Home

Smart Work Package 4.2. Smart field test: experience of users

image

Contents

1. yes no 15 When you don t have linked Smart to your computer what s the reason for this m not interested in Smart I wouldn t know how to link Smart Another reason ViZ 68 16 Do you use Smart to regulate your comfort temperature level at work Yes a few times a day Yes a few times during the week No almost never 17 When you regulate the temperature by means of Smart do you notice the effect of your ac tions Actually I don t use Smart very often so there are no effects to be noticed Although I use Smart I don t notice effects of my actions After regulating the temperature by means of Smart my complaints gradually lessen Regulating the temperature by means of Smart clearly has a positive effect 18 If someone regulates the temperature by means of Smart how does it happen Without consultation skip the next question With consultation beforehand 19 What do you think is the reason for these consultations To verify that someone is not the only one that feels uncomfortable To ascertain that others are not disadvantaged by the desired temperature setting To persuade others to enter the same temperature settings into their computer Other reasons Viis ceriose a a E ECN C 02 094 67 20 When you almost never use Smart to regulate your temperature settings what s the reason Not necessary no time I didn t
2. 10 l 0 Few times aday Few times a Almost never No answer given week Possibility Figure 8 4 Use of Smart We identified the following reasons to be the most important for test users in not using Smart in actions of comfort management Test users either did not think of Smart in the moment they un dertook an action directed at adjusting thermal comfort 40 of the respondents or they ex pected other options they could apply for this purpose to be more effective than Smart 33 of the respondents Almost half of the respondents who did use Smart to regulate thermal comfort did not notice any effect of such action Almost all respondents indicated that they used the ventilation valves about as often as they did before Smart was introduced During the group interview test users made clear that in most cases Smart was not the first op tion they seized for to adjust thermal comfort In order to validate this observation for the larger group of test users questions on preferences for different routes of comfort management were included in the questionnaire Route preferences were measured as the regulatory option used first to adjust thermal comfort That is test users were asked what they did first if they wanted to adjust thermal comfort in two situations i e when they perceived the climate in the office as too hot or as too cold The answers given are plotted in Figure 8 5 100 90 80
3. I pull out a cardigan or vest I open the windows when they are closed I regulate the temperature by means of the thermostat or Smart skip the next question 66 ECN C 02 094 12 If you don t initially use Smart to decrease office temperatures what is the reason for this The actions I take are routine for me I don t think of Smart when I m too hot The actions I take have a positive effect I have the idea that the effect that Smart has is less than the effect of the actions I take Changing my clothing or closing windows also have other positive effects like 13 Some time ago I wasn t possible to open or close the small windows by hand The office temperature became quite high during this period What did you do to decrease the tem perature during this period I didn t do anything because I quite liked the higher temperature I didn t do anything because I wasn t interested enough to perform actions to decrease this temperature I adjusted my clothing after which I did use Smart to change my temperature settings after all I did use Smart to change the temperature setting after which I adjusted my clothing to the higher temperature after all Before I tried to regulate my comfort by the methods described above I first phoned the building operator to ask him to open the windows a m e b O Use of Smart 14 Do you have Smart as short cut or link on your computer
4. 4 70 4 60 4 50 40 4 30 20 4 10 0 P Too cold Too warm O Adjust clothing BUse windows WUse Smart Figure 8 5 Actions taken first for adjusting thermal comfort The results indicate that most respondents when feeling uncomfortable first adjusted their clothing before they took action to adjust the temperature by manipulating the ventilation valves ECN C 02 094 33 or turning to the Smart system These priorities are confirmed by answers to a question on the route preferences being followed in the situation in which the ventilation valves could not be operated Even in this situation only 21 of the users indicated that Smart was the first option chosen for managing comfort When inquired about reasons for neglecting Smart in the first in stance a considerable part of the respondents answered that they deemed the other options more effective whereas they also attributed considerable influence to routines already established be fore the introduction of Smart see Figure 8 6 100 90 4 80 70 4 60 4 50 4 40 4 30 4 20 10 4 0 fal Too cold Too warm O Routine Do not think of Smart E Other action is effective E Additional advantages other option Figure 8 6 Reasons given by test users for not using Smart The results as depicted in Figure 8 5 seem to indicate that the first action users take to improve the comfort situation
5. H repo 281 ulesi wl ISI Vernseken in hopende stempenide r Figure 8 1 Smart user interface The user could enter the following preferences into the system e Thermal comfort Three options were offered to the user indicated by warmer un changed and cooler warmer ongewijzigd and koeler respectively in Figure 8 1 e Energy efficiency The user could urge the Smart system to save energy in realising the pre ferred comfort by ticking the box energy efficient energiezuinig in Figure 8 1 The Smart system informed the user about the number of votes entered during the current voting period within his zone and displayed whether these votes promoted an unchanged increasing or decreasing temperature The number behind requests during the current voting period ver zoeken in de lopende stemperiode see Figure 8 1 indicated the total of votes entered during the current voting period By the box behind tendency tendens the user was informed whether as a result of voting the temperature in the zone was rising or decreasing If the ten dency was unknown this was also communicated de tendens van uw zone is onbekend see Figure 8 1 8 3 Results of the field test A group interview a brief contextual interview with two test users and a questionnaire meas ured the results of the field test In addition Building Operations and Mainten
6. for further explanation of the system and its test About two months later a next e mail an nounced the actual introduction of Smart indicating how the employees could log in on the Smart screen by using their desktop computers This screen supplied the interface by which test users could interact with Smart mainly to enter comfort preferences see below Only at the end of the test it appeared that a few employees had not been able to get access to Smart How ever the large majority of them about 25 persons used the system and participated in different parts of the investigation It is important to emphasise here that Smart was offered to test users as an additional electronic option for handling comfort in their work environment The mechanical provisions already pre gt Fanger a Danish researcher developed a complicated formula relating empirically derived parameters of perceived thermal comfort in buildings such as air temperature humidity draft and clothing of test persons From this widely used formula average settings for comfort management in buildings are derived Fanger 1970 The standard temperature and ventilation settings are chosen such that no more than 10 of all users will complain about the comfort they perceive This 10 level called predicted mean value PMV is a standard value laid down in the ISO standard EN 7730 Fanger 1996 The optimal comfort level is reached if only 5 of all users of a building complain
7. I normally consider the temperature as 7 What do you normally think of the ventilation level on the floor you work on Please tick the appropriate box below for the situation after the introduction of Smart Table C 2 Possible assessments ventilation level within building 42 Not Sufficient Too much it is enough almost like drafting When I enter the building in the morning I normally consider the ventilation as Round noon I normally consider the ventilation as At the end of the day I normally consider the ventilation as 8 Can you mark the air quality on your floor in building 42 You can use marks between 1 to 5 where 1 stands for mushy and 5 for fresh Temperature and comfort setting 9 When you are cold which action do you perform first I pull on an extra cardigan or vest I close the windows when they are open I regulate the temperature by means of the thermostat or Smart skip the next question 10 If you don t initially use Smart to increase office temperatures what is the reason for this The actions I take are routine for me I don t think of Smart when I m too cold The actions I take have a positive effect I have the idea that the effect that Smart has is less than the effect of the actions I take Changing my clothing or opening windows also have other positive effects like 11 When you are too warm which action do you perform first
8. about comfort perception over time revealed that appreciation of air temperature and ventilation decreased during the day Proceeding from morning into af ternoon the number of respondents perceiving the building s climate as too hot increased see Figure 7 1 100 90 80 70 60 50 40 30 20 10 0 r 7 morning around noon end of day m Too warm m Nice and warm O Not too warm but also not too cold Figure 7 1 Assessment of thermal comfort during the day With respect to ventilation a similar but weaker trend was observed i e decreasing comfort during the day see Figure 7 2 below 100 90 80 70 60 50 40 30 20 10 0 morning around noon end of day O Not enough Sufficient m Too much Figure 7 2 Assessment of ventilation during the day 7 3 Use of options for comfort regulation Independent of Smart users of building 42 can regulate their comfort by setting a room thermo stat and opening and closing of the ventilation valves see above Various items in the ques tionnaire addressed the question of how often these regulation possibilities were used in prac tice It appeared that the thermostat was only occasionally set by 12 of the respondents In setting the thermostat no consultation of colleagues beforehand of afterwards took place ac cording to 76 of the respondents Tw
9. about perceived comfort ECN C 02 094 17 sent such as the thermostat and the ventilation valves remained available to them during the test 5 2 Test users The group of test users participating in the study about 25 persons consisted of employees of ECN s administration department The distribution of sex in this group was about equal whereas most members were above the age of fifty for detailed information see Appendix B Part B 3 Computer literacy was supposed to be high because of the kind of office work the test users practised this was fully computer based 18 ECN C 02 094 6 USER REPRESENTATION IN DESIGN Representing users directly in the design process was undermined by the fact that in the early stage of this process the users were still unknown For a long time it remained unclear which department would move into the new building in which Smart was to be installed Nevertheless designing Smart needed an image of its future user to be made explicit and to be fed back into design decisions Especially the open plan character of the office room was an issue of concern in this respect Below we report the deliberations of the design team about user representation and the related design decisions made 6 1 User heterogeneity The use context of the Smart test differs from systems developed for in home comfort manage ment systems such as early applications of COMFY Boertjes et al 2000 Within the home the roles
