient les conditions climatiques les plus tempérées, avaient uniquement un système de refroidissement à panneaux rayonnants ou une ventilation mécanique, et autorisaient des niveaux élevés de contrôle direct par l'utilisateur sans systèmes de verrouillage interrompant la ventilation mécanique lors de l'ouverture des fenêtres. Mots clés: rendement énergétique des bâtiments, refroidissement, qualité environnementale intérieure, ventilation naturelle, satisfaction des occupants, évaluation après occupation, confort thermique, fenêtres


Introduction
In current commercial buildings in the United States, cooling and mechanical ventilation account for over 30% of total energy use, for approximately 20% of electricity use, and for approximately 40% of peak demand. However, before the 1950s, air-conditioning and mechanical ventilation were not yet commercially viable, and so commercial buildings had little choice but to utilize natural ventilation for cooling. Buildings typically had extended perimeter zones so that every office could have access to windows that would open to the outdoors and provide the primary source of light and fresh air. But the availability in the 1950s of large-scale mechanical ventilation and cooling, along with other technologies such as curtain walls and fluorescent lighting (as well as market pressures to maximize floor areas and flexibility of interior space), led to the more common commercial building forms of today that are typically all-glass, flush-skin buildings with large floor plates and no operable windows. These buildings miss out on the large number of documented benefits of operable windows -thermal comfort over a wider range of temperatures based on the adaptive comfort zone (Humphreys, 1975;de Dear and Brager, 1998), reduced energy consumption compared with conventional air-conditioned buildings (Emmerich and Crum, 2005), and fewer Sick Building Syndrome symptoms (Seppänen and Fisk, 2002).
But even with all these potential benefits, there are a variety of concerns and design challenges associated with operable windows. The ability to rely solely on natural ventilative cooling is limited by loads and climate. And given modern day expectations, engineers are often uneasy about the lack of predictability and control over indoor thermal conditions in naturally ventilated buildings. As a result, many innovative engineers are exploring 'mixed-mode' buildings -a way to combine the best features of naturally ventilated and air-conditioned buildings, and essentially extend the range of climates in which operable windows are feasible even when they cannot provide acceptable comfort year round.
'Mixed-mode' refers to a hybrid approach to space conditioning that uses a combination of natural ventilation from vents or operable windows (either manually or automatically controlled) and mechanical systems that may provide ventilation, interior air distribution and/or some form of cooling. For present purposes, a focus is made on buildings with operable windows, and also on systems with active cooling. With this in mind, a well-designed mixed-mode building would ideally allow spaces to be naturally ventilated during periods of the day or year when it is feasible or desirable, and would use mechanical ventilation or cooling only as necessary for supplemental cooling when natural ventilation is not sufficient. The goal is to maximize comfort while minimizing the significant energy use and operating costs of mechanical ventilation and/or cooling.
While mixed-mode buildings are much more common in Europe, it is a relatively newer concept for American engineers. The US building design industry is generally unfamiliar with mixed-mode cooling strategies, and there is a lack of published case studies or design and analysis tools to facilitate their ability to chart new territory. To address this need, the Center for the Built Environment (CBE) developed a web-based library of mixed-mode building case studies, covering a range of climates, design approaches, and control strategies (CBE, 2006). The library offers two levels of information: (1) a database with a broad list of buildings and basic project information; and (2) more detailed case studies. The database includes approximately 150 mixed-mode buildings, with over 60 of them in North America. It is downloadable as an Excel spreadsheet to allow for easy sorting, and includes basic information about each project including location, year built, type of building, owner, architect, engineer, brief comments about the mechanical system, operable windows, control and operation strategies, and web links for more information. The eight case studies provide more detailed narrative and graphic descriptions obtained from literature reviews, drawings and photographs, and interviews with building owners, architects, engineers, and facility managers. The case studies include information about the windows, heating, mechanical ventilation and air-conditioning (HVAC) system, control strategies, building design process (design tools used, commissioning, relevant code issues), cost (where available), and additional green features of the building. The Resources section of the website also includes a more report with 23 new case studies that focus on control algorithms (Brager et al., 2007).
What motivates building owners and the design team to move beyond conventional air-conditioning and design a mixed-mode building? Without question, it is absolutely crucial to reduce energy consumption in buildings, and help avoid the potentially devastating impacts of climate change. But in terms of the building owner's financial budget, energy costs are still relatively small compared with worker salaries, which represent over 90% of the annual operating costs per square foot of a commercial building (Kats et al., 2003). In addition, the cost of worker recruitment and retention is significant (Institute for a Competitive Workforce (ICW), 2001). Therefore, from the building or company owner's point of view, perhaps the most persuasive argument for sustainable design in general -and operable windows in particular, where applicable -is one that makes the connection between a higher quality indoor environment and increased comfort, health and productivity of the workers. There is clear evidence that the health and productivity of occupants is positively correlated with comfort and satisfaction (Leaman and Bordass, 2001). Therefore, if it can be demonstrated that occupant satisfaction can be higher in buildings with operable windows, then that can be a powerful part of the argument to avoid or minimize the use of airconditioning or other forms of centrally controlled mechanical cooling.
So, how does one learn about the quality of the indoor environment? Sadly, very few architects or other members of the design team are likely to know how well their building is working after it is completed and occupied, the fees have been paid, and they are onto another project. Without learning from experience in an objective way, building industry professionals are less likely to make design or economic decisions that will truly enhance the performance and experiential quality of their buildings. Physical measurements can be valuable, but by themselves they also need to be interpreted in terms of how they impact the occupants. Buildings occupants themselves are a rich, yet under-utilized, source of direct information about how well a building is working, but the challenge is how to collect both the positive and the negative subjective feedback systematically and objectively (Vischer, 2008). Detailed thermal comfort field studies that include both physical measurements and subjective surveys are the most revealing, but are also time-consuming and expensive, and therefore the number of buildings that can be investigated is inherently more limiting.
Web-based surveys are an effective way to study building performance from the occupants' point of view. They can be used as a diagnostic tool to help designers, building owners and operators, and tenants evaluate how well their office buildings are working from the occupants' perspective, and to help prioritize investments to improve performance. The surveys can also be used as a research tool for specific projects requiring the assessment of occupant response, or for broader benchmarking and comparative analysis of the performance of particular building design, technologies, and operation strategies. It was with these dual purposes in mind that the CBE developed their survey.

Center for the Built Environment (CBE) Survey
In 2000, the CBE began developing a web-based indoor environmental quality (IEQ) survey and accompanying online reporting tools. Advantages of the webbased format are (1) it is quick and inexpensive to use; (2) it allows for branching questions to get more detailed information where appropriate (in particular, when the occupant indicates dissatisfaction with a certain area), thus avoiding making the survey too long for everyone with overly detailed or inappropriate question; and (3) survey results can be accessed using an automated, advanced reporting tool that allows users to filter, aggregate, compare, or benchmark their data.
In addition to basic questions about demographics and workspace descriptions, the core CBE survey measures occupant satisfaction and self-reported productivity related to nine environmental categories: office layout, office furnishings, thermal comfort, air quality, lighting, acoustics, cleanliness and maintenance, overall satisfaction with the building, and overall satisfaction with the workspace. Satisfaction questions use a consistent seven-point scale ranging from 'very satisfied' (coded as 3) to 'very dissatisfied' ( -3), with a neutral midpoint (zero). A secure standardized query language (SQL) server database is used for collecting and recording the responses. It takes approximately 5-12 minutes to complete the survey, depending on the number of branching questions one receives, and the number of open-ended comments in which one writes. Additional, custom survey modules can be added that gather data about a variety of supplemental topics, depending on available building features or the client's particular interests. Examples include modules on specific issues such as daylighting, radiant cooling, and accessibility, and modules for specific building types such as healthcare facilities, K-12 schools, and dormitories. In addition to the occupant survey, a representative of the building owner or design team fills out a building information form to provide descriptive information about the building and its systems, such as the age of the building, the number of occupants, the type of HVAC systems, and whether the windows are sealed or operable.
CBE also developed an automated web-based reporting tool that researchers and clients can use starting approximately one week after the survey is completed, allowing time to create a final data set where responses of participants who answer fewer than 15 questions are removed. The reporting tool allows one to produce standardized summaries of the responses in a particular building, compare them with the overall benchmarking database, or do more in-depth data mining to compare responses from selected subgroups of people or explore relationships between questions.
The CBE Survey benchmarking database represents the portion of buildings surveyed by the authors that meet certain quality control criteria, such as the number of responses or per cent response rate. At the time of analysis, the CBE Survey benchmarking database included over 370 buildings, with over 43 000 individual responses, and 3.8 million data points.