10. air blown into the building is set according to the Fanger regime Electric convection heat ers in the ceiling can heat up the air further locally in places that are too cold close to windows and corridors The building operators set the temperature of the blown in air as well as the local heaters temperature By setting a room thermostat workers on the office floor can adjust the temperature per block of 8 ceiling heaters within the range set by the operators plus or minus three degrees They also have the possibility to open the valves above the windows to increase the ventilation level within the building for a more extended description of the design of the building and its provisions see WP 4 1 final report Jelsma 2001 During the fall of 2001 building 42 1 came into use About thirty employees of the Shared Service Centre the central administration unit of ECN started their work on the first floor The floor was laid out as an open plan office space with clusters of desks separated by low fences The Smart system was introduced about six months later in April 2002 when the office work ers were well accustomed to the building and its provisions The Smart system was briefly introduced to the unit s employees by an e mail announcing the preliminary investigation In this e mail the employees were invited to join as test users and to collaborate in the test Handing out the questionnaire for the first phase was used as an occasion
11. did not use the options that they did not understand 69 The fact that about half the users did not understand the op tion energy efficient may thus account for the fact that this option was almost never used A further question inquired test users about the effects of ignorance about the way Smart processes votes of users entered during a voting period within a particular thermal zone The results are depicted in Figure 8 7 In most cases lack of transparency was said to lead to a decrease in the use of Smart From this figure and the possible effects mentioned above it can be concluded that a knowledge gap on the actual functioning of a technology may lead to unwanted effects 36 ECN C 02 094 Times mentioned MN wo fF a DN C SS 0 No effect Smart is used Comfort Other effect less preferences are entered more extremely Mentioned effects Figure 8 7 Effects as presupposed by test users to occur in situations in which several users within one segment enter comfort preferences into the Smart system within the same voting period 8 3 5 Possible improvements of the Smart system In this section opinions of respondents about possible improvements of the Smart system are reported These opinions are derived from the group interview as well as from answers to the questionnaire The group interview indicated that negligence of Smart by users mainly had to do with i options for co
12. for representation of end users in the design of the system were distinguished 1 As a group in a climate zone After negotiations the group enters its preferences into the system That is negotiations about preferences are delegated to the group ECN C 02 094 19 2 As individuals Individuals enter their preferences into the system which then calculates and implements an averaged solution In this case negotiation between individual preferences is delegated to the system and is invisible to the users 3 As individuals in a group i e a combination of option 1 and 2 End users are conceived as individual group members who must be enabled by Smart to consider the choice made by other group members in making their own and so influence the outcomes of the choice pro cess in terms of thermal comfort acquired This requires the system to make the choice pro cess transparent to the group members After appraising the merits of all options the design team came to the following conclusion Op tion 1 waters down the goal of Smart too much Moreover negotiation within the group may lead to conflicts Option 2 keeps the suggestion of individual comfort management alive for end users but in reality the preferences entered individually are thrown into an electronic melt ing pot hidden to the users This may undermine the confidence of users in the system There fore from a user perspective option 3 is to be preferred but probably was the most
13. in building 42 1 took a large part of the interview the data from this part have been taken up in Section 5 Asked about his expectations about Smart the building operator described the regu latory possibilities of Smart as limited Since control of ventilation which is related with hu midity control falls outside Smart s range of action temperature remains as the sole comfort parameter users can influence through Smart On the other hand Smart could enhance user sat isfaction by giving users an improved feeling of control that might decrease complaints Though performance of the climate system and changes in settings are continuously registered by the building management system the operator emphasised that building 42 1 is new and lacks a historical record of stored data This might hamper evaluation of the performance of Smart where comparisons to be made require a reference as in measuring changes in energy effi ciency According to the operator supporting the opportunity to optimise energy consumption against energy price by trading on a liberalised energy market is the greatest promise of smart climate systems for buildings 7 2 Comfort assessment and control Comfort as perceived by future test users before the introduction of Smart was assessed by a questionnaire see Appendix B First we asked respondents to rank their perception of thermal comfort on the work floor On average perceived thermal comfort was marked as 6 9 on a
14. in the morning 0 0 19 3 4 Temperature round noon 0 0 13 6 7 Temperature at the end of the day 0 0 10 5 11 Question 10 Table B 11 First questionnaire answer question 10 Possibility Times mentioned Thicker clothing at work than at home 0 Same clothing at work as at home 20 Thinner clothing at work than at home 2 Part of clothing taken off during the day 4 Question 11 In the following table it is indicated how often the various ventilation assessments were made Table B 12 First questionnaire answer question 11 Notenough Sufficient Too much it is almost like drafting Ventilation in the morning 5 20 0 Ventilation around noon 7 18 0 Ventilation at the end of the day 8 17 0 Question 12 The mark given was 2 9 with a standard deviation of 0 9 scale 1 5 Question 13 Table B 13 First questionnaire answer question 13 Frequency setting temperature by means of the thermostat Times mentioned Quite often 0 Sometimes 3 Almost never 23 62 ECN C 02 094 Question 14 Table B 14 First questionnaire answer question 14 Consultation related to temperature setting Times mentioned No consultation Consultation before the temperature is set Consultation after the temperature is set 20 5 1 Question 15 Table B 15 First questionnaire answer question 15 Reasons not to adapt the temperature setting Times mentioned Not necessary no time Unknown possibility Because of
15. is interviewed as an expert on the use of the system During the interview the user explains why and how he or she uses a cer tain technology or system While the user uses the system or technology the interviewer can interfere by asking question why the user acts the way that he does By using contextual inter views real time information on how systems are actually used can be obtained Contextual in terviews were used on a modest scale to obtain detailed data on an individual level about the daily use of the Smart system 4 2 5 Participant observation In participant observation the observer becomes a member of the community he wants to ob serve After some time the community will not notice the observer anymore so that he can ob serve the daily state of affairs without distorting it Participant observation was used in the in vestigation to obtain background information for formulating the questionnaire and to obtain additional information on the actual behaviour of test users directed at comfort management ECN C 02 094 15 4 2 6 Brainstorm session Brainstorming was used as a technique to generate ideas about ways to save energy through Smart One session was run for which a group of internal experts was recruited 4 3 Research scheme During the first phase before the introduction of Smart the following methods or instruments were applied e Participant observation of the users in building 42 1 e Interview with one of the b
16. it appeared that the assess ment of thermal comfort at the test site changed during the day By the majority of users the temperature in the morning was indicated as being not too warm but also not too cold At the end of the day the opinions on thermal comfort had changed considerably about half of the test users assessed the temperature as too warm see Figure 8 2 100 90 4 80 70 4 60 50 4 40 30 20 10 0 A morning noon end of day E Too warm E Nice and warm O Not too warm but also not too cold E Nice and cold Figure 8 2 Assessment of thermal comfort during the day Respondents assessment of ventilation exhibited a similar trend The majority assessed the ventilation level as sufficient during the morning but in the course of the day satisfaction de creased An overview of the assessment of the ventilation level during the day is given in Figure 8 3 ECN C 02 094 31 76 100 90 80 70 60 50 40 30 20 10 morning noon end of day ONot enough E Sufficient HToo much Figure 8 3 Assessment of ventilation level during the day When asked to compare comfort levels before and after the introduction of Smart most respon dents 94 indicated not to have noticed any change in comfort level since Smart had been in troduced 8 3 2 Use of options for comfort regulation During the field test the test users cou
17. means of a thermostat ventilation by means of opening or closing windows heat originating from sunshine by means of blinds light originating from sunshine by means of blinds light switching on or off of artificial lights other aspect of comfort V1Z ce cece ees 56 ECN C 02 094 Inner climate building 42 9 What do you normally think of the temperature on the floor you work on Can you please tick the appropriate box below for the situation on an average working day in this season Table B 1 Possible assessments temperature Situation Too Niceand Nottoo warm but Niceand Too cold cool also not too cold warm warm When I enter the building in the morning I normally consider the temperature as Round noon I normally consider the temperature as At the end of the day I normally consider the temperature as 10 How are you normally clothed when you re working The clothes I wear at work are warmer than the clothes I wear at home I wear the same clothes to work as I do at home The clothes I wear at work are thinner than the clothes I wear at home I often take off some of my clothes cardigan vest etc when I m at work 11 What do you normally think of the ventilation level on the floor you work on Can you please tick the appropriate box below for the situation on an average working day in this season Table B 2 Possible assessments ventilation l
18. of existing potential is generally far from being optimal At the same time information that is or can be made available through networks such as Internet and power lines is seldom exploited fully to enhance the performance of such systems To improve this situation a new generation of control systems is being developed for use in the operation of public commercial and residential buildings The general objective of the Smart project is to formulate requirements for such new systems The technology of these systems will be based on agent mediated communication over local networks the Internet and power lines Such novel control systems not only should improve the performance of the building but should also offer opportunities to users i e building operators as well as residents and workers for linking comfort management with improving energy efficiency and the application of renewable energy Along with energy saving cost saving could be reached by automated selling and pur chasing of energy through agent mediated electronic marketplaces Kamphuis et al 2001 One of the more specific goals of the Smart project is to gain experience about implementation of requirements for such a system in an experimental setting Right from the start aspects of use such as user acceptance and embedding the system in a practice of use were included in the development of system requirements to be explored in the field test During the course of the Smart
19. or down is mapped to the desired decrease or increase in comfort level in the Fanger model by a temperature shift in the Fanger index Fanger 1994 calculation By the vot ing procedure a time dependent modulation of the comfort level as defined by the building manager is obtained The building manager is then able to set the modulated comfort level de fined in this way as the new comfort level after the votes latency period In this way the in volvement of the user in management of comfort is expected to decrease in time that is after a while the new adapted comfort level set matches the level as preferred by the user The latter comfort level is used for calculating the strategy for comfort aspect controllers during the next optimisation cycles The trend as determined during the field experiment is depicted in Figure 10 3 ECN C 02 094 43 North segment Comfort 0 2 O pean ee Feet pee eee ee ep a ptt me hat tee 7 g7 10 aii aa 14 16 18 f K 0 2 5 E i 1 2 H E a fo G oO 0 4 T S x 2 0 6 i r o yS z oe 0 8 3 a 1 1 2 Time BM Setting Fanger Index Figure 10 3 Modulation of the Fanger index as derived from the votes issued over time The figure shows that the largest comfort deviation occurs at around 1 30 p m Users then per ceive the ambient air temperature as higher than desired In the early morning the opposite is true i e the
20. project the user setting gained an even more prominent role in the design of the Smart system Meant to improve comfort for individual users the system however had to reckon with a collective of users in one large open plan office floor at the test site which introduced severe complications for its design The present study reports about experience of test users with a prototype of the system and assesses the results in relation with design constraints The Smart project as a whole consists of a number of research and development activities di vided over several work packages For an elaborate project description and work plan we refer to the overall Smart project proposal Projectplan 2000 The present report is the final report of work package 4 2 The complete list of work packages can be found in Appendix A of this re port 10 ECN C 02 094 2 APPROACH For this work package we applied the same general approach as for WP 4 1 see ECN C 02 008 Summarised briefly this implies that we conceive systems the building its technical pro visions including Smart and its users to be ensembles or networks of actors interacting i e acting and reacting in the production of effects which can be observed measured and assessed to optimise the original design of the system For the system to produce effects on the behaviour of its users factors of most importance are e The strength and weaknesses of scripts i e the extent to which
21. specifics of the system guide intended as well as unintended forms of use e Overlaps and misfits between design logic and use logic i e between interests needs values and conceptions of designers on the one hand as inscribed in the hardware and software of the system and of the users on the other In the test situation investigated the material landscape in which the users navigated to optimise their comfort did not consist of the Smart system only Users had also direct access to other equipment mediating comfort such as ventilation valves above the windows which could be opened and closed manually and thermostats for setting temperature Thus users were able to manage their comfort via different routes through the provisions available to them In fact the Smart system had to compete with these alternatives to enrol users Therefore we did not only study the interactions of users with the Smart system per se but included in our study the inter actions with other agents for comfort mediation Choices by users to take one or another route for comfort management may be influenced by the interaction of scripts acting out the underlying design logic of the different technical options offered on the one hand and by use logic on the other That is the different options available may appeal in different ways to the logic of users in terms of desirability comprehensibility et ceteras The logic which users often tacitly mobilise in making s
22. ten point scale 10 indicates perfect comfort Air quality was marked as 2 9 on a scale from 1 5 on which 5 means fresh Sec ond we requested respondents to rank determinants of comfort perception in order to be able to appraise possible dissatisfaction with Smart later on during the field test Such dissatisfaction might appear to be related to one or more high ranking parameters that were not under the con trol of Smart In addition this measurement yielded a check on how comfort was appreciated by the future test users in relation to the parameters derived from the Fanger doctrine which was the theoretical ground on which comfort management in the test building was based If Fanger is right comfort in the building was expected to be perceived as satisfactory If so this would con strain the possibilities of Smart to improve comfort Respondents ranked ambient air temperature i e whether the air feels comfortable or not and draft as the most important comfort determinants by far Besides being probed about determi nants of comfort assessment as such respondents were asked which aspects of comfort they wanted to have influence on Besides ambient air temperature and ventilation respondents mentioned lighting artificial light as well incoming sunlight as important factors they should ECN C 02 094 23 be able to control Air temperature preferred by respondents ranged between 19 9 plus or minus 0 8 degrees Answers to questions
23. terms of the real room temperature Displaying the current temperature in the Smart screen would urge him he thought to check whether the ventilation valves needed opening or closure or make him aware of a cold he might have caught Y had never used Smart up to the moment of the interview This did not imply that she was in different about comfort to the contrary She used to be active in setting the room thermostats and handling the ventilation valves in which she involved colleagues by consulting them Y was rather dissatisfied with the comfort situation in the office and she regularly called the building operators to complain about it After reading the introductory mail she successfully in stalled the Smart screen Her first impression of the screen was favourable Her first vote was a vote for unchanged After entering this vote correctly the figure in the box requests in the cur rent period changed from 0 to 1 This change she misinterpreted as someone else having en tered a preference different from hers Like X Y neglected the box energy efficient which she explained by referring to the clothes she wore Y regretted the fact that Smart offered no op portunities to communicate complaints about the current comfort situation directly to the build ing managers Comprehensibility as tested by the group interview and the questionnaire The group interview confirmed that not all aspects of Smart were completely clear to t
24. the com fort level they perceived That is there was no Smart related perception of improved com fort by test users who faced a concrete comfort deficiency in their office environment At the end of the field test about 50 of the respondents did not have an opinion on the usefulness of Smart As far as opinions on the usefulness of Smart were expressed these were mostly negative Most probably this low appreciation of Smart results from its per ceived and real lack of effectiveness Nevertheless from the beginning of the test the gen eral attitude among test users concerning Smart was positive presumably because Smart was seen as giving users a voice in comfort management see also next point Test users made clear that they wanted to be in control of their own comfort The fact that Smart invited them to vote for thermal comfort was welcomed as an improvement compared to the existing situation In the same vain a majority of users wanted to keep control over ventilation by operating the valves instead of delegating this function to Smart We see us ers attitude toward saving energy as a further confirmation of this preference for control Users were willing to save energy in the office only on the condition that they could decide on any related decrease in comfort Smart did not meet the demand of transparency for users in all respects Especially the way in which Smart processed preferences entered by different voters and reacted to
25. the consultation process The temperature will be set by someone else shortly afterwards 10 10 2 1 Question 16 Table B 16 First questionnaire answer question 16 Frequency opening or closing the small windows to regulate the ventilation level Times mentioned Quite often Sometimes Almost never 0 Question 17 Table B 17 First questionnaire answer question 17 Consultation related to opening and closing the windows Times mentioned No consultation Consultation before the temperature is set Consultation after the temperature is set 9 14 3 Question 18 Table B 18 First questionnaire answer question 18 Reasons not to open or close the small windows Times mentioned Not necessary no time Unknown possibility Because of the consultation process The temperature will be set by someone else shortly afterwards 10 3 3 3 Question 19 Periods of time when it is not possible to regulate the office temperature sufficiently The following time periods were mentioned e Monday mornings after a cold weekend Afternoons Mornings on cold winter days ECN C 02 094 Summer days with temperatures exceeding 25 degrees Celsius The whole day because other people have different temperature preferences 63 e The whole day when the wind is strong additional information from someone who is sit ting next to the door e Between 12 and 15 in the afternoon too
26. up screen No opinion Not answered NU VDU 76 ECN C 02 094 Question 27 Table C 27 Second questionnaire answer question 27 Preferred aspects to be regulated by Smart Times mentioned temperature only 10 Temperature and small windows directly 2 Temperature and small windows indirectly 6 Temperature and internal ventilation 1 No answer given 3 Question 28 Various reasons were given Question 29 Table C 28 Second questionnaire answer question 29 Assessment user friendliness screen Times mentioned User friendly 14 Quite use friendly 2 Not user friendly 0 No answer given 6 Question 30 In the following table the percentage of people that understood a certain concept is given The percentages measured for the situation just after the introduction of Smart and a few weeks later are more or less the same Therefore only the percentage for just after the introduction of Smart are given Table C 29 Second questionnaire answer question 30 Concept Measured percentage of people indicating that they completely understood the concept used Segment 94 Warmer 94 Unaltered 89 Colder 94 Energy saving 50 Requests in the present voting period 78 Tendency 56 Annul 89 OK 94 Question 31 Table C 30 Second questionnaire answer question 31 Reaction on not completely understanding the concepts used Times mentioned Did not vote 4 Did vote but without using possibilities that were n