For more information about the CBE Survey, see Zagreus et al. (2004). One previous study focused on Occupant satisfaction in mixed-mode buildings comparing the performance of green and LEED 1 buildings with the overall database (Abbaszadeh et al., 2004), where it was found that there was not necessarily a correlation between buildings with a large number of LEED IEQ points and IEQ performance from the occupants' perspective. Another focused on the role of air movement and personal control in influencing thermal comfort and perceived air quality for the database overall (Huizenga et al., 2006). That study found that satisfaction with both thermal comfort and air quality increases significantly in buildings that provide people with some means of personal control over their environment, such as thermostats or operable windows.

Mixed-mode buildings
The purpose of this analysis was to examine occupant satisfaction in mixed-mode buildings, with the aim of comparing patterns with those found in the other buildings in the database (which are primarily sealed with mechanical cooling). The twelve mixed-mode buildings analysed for this study were identified from the CBE Survey database. A representative of the building fills out a 'building characteristics' form, which helps one identify basic descriptive information about the building. Unfortunately, this form is not always filled out fully or consistently. Therefore, while there are possibly additional buildings in the CBE database that may have operable windows, only included in the study are those for which there was sufficient information about the building from the characteristics form, and where one could find other available case study material confirming that they all were mixedmode buildings. For similar reasons, the authors had difficulty identifying buildings in the database that rely on natural ventilation exclusively, and so were unable to compare their performance. Finally, to avoid the potential bias associated with low response rates (which are more common in internet surveys), only surveys were included that had a response rate of over 50% for buildings with fewer than 50 occupants, or a response rate of over 25% for buildings with 50 or more occupants. Table 1 summarizes some basic characteristics of the twelve mixed-mode buildings compared with the other 358 buildings in the database, as well as the total number and rate of responses in each group. Overall, the mixed-mode buildings are relatively newer and smaller, but not necessarily less dense. The buildings were more likely to incorporate other green building features (75% were LEED-certified compared with only 12% of the general building stock), including innovative mechanical cooling systems such as underfloor air distribution. Table 2 identifies the location and general control scheme for the twelve mixed-mode buildings. The group of mixed-mode buildings analysed in the study represents a broad range of climates, building types, sizes, and uses. They range in size from 1100 to over 14 000 m 2 . The buildings also ranged in the number of occupants; the buildings often had a significant transient occupancy, especially in educational buildings. However, the survey was only offered to employees and so the occupancy numbers in Table 2 reflect the number of employees rather than all occupants of the building.
The buildings also represent a variety of different organizational or control strategies, including zoned systems (where the natural ventilation and mechanical cooling essentially occurs in different areas), changeover systems (where the mechanical cooling is shut off when the windows are open), concurrent systems (where the windows and mechanical cooling can be operated simultaneously), and 'red light/green light' systems (what is called here 'informational controls', where indicator lights controlled by temperature and humidity sensors tell occupants when they can open windows). In many cases, buildings can combine more than one of these operational strategies.
Three of the twelve mixed-mode buildings have airconditioning only in very limited areas (that is, zoned strategy). Instead, they use natural ventilation in the primary occupied spaces, and in two of the cases mechanical fans bring unconditioned outdoor air into the rooms directly. In these three buildings, which are all educational spaces, air-conditioning is only provided in spaces that have a programmatic need for more cooling (for example, laboratories, large assembly rooms, etc.).

Results
Overview of indoor environmental quality (IEQ) scores Figure 1 shows a summary of the average (mean) scores in each IEQ category for the twelve mixed-mode buildings (a total of 520 individuals) and the full survey database (a total of 42 700 individuals). One thing commonly seen is that people may give a building low scores in individual categories, but will tend to give better marks for general satisfaction with the building or workspace. Across the entire database, thermal comfort, air quality, and acoustic quality received the lowest scores. These are also areas that can be potentially and significantly impacted by operable windows and mixed-mode systems, and so the paper will focus on these areas.
On average, mixed-mode buildings perform significantly better than the remaining (primarily air-conditioned) buildings in the database in nearly every category except lighting and office layout (where performance is close to being equal), and acoustics (where performance is only marginally better). The improvement in office furnishings is most likely attributable to the mixed-mode buildings being newer.