27. C 02 094 8 RESULTS OF THE SMART FIELD TEST According to the original plan the field test was scheduled to take place during the winter of 2001 2002 However due to design problems see Jelsma 2001 only in the spring of 2002 a prototype of the Smart system became available to be installed at the test site The field test proper lasted from April 25th until June 11th A disadvantage of the test being shifted to a warmer season was a diminished need to operate the space heaters the Smart system utilised for managing thermal comfort Moreover a higher outside temperature might induce discrepancies in comfort perception compared to the winter season in which the preliminary investigation had been carried out Such discrepancies might affect the referential value of the preliminary inves tigation For instance increased sunshine in the daytime might heat up the building and lower comfort although this effect was counteracted by sunblinds which had been installed mean while The comparability of the reference situation and the test situation might also be affected by a temporary failure of the ventilation valves during the test i e the valves could not be opened However this inconvenience lasted only about a week 8 1 Introduction of the field test In order to participate in the Smart test each employee had to reply to an email send by Building Operations and Maintenance When they did so participants received the address of the web page o
28. October 2002 ECN C 02 094 SMART WORK PACKAGE 4 2 Smart field test experience of users and technical aspects J Jelsma ECN Beleidsstudies A Kets ECN Beleidsstudies I G Kamphuis ECN DEGO W Wortel Kropman BV Acknowledgements The SMART project Smart Building Multi Agent InteRnet Technology is sponsored by SENTER within the framework of its BTS programme This report is registered under ECN project number 7 7425 04 02 The authors wish to thank all persons who have been interviewed participated in the group interviews or filled in the questionnaires for their collaboration and patience Abstract Agent based smart systems for energy management in buildings show great promise for en hancing comfort of individual users for saving energy and for enabling cost effective building management The study is part of a larger project to design implement and test an experimental smart system called Smart in a new office building at the ECN site near Petten The Nether lands The objective of this part was to investigate in a field test a prototype of Smart with spe cial attention for the user aspects In the study user comfort perceived office temperatures and ventilation levels and the usage of comfort regulation systems were assessed before and after the introduction of Smart by means of questionnaires For a more extended disclosure of these and other user aspects also inter views group interviews contextual intervie
29. Question 5 The mark given was 6 5 with a standard deviation of 1 2 Question 6 In the following figure it is indicated how often the various temperature assessments were made 72 ECN C 02 094 Table C 8 Second questionnaire answers question 6 Too Niceand Nottoo warm but Niceand Too cold cool also not too cold warm warm Temperature in the morning 0 2 13 1 5 Temperature round noon 0 0 12 2 7 Temperature at the end of the day 0 0 9 2 10 Question 7 In the following figure it is indicated how often the various ventilation assessments were made Table C 9 Second questionnaire answers question 7 Not enough Sufficient Too much it is almost like drafting Ventilation in the morning 5 14 0 Ventilation around noon 9 10 0 Ventilation at the end of the day 10 8 1 Question 8 The mark given was 2 6 with a standard deviation of 0 8 Question 9 Table C 10 Second questionnaire answer question 9 Action Times mentioned Adjust clothing 10 Open windows 6 Temperature regulation by means of Smart or thermostat 3 No answer given 3 Question 10 Table C 11 Second questionnaire answer question 10 Reason not to use Smart Times mentioned Action taken is routine I don t think of Smart when I want to regulate my temperature level The action taken has a positive effect The action taken also have other advantages 10 5 7 4 Here it must be noted that some respondents filled in more than on
30. about a week because of a failure in the electric controls see above 34 ECN C 02 094 general information on the possibilities to save energy through Smart or on comfort decreasing effects of their own energy saving behaviour 8 3 4 Comprehensibility of Smart Comprehensibility as tested by contextual interviews About a month after the introduction of Smart two contextual interviews were carried out with test users X and Y Observation made clear that X had made a shortcut to the Smart screen in his office bar While promoting his preference for improved comfort by correctly entering a vote through the Smart screen he neglected the box energy efficient He explained this by saying that he did not know what the effect of this action would be on the comfort level he was voting for Further he indi cated that information given in the box requests in the current period was unclear to him too What was current period supposed to mean an hour a day a week and who could have done the requests X others within the comfort zone everybody on the floor By co ordinating his voting behaviour with that of his neighbour he had found out that in any case persons within the same zone could make requests in a current period Further he had observed that after voting the OK button disappeared and that voting was impossible for the next hour He added that he preferred to have a reference for this own voting behaviour in
31. adiators or windows in the building the freshness of the air in the building fresh as an opposite of musty or stale the moistness or dryness of the air in the building if I have cold feet or not if I have cold hands or not if I have to perspire or not if I feel drafts or not other aspects of the inner climate viz 6 Which aspect of the aspects mentioned below that you feel is the most important aspect to consider if you have to indicate if you work in a comfortable environment my perception of the temperature of the air in the building the heat or cold radiated by the radiators or windows in the building the freshness of the air in the building fresh as an opposite of musty or stale the moistness or dryness of the air in the building if I have cold feet or not if I have cold hands or not if I have to perspire or not if I feel drafts or not other aspects of the inner climate viz 7 Which working temperature would you like to have in building 42 about 16 degrees Celsius or colder about 17 degrees Celsius about 18 degrees Celsius about 19 degrees Celsius about 20 degrees Celsius about 21 degrees Celsius about 22 degrees Celsius about 23 degrees Celsius about 24 degrees Celsius or warmer I don t know 8 Which comfort aspects of your working environment would you like to regulate temperature by
32. aggregates a number of sensors and actuators belonging to part of the installation In the current setting a protocol was defined to handle all data in the office space that were considered to be relevant for the field test Signals per climate zone included the measured temperature output only and the setpoints input and output Given the amount of data to be transfered a data collection frequency of once per 15 minutes was the maximum achievable Within the software of the InsiteView server a driver was written implementing a set of protocol inquiries The InsiteView server was directly connected to the internal local area network using central Internet technology To get access from the InsiteView server to the SMART software an interface based on Micro soft COM was written In COM remote procedure calls can be made after a marshalling proce dure for function arguments and returned values has been performed In this way InsiteView function calls and data structures available from Delphi ObjectPascal were made transparent through the network and available to the Java code from the SEBOS shell The mechanism is described in Kamphuis 2002 Thus the real time data needed per segment could be obtained combined with the other information and the setpoint values could be set from the SEBOS shell according to the strategy calculation 10 1 2 Measurements and control settings The existing building management system provided us with
33. ambient air temperature is perceived as slightly too low 10 2 Measurements during the field experiment 10 2 1 Temperatures and settings Temperatures as measured during the winter period November 2001 March 2003 in the north segment are depicted in Figure 10 4 44 ECN C 02 094 Temperature North segment Winter period C 23 22 5 0 4 8 12 16 20 24 Time Mon Tue Wed Thu Fri Figure 10 4 Temperature north segment The temperature curves indicate that the large heat capacity of the building leads to a very slow pace of reaching a maximum temperature of 22 degree Celsius on Monday lowest curve On all workdays this maximum temperature over 22 degree Celsius is reached only in the after noon Temperature set points for the controllers of the local heating units in the ceiling are shown in Figure 10 5 These setpoints are the result of the building manager s baselevel settings and of possible adjustments by users setting wall thermostats in the office room The figure clearly demonstrates the current heating regime and how the building reacts to it Early in the morning the setpoint is increased by the BMS to preheat the climate zones Just as shown by Figure 10 4 the building achieves the maximum temperature only after a period of 10 hours This is further shown in Figure 10 6 where the setpoints and realised temperatures are shown for a whole week It can be seen that t
34. an 1997 DUCOZT A prototype system for democratic user control of zonal temperature in air conditioned offices CIBSE 97 Virtual Conference Quality for People pp 1 15 ECN C 02 094 53 APPENDIX A WORK PACKAGES List of work packages of the Smart project WP 1 Multi agent frame work WP 1 1 Agent rolls and utility functions WP 1 2 Defining agent platform optimising strategy and tasks WP 1 3 Evaluation of electronic market mechanisms WP 1 4 Final report multi agent technology WP 2 Model studies and scenarios WP 2 1 Energy and comfort scenario analyses WP 2 2 Evaluation of comfort strategies and of concepts for building operation WP 3 Design of software application and field test WP 3 1 Mapping the local context WP 3 2 Interface design WP 3 3 Development of multi agent software application WP 3 4 Management and monitoring of field test WP 4 Study and evaluation of design and use WP 4 1 Comparing Smart logic with design logic of test building WP 4 2 User aspects of Smart 13 This final report concerns WP 4 2 54 ECN C 02 094 APPENDIX B FIRST QUESTIONNAIRE In this appendix the questionnaire presented to the users of building 42 is presented The ques tionnaire was originally drafted in Dutch First the questionnaire is presented After the questionnaire the answers given to the questions in the questionnaire are presented B 1 First questionnaire Questions The first questionnaire consisted of two
35. an be concluded that no significant change in the assessment of thermal comfort or ambient air quality between the two situations occurred This outcome is confirmed by the responses given to a straightforward question in the second questionnaire inquiring users whether they had per ceived any change in comfort level after the introduction of Smart Most respondents 94 re ported that they had not perceived any difference Somewhat earlier the group interview had shown the same negative outcome However in the case of the group interview the participants held strong opinions that this negative outcome did not prove that any comfort improvement re alised by Smart was lacking There might have been an undetected improvement they thought They judged it most probable that they would observe any deficiency in comfort while im provements stood a high chance to remain unnoticed 9 2 Saving energy attitude of users and use of the energy saving option The first questionnaire revealed a positive attitude towards saving energy in the office On the condition that responsibility for saving energy was left to the users a considerable percentage of users were willing to adapt their behaviour in order to save energy 10 Most probably this difference points to the temporary failure of the ventilation valves see preceding chapter ECN C 02 094 39 From the group interview and the second questionnaire a different picture arose Before adjust ing the t
36. anagement in buildings show great promise for enhancing comfort of individual users for saving energy and for enabling cost effective building management The present study is part of a larger project in which different aspects of the design and implementation of such a system called Smart in an office building were investigated It reports about the user aspects of the implementation of Smart in a field test and its related technical aspects In the springtime of 2002 a prototype of Smart has been installed and tested in an open plan office setting in a new building at the ECN site near Petten The Netherlands For the sake of the test the office space was divided in five comfort zones in each of which temperature control was possible to a certain extent During six weeks a test population of about 25 office workers had access to Smart by an interface installed on their personal desktop computers Through this interface the test users were able to enter personal comfort preferences relating to perceived room temperature once per hour by means of a voting protocol For each comfort zone the individual votes were processed regularly by the Smart optimiser in which software agents represented the interests of individual users The outcomes of optimising in terms of optimal temperature settings per zone steered the heating of the office space by setting the controls of dispersed heating units in the ceiling of the room The present study reports a
37. ance supplied re sults of measurements of physical parameters made during the test whereas Smart logged ac tions of users such as voting behaviour 22 persons filled in the questionnaire As some of the respondents unfortunately had no access to Smart see above not all questions were answered by all respondents This has been accounted for in processing responses An overview of all re sponses to the questionnaire i e responses given or left open by test users can be found in appendix C In the following sections the results of the Smart field test are described regarding the following aspects 30 ECN C 02 094 Perceived comfort Use of different options for comfort management Saving energy by Smart Comprehensibility of Smart Possible improvements of Smart Overall assessment of Smart 8 3 1 Comfort assessment To assess comfort at the test site as perceived by test users the questions used in the question naire were the same as those used for this purpose in the preliminary investigation That is on the one hand respondents were asked first to mark overall thermal comfort ventilation and air quality perceived during the complete test period on a ten point scale Second an assessment of the temperature and ventilation during various times of the day was requested Overall thermal comfort level was marked as 6 5 on a ten point scale while ambient air quality was marked as 2 6 on a five point scale From the second question
38. ant task is the diagnosis of system malfunctions and planning and surveillance of repair actions A Sauter system internally uses proprietary physical and software connections to transfer data between the central BMS and the substations local controllers Its communication scheme has severe bandwidth limitations Only changes of data exceeding a certain magnitude given rise to data traffic This scheme hinders flexible data collection as compared to polling and broadcast techniques with predefined frequencies especially during tuning the building SEBOS is the software supporting Smart functions i e feeding the building management system BMS with internet data such as weather and price forecasts and providing communication between users and the BMS ECN C 02 094 41 management system and in the process of calibration of sensors this leads to heavy data traffic Furthermore the building operator diminishes the accuracy of data collection in order to avoid data transfer problems For remote access to the data a communication channel of Sauter EY 2400 was used by means of an RS 232 connection Using this connection scheme the InsiteView building management system was connected with a modem for remote access or a null modem The Sauter system has a protocol language consisting of commands and predefined answer formats for retrieving data and setting values Geveke 1994 Central in this is the definition of a protocol or macro which
39. as 6 9 with a standard deviation of 0 9 scale 1 10 Question 5 Table B 7 First questionnaire answer question 5 Comfort aspect Times mentioned max 26 Temperature perception 25 Heat or cold radiated 2 Freshness of air 17 Moistness or dryness 18 Cold feet 8 Cold hands 10 Perspiring 11 Draft 20 Other aspects 7 The other aspects mentioned were e Light mentioned 4 times e Not enough oxygen e Dust e The smell of the air Question 6 Table B 8 First questionnaire answer question 6 Comfort aspect considered Times mentioned most important max 26 Temperature perception 10 Heat or cold radiated Freshness of air Moistness or dryness Cold feet Cold hands Perspiring Draft Other aspects of OrFOCOOrFrO Question 7 The average preferred working temperature is 19 9 degrees Celsius with a standard deviation of 0 8 degree ECN C 02 094 61 Question 8 Table B 1 First questionnaire answer question amp Comfort aspect Times mentioned Temperature 18 ventilation 18 Heat from the sun 14 Sun light 18 Artificial light 21 Other aspects 1 The other aspect mentioned was the possibility to direct light to individual working places Question 9 In the following table it is indicated how often the various temperature assessments were made Table B 10 First questionnaire answer question 9 Too Niceand Nottoo warm but Nice and Too cold cool also not too cold warm warm Temperature
40. bjective of the Smart project 42 ECN C 02 094 Centrally Controlled j Local DDC Air Temperature Tein Controlled Watertemperature T water el Convective Heating Unit Ambient on Air In T airln T airOut Figure 10 2 Convective heater model SEBOS expresses the control strategy by administering a dose of heat to a segment for every period in the simulation From the model a setpoint had to be generated for the PID controller of the ceiling unit given the measured central heating water temperature and the temperature of the pre heated air This was realised for one part by mapping the heat flow to a temperature dif ference and for another part by using the predicted model temperature 10 1 3 Comfort index calculation as a function of time over the day User comfort votes were collected using a WEB browser interface The preferences were stored in the vote file format as described in Kamphuis Warmer 2002 Each user had a votefile maintained by the InsiteView WEBServer This votefile retained for every user the vote direction and the voting time The SEBOS shell has access to the votefiles and to a user table in which information about which user resides in what building segment is contained During each SEBOS run all vote files of all users are processed Votes are valid if they are within the vote s latency period During the experiment this period was defined to be a week A vote for the tem perature to go up
41. bout aspects of comfort in the test situation as perceived by test users and about the ways in which test users practised and assessed comfort regulation before and after the introduction of Smart For investigating these aspects questionnaires personal interviews group interviews contextual interviews and participant observation were used The main outcomes of the study are the following e The building manager expected most of Smart like systems for optimising energy con sumption against energy price by trading in liberalised energy markets His expectations of Smart s potential for comfort management and the saving of energy in buildings with high inherent energy efficiency were low e We found no Smart related changes in the appreciation of comfort by office workers The test users reported no perceptions of improved comfort after the introduction of Smart e Initial interest in Smart among test users dropped quickly and over the complete test period Smart was used only modestly During the test period of six weeks 18 of the 23 test users voted 98 times e The test users turned to Smart only as a second best option for adjusting their comfort level Most test users preferred alternative routes over Smart for comfort management i e by ad justing ventilation or clothing e About 50 of the test users did not have a clear opinion on the usefulness of Smart As far as opinions on the usefulness of Smart were expressed these were mostly
42. complicated option to realise technically Inspired by a pilot study with a similar system the DUCOZT system see Oseland et al 1997 the team decided to realise option 3 by developing an individual voting system for Smart The voting system implied that every user in a thermal zone could enter his vote warmer colder within a voting period e g one hour while seeing the aggregated voting of other users at the moment of voting A user interface materialising this idea would look like this cf Oseland et al Previous requests in your zone warmer 63 no change 13 cooler 25 time to next vote 15 mins current time 10 15 6 2 Control trust and access Further user aspects relate to control trust and access Users should put trust in the system and have a feel to be in control of the system where needed The following minimum requirements for the design of the system and its interface were thought to apply e Transparency the system should make clear to users which actions they have to perform to achieve the outcomes they desire The user must be able to check the system over time on the effects of an instruction entered and to compare these with the comfort perceived in or der to be able to formulate the next instruction This means that the system must have forms of feedback toward users The most important feedback is the perception of a comfort change after voting for such a change For some functions feed forward ma
43. d on a physical de scription of the relevant climate dynamics within the building it manages These views enable the SEBOS optimiser to calculate effects of management actions in terms of realised comfort on the provisions that have an influence on the thermal parameters In the physical description not only local controllers but also centrally controlled parts of the installation are taken into ac count SEBOS further models the user in terms of position in the building and voting behaviour For obtaining data pertaining to the real comfort parameters and for giving out setpoints to parts of the comfort installation connections had to be made to the existing building management system BMS The connection scheme is depicted in Figure 10 1 The existing building man agement system was a Sauter EY 2400 system which consists of a central computer with local dedicated control units These control units receive their setpoints either from local settings on switch panels partly accessible by the user or from the central system The building manage ment system has a supervisory role and also helps in data acquisition for evaluating purposes Medium Sauler RS 232 Rata BMS network SingleChannel pany Intranet PC Client SEBOS Local BMS PC Server DDC EY 2400 InsiteView Role Data acq Sauter Protocol Remote amp Control Message exchange Procedure Calls Figure 10 1 Data transfer from DDC to SEBOS Another import