The biggest IEQ improvements in mixed-mode buildings compared with sealed, air-conditioned buildings were for thermal comfort and air quality. Even the slight improvement in acoustics is surprising given that it might be anticipated that open-plan offices that facilitate natural ventilation often contribute to poorer acoustic environments, and outside noises are often perceived to be a barrier for operable windows -but evidence suggests that these were not problems for these particular sites. The better acoustics performance may also be attributable to the lower level of occupancy of the mixed-mode buildings compared with the database. Also surprising was the higher rating for cleanliness and maintenance in mixedmode compared with sealed buildings. This is often perceived as a problem with operable windows, but perhaps the mixed-mode buildings being newer offset this fact.
As noted in Table 1, the twelve mixed-mode buildings differed from the overall database in several characteristics. On average, they were smaller in size (gross floor area), newer, and also included a larger percentage of educational buildings. To examine the extent to which these differences affected satisfaction scores, thermal comfort and air quality satisfaction in the mixed-mode buildings were compared with subsets of the larger database that were comparable in these three characteristics. The findings, summarized in Table 3, reveal that the mixed-mode buildings still had clearly higher levels of thermal comfort and air quality satisfaction even when these characteristics were comparable.
A significant majority of the buildings surveyed were office buildings. Although the educational buildings represent a much smaller sample size, there is a clear trend that satisfaction with both thermal comfort and air quality was higher in these buildings compared with the overall database (and this holds true for both the mixed-mode and other educational buildings). As noted previously, three of the four mixed-mode educational buildings had air-conditioning only in limited spaces such as laboratories and large assembly rooms. Therefore, the rooms more commonly occupied by  survey respondents in those three buildings were likely relying exclusively on natural ventilation.
For the non-mixed-mode buildings, satisfaction with thermal comfort and air quality improved slightly in the subset of smaller buildings, and even more significantly in the newer buildings (the improvement in air quality satisfaction was the most significant). Therefore, while the mixed-mode buildings were still scoring higher, the difference between them and the other buildings in these subsets was less pronounced.

Thermal comfort
Thermal satisfaction was assessed with the question: 'How satisfied are you with the temperature in your workspace?' Looking first at the overall database (that is, primarily air-conditioned buildings, with the twelve mixed-mode buildings removed), when analysed by individuals' responses, 41% of all workers expressed some level of dissatisfaction with the thermal environment ( Figure 2). This is a significant shortcoming from the goal of thermal comfort standards, which aim to create environments in which no more than 20% of the people are dissatisfied. This can be compared with Figure 3, the frequency distribution of thermal satisfaction responses in the twelve mixed-mode buildings. Dissatisfaction has dropped to 25%, which is significantly lower, but still somewhat above the acceptability criteria in the thermal comfort standards.
The current authors wondered if there was perhaps just a small fraction of poorly performing buildings contributing to the low satisfaction. Therefore, the analysis was performed again for the non-mixed-mode buildings, using the building as the unit of analysis (Figure 4). These results were equally concerning and revealed that only 11% of the buildings in the CBE database met ASHRAE Standard 55's 80% acceptability threshold (where 'acceptability' here is defined as votes of greater than or equal to 4, or neutral plus the top three categories of satisfaction). This is rather convincing   evidence that the standard practice of air-conditioned buildings is not reliably providing occupants with a satisfactory thermal comfort.
When the results were considered from the branching questions, enquiring about reasons for dissatisfaction ( Figure 5), the top reasons were about spatial nonuniformity ('my area is hotter/colder than other areas'), control ('thermostat is inaccessible' or 'adjusted by other people'), a lack of air movement ('air movement too low'), and speed of response ('heating/ cooling system does not respond'). Only 3% of the dissatisfied respondents referred to drafts from windows. Figure 6 is a cumulative frequency graph showing the percentile ranking of all 370 buildings in the database, based on the building's mean score for the 'thermal satisfaction' question. Triangles represent buildings that are mixed-mode, while diamonds represent the remaining buildings in the database. The median Figure 5 Reasons for thermal dissatisfaction (percentage of the total thermal dissatisfaction votes).Top 5 reasons for dissatisfaction were related to spatial non-uniformity, control, lack of air movement, and speed of response  Occupant satisfaction in mixed-mode buildings satisfaction score for each building set is shown as coloured symbols on the y-axis. The mixed-mode buildings were all in the top half of the percentile ranking, with a few being among the very best performers. In fact, eight out of twelve mixed-mode buildings are in the top quarter of the percentile ranking.