44. d test is to be preferred and test users must be better informed during the test By early involvement of users the designers of the Smart system can easier decide which options users are willing to trust to Smart and how the interface can best be designed Careful communication with the users about the system at the introduction and during the test will also help to make it more transparent for users where needed and thereby enhance its use and effectiveness 8 ECN C 02 094 Finally users were asked opinions about possibilities to improve the functioning of the Smart system The inclusion of additional options for comfort management in the Smart system the use of a pop up screen and the idea of grouping the users on the office floor according to their comfort preferences were assessed The assessment of the first two options was ambiguous while the use of the last option was clearly rejected because of being incompatible with work relations As the acceptance or rejection of such possible improvements is probably user de pendent this underlines the relevance of early user involvement and increased communication during the field test ECN C 02 094 9 1 INTRODUCTION Existing control systems for public and commercial buildings under utilise new technical op portunities that emerge from computer networks which become increasingly fine meshed Espe cially in organising access to distributed generation and calculation capacity the utilisation
45. des and actions How did users assess the Smart system in terms of visibility usefulness transparency trust Were there any misfits between the logic of the system and the logic of the users than should be repaired in the next version of the Smart system ECN C 02 094 13 4 PLAN METHODS AND INSTRUMENTS 4 1 Plan The study was carried out in two phases Phase 1 Preliminary investigation February 2002 This phase was carried out shortly before the Smart system was introduced and made available to test users The objective of this part of the study was twofold First it aimed to obtain infor mation that could be used during the next step of the investigation The second purpose was to create a reference situation i e to gain relevant information before the Smart system was made available to the test users to be compared later with the data collected at the end of the test Comfort indicators checked with users were explored as an aspect of use logic user assessment of the existing comfort was measured as well as attitudes relating to energy saving in the office and expectations concerning the Smart system Additionally in this first stage considerations about user representation have been developed and communicated with the designers of the Smart system These helped to shape those parts of the system that would be in immediate con tact with users Finally ideas were explored about possible motivations of users for saving en er
46. do so e Energy is only a small share of the costs faced by the employer Due to productivity loss dissatisfaction about comfort is a much more costly risk of increasing energy efficiency e Giving feedback i e confronting users with long term developments in energy consumption generally increases awareness which generates incentives for saving energy by changing behaviour This appeared from a campaign targeted at Dutch households in Dutch Zuinig stoken zuinig aan Goal setting may help consumers to maintain incentives over a longer period Here again the question arose whether the results found within households apply in a work environment Part of the experts low expectations about energy saving potential in the office stemmed from the realisation that comfort levels were already tuned to user needs by the Fanger criteria on which the building s central comfort management is based Complaints by office workers more or less forced a further fine tuning of the system to user needs During the session the building manager present indicated that for an operator the easiest way to satisfy complaints about cold is to raise the building temperature Those who are getting too warm open the window In this way an optimum comfort is realised which users will not give up easily he thought However this practice leads to an average temperature in ECN buildings close to 23 degrees C a tem perature considerably higher than most workers wer
47. e expected to have at home The experts agreed that in this temperature difference a considerable potential for saving energy is looming since a 1 degree decrease in room temperature implies a 10 saving of energy consumption Behavioural change supported by Smart could support an ECN wide policy setting targets for bringing down office temperature Smart could help in giving feedback to users about their contribution in catching these targets the experts argued This approach could be successful only if the employees could be convinced to participate and such participation would be vol untary The latter condition is difficult to realise in an open plan office setting It was decided that the participants in the Smart test would be consulted about such an initiative 22 ECN C 02 094 7 RESULTS OF THE PRELIMINARY INVESTIGATION Various research instruments were used to gain information on the situation in building 42 be fore Smart was introduced see Section 4 4 The information obtained can be divided into in formation on the following subjects e Interview with building operator Perceived comfort Use of options comfort regulation building 42 Expectations and opinions on Smart Attitude and conditions on energy saving 7 1 Interview with building operator One of the members of the department of ECN Building Operation amp Management FAC was interviewed Explaining the functioning of the current climate provisions and management
48. e for weekdays in January 2002 27 Figure 8 1 Smart user interface 30 Figure 8 2 Assessment of thermal comfort during the day 31 Figure 8 3 Assessment of ventilation level during the day 32 Figure 8 4 Use of Smart 33 Figure 8 5 Actions taken first for adjusting thermal comfort 33 Figure 8 6 Reasons given by test users for not using Smart 34 Figure 8 7 Effects as presupposed by test users to occur in situations in which several users within one segment enter comfort preferences into the Smart system within the same voting period 37 Figure 8 8 Assessment of the usefulness of Smart 38 Figure 10 1 Data transfer from DDC to SEBOS 41 Figure 10 2 Convective heater model 43 Figure 10 3 Modulation of the Fanger index as derived from the votes issued over time 44 Figure 10 4 Temperature north segment 45 Figure 10 5 Local controller setpoints on working days north segment winter period 46 Figure 10 6 Setpoints and temperature for north segment The lower lines indicate the sigma of averaging 46 Figure 10 7 Average central heating water temperature winter period 47 Figure 10 8 Temperature of in blown air 47 Figure 10 9 Voting behaviour north segment and average temperature 48 Figure 10 10 East segment comfort votes and average temperature 49 Figure 10 11 Managing to an optimal percentage of users satisfied 50 ECN C 02 094 5 ECN C 02 094 SUMMARY Agent technology based smart systems for energy m
49. e option Question 11 Table C 12 Second questionnaire answer question 11 Action Times mentioned Adjust clothing Open windows Temperature regulation by means of Smart or thermostat No answer given 8 8 2 4 ECN C 02 094 73 Question 12 Table C 13 Second questionnaire answer question 12 Reason not to use Smart Times mentioned Action taken is routine 7 I don t think of Smart when I want to regulate my temperature level 5 The action taken has a positive effect 9 The action taken also have other advantages 3 Question 13 Table C 14 Second questionnaire answer question 13 Actions taken Times mentioned No action taken because of positive assessment of the high temperature 0 No action taken due to low interest 5 First action clothes adjustment 6 Action taken afterwards temperature regulation by means of Smart First action temperature regulation by means of Smart 4 Action taken afterwards clothes adjustment First action complaint to building operator 4 No answer given 3 Question 14 Table C 15 Second questionnaire answer question 14 Link or short cut to Smart Times mentioned Yes 11 No 9 No answer given 2 Question 15 Table C 16 Second questionnaire answer question 15 Reasons Smart not linked to computer Times mentioned Not interested in Smart 2 Procedure to do so unknown 7 Other reasons 10 No answer given 3 Question 16 Table C 17 Second questionna
50. ed for the group interview possible participants have to be characterised The following questions are meant to make this characterisation possible 1 We would like to have an impression of you on certain areas Can you indicate if the follow ing prepositions apply Table B 3 Possible characterisations Preposition This applies to This doesn t This sometimes me apply tome applies to me I m technically interested generally speaking I m interested in computer possibilities generally speaking I like to use and explore new computer programs once in a while I m able to do my work with the computer but for the rest I m not very good with computers 2 Can you indicate which of the possibilities below does best apply to you l m normally not the one who is the first to wants to change the temperature within the room The temperature settings are thus not very important for me at most of the times I think I like the same temperature as most of the people in my environment Most people would agree with the temperature settings I would like to apply I m always one of the first to be cold or warm in a certain environment Therefore I would like to have a temperature setting of the system that is close to my temperature preferences I would like to work at different temperatures than most people in my environment ECN C 02 094 59 3 Did you ever complain to FAC building operation and maintenance abo
51. em in an office building In addition behaviour of the building and of the test users during the field test was mapped by physical measurements From the main findings of this investiga tion we draw the following conclusions Smart use by the test population was rather limited In all 18 out of the 23 test users who had access to Smart voted 98 times during the 1 5 month test period Mainly for this reason we did not see any specific routines of Smart use developing Rather test users preferred sticking to old routines of opening ventilation valves and adjusting clothing to remedy de creases in perceived comfort We estimate the fact that Smart had to compete with other options for comfort management deemed more effective by users and the relative invisibil ity of Smart no pop up screen was used to be important reasons for the failure of Smart to become an obligatory passage point for comfort management by users The current central heating regime in combination with the building s great mass resulted in decreasing comfort during the day Both the physical measurements the votes entered and the comfort perceptions reported point in the same direction i e the office space became too hot in the afternoon This situation was not improved by installing Smart Participants in the test did not perceive any change in comfort after Smart had been installed Neither did they detect any relation between voting entering preferences into the system and