Of the mixed-mode buildings that had relatively lower scores, open-ended comments referred to complaints about conditions being too cold (sometimes referring to winter, other times suggesting that the air-conditioner was on when it did not need to be), complaints about drafts from vents, or thermostats not working. Only one building had a few complaints related to the windows, and respondents did not like that the mechanical air distribution was turned off when a window was opened because sometimes only a limited number of people got the benefit of the window while the air circulation was shut off to a larger zone.
Using these cues, the authors looked further into the thermal comfort scores for the group of mixed-mode buildings, testing for a variety of different indicators that might be contributing to the high satisfaction scores, including the size of the building, the number of occupants, the year of completion, and climate. The relationship to size and the number of occupants was not significant within this group of twelve buildings, but climate and age did reveal a pattern.
Using annual heating degree-days (HDD) (assuming a base temperature of 658F, or 188C) as a simple metric for the severity of climate, there was a correlation with the climate in which the buildings were situated. Figure 7 shows the relationship between HDD and the thermal comfort scores reported in the mixed-mode buildings.
The cluster of six buildings in more moderate climates had mean satisfaction scores over a wide range, and also included the buildings with the highest thermal satisfaction scores. But it can be seen that most of the buildings in the colder climates (a higher HDD) have somewhat lower satisfaction scores than warmer ones. Looking through comments and sources of dissatisfaction for the colder climate group, no evidence was found that occupants were opening windows during the winter. As noted above, problems generally focused on thermostats that were not working, drafts from vents, etc.
In addition, there was a positive correlation between thermal comfort and the year the building was built (Figure 8), which is promising for the future of mixed-mode buildings. In the database overall, there is a similar, but much less pronounced, trend towards higher thermal comfort satisfaction levels in newer buildings.
There was some commonality with the types of systems in the lower scoring mixed-mode buildings. The systems are primarily 'changeover' systems, where the HVAC system has been interlocked with the windows so that when the windows are opened, the HVAC system turns off. In these cases, there were specific complaints that when one occupant would open a window, the subsequent HVAC shutdown would make others uncomfortable.
Finally, there was also some commonality with the systems in the higher scoring mixed-mode buildings. None of the top five had air-conditioning systems in the commonly occupied parts of the building; instead, two has radiant cooling systems, and the other three relied on ventilation systems (both natural and mechanical) for cooling.

Air quality
Air quality satisfaction was assessed with the question: 'How satisfied are you with the air quality in your workspace?' Air quality fared slightly better than thermal satisfaction. Again, when responses by Figure 7 Thermal comfort versus heating degree-days. Six buildings in moderate climates had highest scores, but also covered a wide range individuals were analysed, overall 31% of workers were dissatisfied with air quality in the non-mixedmode buildings ( Figure 9) compared with only 14% in the mixed-mode buildings ( Figure 10). When the non-mixed-mode responses were analysed by building, only 26% of the buildings met the common 80% acceptability threshold ( Figure 11). The most common complaints from dissatisfied occupants were that the air was 'stuffy/stale', 'not clean', and was 'smelling bad'.
The cumulative frequency graph for air quality satisfaction ( Figure 12) shows that mixed-mode buildings typically perform very well, with all but two falling in the upper quartile. Complaints from occupants of one building that did not perform as well spoke frequently of dryness, while occupants of another building referred to smells from a nearby cafeteria and the lack of fresh air.
In comparing air quality satisfaction with annual heating degree-days (Figure 13), the same general trend was found: a slightly lower satisfaction level as heating degree-days increase. In particular, the two outliers with the lowest satisfaction scores were both in cold, dry climates. In both cases, occupants complained that the air was too dry and stuffy and that there was not enough fresh air in the building. The comments indicated that the windows might not be open often (especially in the winter), so the air quality problems were more likely to stem from the lack of humidification in the mechanical systems of the building. There was also a positive correlation between the year of completion of the building and air quality satisfaction ( Figure 14). This relationship was more pronounced than the similar thermal comfort correlation ( Figure 8).

Acoustics
For the entire database, acoustics received the lowest mean satisfaction scores. Figure 15 shows that mixed-mode buildings performed only slightly better on average, but they covered the full range, from the best to the worse. This is where the comments can be particularly revealing. In one building, people spoke of being under a metal roof that was noisy during heavy rain and snow. In another, they complained about the public address system and the disruption of tours going through the laboratory. Only one person in all the mixed-mode buildings complained of acoustical issues that were related to open windows -a teacher in a classroom situated next to a noisy playground. This does not imply that outdoor noise is never a problem in buildings with operable windows, just that it was not in this particular set, and the issue is clearly very site-dependent.