52. emperature by means of Smart or the thermostat most users acted environmentally friendly by first adjusting their clothing or closing the ventilation valves before adjusting the temperature However when users decided to adjust the temperature by means of Smart they neglected the option energy efficient almost completely This non use can at least partly be explained by the incomprehensibility of this option as revealed by the questionnaire see Para graph 8 3 4 9 3 Expectations and opinions on Smart and the assessment of the Smart system The attitude towards the Smart system as measured in the first questionnaire was compared with the assessment of the effects and the usefulness of Smart as measured in the second question naire The initial attitude towards the Smart system was predominantly positive although quite a few people made conditional statements The assessment of the actual usefulness of Smart as tested in the second questionnaire was far less positive As indicated in Paragraph 8 3 6 most users did not have an opinion on the usefulness of Smart while most users who had an opinion judged Smart to be useless This assessment is in line with the lack of any perceived effect brought about by Smart 40 ECN C 02 094 10 TECHNICAL ASPECTS AND PHYSICAL MEASUREMENTS 10 1 Technical aspects 10 1 1 Connection of Smart to the existing building management system SEBOS has a number of views on managing the building which are base
53. ency is actively heated up only during a few hours in the early morning Shortly be fore arrival of the workers around 8 30 a m the central heating is switched off The building then continues to rise in temperature slowly by the presence of people switched on appliances computers electric light and sunlight 7 6 Main findings and conclusions Although the preliminary investigation was undertaken to acquire a reference for the future field test it yielded some salient findings which are worth to be summarised e The perceived overall comfort in the building before introduction of Smart was assessed as satisfactory on average ranked as almost 7 on a ten point scale Ambient air quality how ever was appreciated considerably lower ranked as 2 9 on a five point scale e Respondents ranking of comfort determinants for the building s climate is consistent with those used in Fanger s doctrine Lighting ranked high as an additional comfort indicator Most probably this had to do with the new building s lack of blinds Participant observation on the office floor during two summer days made clear that sunlight was an important com fort issue especially for workers sitting near windows which they blinded using sheets of paper and ECN flags some time after conclusion of the investigation sun blinds were in stalled e The ambient air temperature respondents indicate as the preferred one 19 0 8 degree Cel sius differs considerabl
54. enter the temperature they preferred to have in building 42 The preferred temperature mentioned was 19 9 degrees Celsius standard deviation 0 8 de grees This temperature was compared with the temperature measurements carried out by Building Operation amp Maintenance FAC In Figure 7 4 below the temperature curve recorded during weekdays in January 2002 is given 4 It is possible that some of the people answering No to this question did so because they deemed a positive answer not to be socially correct gt The curve is based on the average of temperatures measured by the sensors in the office room at about 1 5 meters above floor level communication with building operator 26 ECN C 02 094 Temperature degrees Celsius A a cea ee ca 22 20 18 16 J eenen dal obec ciotaantttacneeataedien ieee ee 10 T T T T T T T T T T T T T YF PF HF PF PM FM aA HM A HM A PM A S LP HP HP HM HD HD HP HP HP HP HP HP HM YH Time Figure 7 4 Average daily temperature at the test site for weekdays in January 2002 The average temperature during working hours 9 AM to 5 PM amounted to 22 5 degrees Cel sius a value considerably higher than the preferred average as indicated by test users in the questionnaire During a discussion about this outcome in the design team the FAC representa tive accounted for the curve by explaining that in practice the building because of its high en ergy effici
55. ere willing to do so In the following paragraphs the re sponses to these questions are given With respect to users willingness to save energy it became clear that e Most users found it appropriate that the responsibility to save energy is allocated to the us ers 92 in favour e About a fifth of the respondents were inclined to allow a decrease in their comfort level in order to save energy e Almost three quarters of the respondents considered a pre set energy saving goal as an in teresting challenge e About a third of the respondents were willing to adjust their clothing to enable energy to be saved e Most users were willing to join discussions with colleagues in order to save energy 85 in favour Concerning the conditions under which the users of building 42 were willing to save energy the following data were collected e Users wanted to be in control of the comfort being handed in for saving energy 88 in fa vour e Energy saving had to be voluntary 73 in favour Feedback was appreciated 46 of the users was interested in saving energy only if feed back was given about the quantity of energy saved e A reward for saving energy was not deemed necessary Only 8 of the respondents an swered to be interested in energy saving only if they would receive a reward 7 5 1 Temperature preferred by users and temperature as measured by Building Op eration amp Maintenance In the questionnaire users were asked to
56. ergy saving is marked No I don t have to know what happens when I check the option energy saving Yes I would like to know what Smart does when the option energy saving is checked 40 Would you like to know the energy saving effects of your actions No Smart can take care of energy saving Yes I would like to know how much energy I save Yes I would like to know how far my comfort level is decreased to enable energy saving All together What do you think of Smart 41 What is your opinion on the usefulness of Smart in view of your experiences with the sys tem Smart is not useful DeCauUse 0 ccc ccc cece risen e nee e eee n nee eee e senda E ee eben EEE Smartits not Useful DECAUSE oir anren cick cee 0d Ved a ania odo a bah 4b 6 Made E ENON NS coe u eb N J Smartis isel ss because 65 26 fc asco a en ais a waite cae Debates e tea bane ed eenes I don t know 42 Did in your opinion your comfort level increase when Smart was introduced No my comfort level decreased when Smart was introduced the office temperature is now too low for me lower than it was before Smart was introduced No my comfort level decreased when smart was introduced the office temperature is now too high for me higher than it was before smart was introduced No my comfort level is more or less the same as before the introduction of Smart Yes my comfort level increased when Smart was i
57. erybody votes for a lower temperature e More information will be appreciated e A lot of people complain about drafts 80 ECN C 02 094
58. especially when they perceive the climate as being too hot is opening the ventilation valves 8 3 3 Saving of energy By means of the questionnaire the following aspects of saving energy through Smart were in vestigated e Use of the energy saving option in the interface e Test users need for information about the energy saved While entering their comfort preferences by the Smart screen the test users could indicate that energy should be saved by ticking the box energy efficient The group interview made clear that not any of the interviewees ever used this option This outcome was affirmed by answers in the second questionnaire more than 80 of the users did not tick the option energy efficient while entering their comfort preferences A possible reason of this behaviour is given in Section 8 3 4 below Questions probing the need among test users for more information on energy saving through Smart either in general or more directed towards the effects of their own energy saving behav iour i e by ticking the box confirmed the results gained by the first questionnaire preliminary investigation By means of this first questionnaire 46 of the users made clear they were only interested in saving energy when feedback was given on the amount of energy saved In the second questionnaire about 50 of the users indicated that they would appreciate to receive 8 Within the test period the ventilation valves did not work during
59. evel Not enough Sufficient Too much it is almost like drafting When I enter the building in the morning I normally consider the ventilation as Round noon I normally consider the ventilation as At the end of the day I normally consider the ventilation as 12 Can you mark the air quality on your floor in building 42 You can use marks between 1 to 5 where 1 stands for mushy and 5 for fresh Possibilities to influence the inner climate in building 42 13 Do you make use of the thermostat in building 42 to adapt the temperature settings yes quite often yes sometimes no almost never 14 What happens when someone does use the thermostat to change the temperature settings No consultation takes place before the temperature settings are adapted consultation takes place Consultation takes place after the temperature settings are changed The consultation process ultimately leads to a compromise ECN C 02 094 57 15 If you never do so why don t you adapt the temperature settings by means of the thermo stat I don t find it is necessary I don t have time to care about the temperature settings I didn t know it was possible Because of the necessary consultation processes When I adapt the temperature setting the setting is often altered by someone else within a short period of time Setting the temperature is thus not very useful 16 Do you open or close the
60. f the system to en rol users can be dosed by varying the openness of its script 1 2 Open script users have to activate the system themselves Half open script the system offers it service to users automatically for instance with regular intervals the Smart interface pops up but users may switch it off unused Closed script the Smart interface pop up screen will only disappear after the user has en tered preferences 3 A pilot study with the smart DUCOZT system revealed that a majority of the end users 69 preferred option 2 Oseland et al 1997 ECN C 02 094 21 6 5 Energy efficiency To generate ideas about the design question how Smart could promote the saving of energy by changing behaviour of users a brainstorm session with internal experts was held The discus sion at this meeting revealed the following opportunities and constraints for an energy saving feature to be effective e Considerable saving of energy is delegated to the building itself i e the building is energy efficient by design see Jelsma 2001 This implies that its users have little opportunity for saving energy which may undermine efforts to do so e Jna utility environment the saving of energy does not reward users but the employer Users loose nothing by neglecting energy efficiency e Experience in the home environment demonstrates that support for extra behavioural effort to save energy is low only 10 of residents are motivated to
61. gy in the office through the Smart system These ideas were to be verified in the first phase Phase 2 User aspects of Smart field test April 25 June 11 2002 Phase two was carried out during the field test The objective of this part of the study was to produce data to answer the research questions i e e To measure frequency and ways of use of the Smart system e To measure effects on the practice of comfort management by the test users and on comfort experience To inquire on satisfaction of users with the system e To produce some insight in the underlying logic of the outcomes 4 2 Survey of research methods and instruments To tackle the research questions different types of research methods or instruments were needed each of which was used to answer certain aspects of the research questions In the sec tions below a survey of the research methods employed is presented In this survey it is ex plained which kind of data was collected by means of the methods described The chapter ends with an overview of the research scheme describing the order in which the instruments were ap plied 4 2 1 Questionnaires Questionnaires are written lists of fixed sets of questions with pre determined response options Questionnaires can be used to collect data among a larger group of people The data to be col lected must be concrete and easy to define Further questionnaires lend themselves for measur ing difference such as difference in c