Given that mixed-mode buildings generally outperform the larger database of buildings in many areas, it was notable to see that they received scores closer to the overall database average in this area. Most of the mixed-mode buildings are 'green' buildings that have been designed for daylighting and good indoor air quality, leading to open-plan offices that contain many hard-finish materials. Comments from respondents indicate that these factors may be the cause of the higher levels of acoustical dissatisfaction in these buildings; respondents often note that there is little acoustical privacy in their space. For instance, one occupant noted this relationship specifically, saying: There is a certain openness to the building and that lends to the acoustics not being very effective.  (ASHRAE, 2004a(ASHRAE, , 2004b. However, satisfaction scores from the buildings in the CBE database (excluding the mixed mode), most of which have conventional air-conditioning systems, indicate that buildings are falling far short of these standards. It was disturbing to find that only 11% of the these buildings met the intent of the thermal comfort standard, with an overall average of only 59% of occupants expressing satisfaction with the thermal environment. Thermal dissatisfaction was most commonly related to people feeling that they did not have enough control over their environment, in addition to complaints about air movement being too low. This is particularly interesting given that thermal comfort standards are geared towards limiting air movement, mistakenly believing that drafts are a more common problem.
Mixed-mode buildings are performing much better than the overall building stock in the database, particularly with regard to thermal and air quality satisfaction. Using a seven-point satisfaction scale of þ3 (very satisfied) to -3 (very dissatisfied), the mean thermal satisfaction in mixed-mode buildings was 0.81 compared with -0.13 for the overall database (a difference of 0.94 points). The difference was even larger for air quality, with a mixed-mode mean of 1.71 compared with 0.28 for the overall database (a difference of 1.43 points). When mixed-mode buildings were compared with smaller subsets of the database that had comparable characteristics of size and building age, mixed-mode buildings continued to perform better, although the differences were reduced somewhat (particularly for air quality in newer buildings). For both thermal and air quality satisfaction in mixed-mode buildings, results indicate a relationship between climate (the highest scoring buildings were in more moderate climates, while buildings in colder climates scored lower on average, particularly with regard to air quality), and age of buildings (there was greater satisfaction in the newer buildings, again even more pronounced for air quality).
While the trends in occupant satisfaction are clear, the exact causal mechanisms are less so, and would require more rigorous case studies and field monitoring beyond the scope of the survey methods used in this research. Occupants' comments in the surveys, combined with findings from other research in the field, suggest that people value operable windows for a wide variety of reasons -personal control of their thermal environment, increased air movement, perceived fresh air, and connection to the outdoors. Mixed-mode buildings can potentially provide these benefits, while still offering a higher degree of thermal control through mechanical means compared with buildings that rely on natural ventilation exclusively.
In the group of mixed-mode buildings studied, some general trends were seen related to the types of mechanical systems and controls. The best performing buildings had either radiant cooling or only mechanical ventilation, but no air-cooled systems in the spaces primarily occupied by workers (that is, they may have had air-conditioning in large assembly rooms). The lowest performing mixed-mode buildings tended to be changeover systems, where there were problems with the window interlock systems. This suggests the importance of a well-integrated design where the mechanical and natural systems can work well together, and occupants have the ability to override automated controls as needed or desired. Occupants who have taken the survey can often provide very useful cues for understanding how the building is working not just at their individual desks, but for the building as a whole. In mixed-mode buildings, comments indicate that the relationship between the mechanical and natural systems in the building are not always working as planned, which can lead to the windows being shut more often than necessary. As one respondent who worked in the building that scored lowest on thermal comfort noted: I do wonder why they put windows in then told us not to open them, as they would mess up the air system. This type of disparity between how the buildings were designed and how they are running needs to be actively addressed through building commissioning and clear and robust communication to building occupants. While these issues are, of course, important in conventional air-conditioned buildings as well, they are particularly critical given the unfamiliarity of mixedmode design and operation.
Providing workers with a quality indoor environment should be a goal of any building design, but is particularly important for green buildings that claim to be more responsive to supporting occupant comfort, health and productivity. Improving the quality of buildings critically depends on accountability and learning from experience -what works, what does not, and what choices about building design or operation can make the biggest difference. The voices of the occupants are an invaluable component of that assessment. As the construction and property industry moves towards embracing high-performance, green buildings as the industry standard (as one must), the industry must also insist that post-occupancy evaluations be a natural part of that process. In the end, everyone benefits from learning how a building performs in practice.