62. he building manager in order to reach a acceptable temperature on Monday morning starts warming short after midnight The other days warming is started at 5 30 ECN C 02 094 45 23 22 21 20 TNorth Workingdays SPNorthWorking Figure 10 5 Local controller setpoints on working days north segment winter period C Week pattern Nov March 24 22 20 Day number Thu 0 SP North SSP North T North Sig Tnoord Figure 10 6 Setpoints and temperature for north segment The lower lines indicate the sigma of averaging Figure 10 6 shows the impact of actions of users on comfort by setting the wall thermostats to be very limited It is hardly possible for users to have any impact on the momentous comfort level through setting the room thermostats Reacting to a low temperature comfort situation early in the morning by increasing the setpoint only leads to a further overheated temperature comfort situation in the afternoon In Figure 10 7 and figure 10 8 the comfort actuator parameters TWater and TcIn which determine the heat dosing mechanism see Figure 10 2 of a climate zone are shown The decrease at 6 00 and increase at 18 00 hour correspond to the 46 ECN C 02 094 switch on off points of the central ventilation system which is only operational during working hours C TcAir 55 35 30 1
63. he material presented we draw the following conclusions In the present test environment neither the building and installation architecture nor the SMART design appear to offer the building users good opportunities to influence thermal com fort in their climate zone The intrinsic inertia of the building due to its physical parameters and comfort provisions hinders fast feedback on user control actions Further we expect SMART to deliver a better performance if the time horizon over which the comfort level is optimised on the basis of data on the building physics and environmental parameters is extended beyond the limit 24 hours which was practised in the present field test From the temperature measurements it follows that there is energy saving potential looming in the difference between the preferred temperature as indicated by users in the questionnaire compared to the real office temperature The difference amounts to approximately 2 degrees which could result in a saving of up to 20 in heating cost Our measurements seem to support the expert group conclusion that building managers tend to react to complaints about temperatures in a building being too low by adjusting the building s heating curve upwardly In Fanger terms the voting procedure suggests that if the building is managed using a Smart system towards the target at the other side of the bell shaped distribu tion as presented in Figure 10 11 the percentage of people dissatisf
64. he test users For this reason the comprehensibility of the Smart interface and the following aspects of the functioning of Smart and the possible effects thereof were assessed through the question naire The interface and its concepts The interface was assessed as user friendly by most of the respondents 14 of the 16 respondents who answered the question However some of the concepts used in the interface appeared to be unclear to parts of the test population Table 8 1 shows the shares of the test population who understood the concepts concerned Respondents who did not answer the question were not in cluded in the calculation As we found out about half of the test users did this despite the fact that this was not recommended in the introductory mail The presence of short cuts indicated that at least those users who made them had the intention to use Smart ECN C 02 094 35 Table 8 1 Percentage of respondents that completely understood the concepts used in the Smart interface Concept Measured percentage of people indicating that they completely understood the concept used Segment 94 Warmer 94 Unaltered 89 Colder 94 Energy saving 50 Requests in the present voting period 78 Tendency 56 Annul 89 OK 94 The table demonstrates that especially the meaning of the concepts energy efficient requests in the present voting period and tendency were unclear to at least some of the inhabitants of buildi
65. hermal comfort ISO 7730 1994 see www annova dk Geveke Electronics 1994 EY2400 Building management system technical manual and protocol description Gebouwbeheerssysteem technische handleiding en protocolbeschrijving in Dutch Jelsma J 1999 Domestic energy use from better behaviour to a better design an onset to an integral approach Huishoudelijk energiegebruik van beter gedrag naar beter ontwerpen een aanzet tot een integrale benadering in Dutch Novem Jelsma J 2001 Smart work package 4 1 final report the Smart system and its test building matching design logics ECN report ECN C 02 008 Petten NL McCalley L T C J H Midden 1998 Computer based systems in household appliances The study of eco feedback as a tool for increasing conservation behaviour Research report Faculty of Technology and Management Technical University of Eindhoven Eindhoven I G Kamphuis 2000 SMART Smart Building Multi Agent InterNet Technologie in Dutch Project plan August 2000 Kamphuis I G C Warmer H Akkermans 2001 Smart Innovative e Services for Smart Buildings paper for ISPLC Conference Birmingham Kamphuis I G C J Warmer 2002 SEBOS A Smart Enhanced scope Building optimiser shell User manual ECN report ECN CX 02 103 Petten NL Morgan D L R A Krueger 1997 The focus group guidebook Part 1 SAGE publications London Oseland N A R Fargus D K Brown A D Charnier A J Leam
66. hot and not enough ventilation Question 20 Table B 19 First questionnaire answer question 20 Assessment of the idea that Smart enables participation in comfort management Times mentioned Positive assessment Negative assessment Conditional assessment No assessment M 0 W O Question 21 Table B 20 First questionnaire answer question 21 Most appropriate preposition Times mentioned Energy saving doesn t concern employees Energy saving by employees is appropriate 2 24 Question 22 Table B 21 First questionnaire answer question 22 Most appropriate preposition Times mentioned Comfort to be surrendered is to be determined by me 23 Comfort to be surrendered may be determined by the building operators 3 Question 23 Table B 22 First questionnaire answer question 23 Preposition Percentage that agreed with the preposition Energy saving has to be voluntary 73 Energy saving at work may decrease my comfort 19 I m only interested in energy savings when I get feedback 46 I consider achieving a pre set energy saving goal as an 69 interesting challenge I m only interested in energy savings when I get rewarded 8 I m willing to wear warmer clothes such that the central 35 heating can be turned down I m willing to join discussions with colleagues in order to 85 save more energy as a group 64 ECN C 02 094 APPENDIX C SECOND QUESTIONNAIRE In this appe
67. ied remains unchanged This is especially true in view of the voting behaviour during the day and maintaining a satisfactory comfort level within a scope of a day We expect that lowering the heating curve of the building will deliver better comfort during the whole day but especially during the afternoon hours ECN C 02 094 49 SMART Target Cost effective Target Percentage Satisfied BMS Target ComfortIndex Figure 10 11 Managing to an optimal percentage of users satisfied The data presented above confirm the findings of the design phase study Involving users in the management of their comfort requires a short feedback cycle i e users should see systems to react to their preferences within a short time This should be reflected in the response of comfort aspect controllers The voting procedure as applied here which is also used in other cases Oseland 1997 yields consistent results with respect to the comfort perceptions in different climate zones The trends in the perceptions measured in this way also coincide with overall trend as measured by the questionnaires see Chapter 8 Even day type specific trends are reproducible 50 ECN C 02 094 11 CONCLUSIONS AND RECOMMENDATIONS We investigated the use of possibilities for comfort management and compared the comfort situation as perceived by test users before and after the introduction of a prototype of the Smart syst
68. ire answer question 16 Use of Smart to regulate comfort level Times mentioned Yes a few times a day Yes a few times a week No almost never No answer given 2 7 10 3 Question 17 Table C 18 Second questionnaire answer question 17 74 ECN C 02 094 Effect noticed Times mentioned Never uses Smart so no effect noticed Although Smart is used no effect noticed Complaints disappear gradually when Smart is used Effect is clearly noticed No answer given Ww OwWon7 Question 18 Table C 19 Second questionnaire answer question 18 Consultation process when Smart is used Times mentioned Without consultation With consultation beforehand No answer given 11 6 3 Question 19 Table C 20 Second questionnaire answer question 19 Assumed reason consultation process Times mentioned Check own feeling To ascertain that others are not disadvantaged To persuade others to enter same temperature setting Other reason No answer given 2 3 1 0 16 Question 20 Table C 21 Second questionnaire answers question 20 Reason not to use Smart Times mentioned Not necessary no time Unknown Smart not on top of head when temperature I regulated Consultation process Expectation that smart has less effect than possible actions Other reasons No answer given NYNNODODOCWN Question 21 Table C 22 Second questionnaire answer question 21 Freque
69. ithin one segment enter their comfort preferences into the Smart system No I think so but I m not sure I m quite sure what Smart does when more people enter their comfort preferences into the Smart system 33 What do you think what Smart does when several comfort preferences are entered 34 What is the effect of not being sure of what happens when more than one comfort prefer ence is entered within one segment This has no effect I don t use Smart as often if I would do otherwise The way I enter my comfort preferences becomes more extreme because of this uncertainty E g I enter my comfort preferences more frequently This has another effect viZ scons sacsssaveloi does ranae rt EK DIEE NO EE ERE A E 35 Do you have further remarks on the way comfort preferences can be entered into the Smart system 70 ECN C 02 094 Energy saving 36 Do you use the option energy saving when you enter your comfort preferences into the Smart system i Yes DE CAUSE Bho Setesesani lied sare ccudieatajaea ards a a JNO bDeCause s ere Ae see Rei Sometimes it depends oNn ccc cece cence eens 37 Do you have an idea what Smart does when the option energy saving is checked Yes No 38 What do you think are the actions taken by Smart when the option energy saving is checked 39 Would you like to receive more information on what Smart does when the option en
70. know it was possible I don t think of mart at the moment want to regulate my comfort settings Because of the necessary consultation process Because of my expectation that voting by Smart has effect on my comfort level Other reasons Vizio eee ee tid ees 21 Do you close or open of the small windows more or less often since the introduction of Smart Yes I open or close the windows more often since Smart is introduced No Yes I open or close the windows less often since Smart is introduced 22 Are there periods of time when it is not possible to regulate the temperature on the office floor sufficiently If applicable please detail your answer which time of the day which kind of weather is it too cold or too warm etc Regulation possibilities 23 How would you like the idea of Smart manifesting itself by means of a pop up screen Convenient this way I don t forget to vote Convenient but I want to be the one to decide how often the screen appears on my computer Inconvenient it will braek my concentration Inconvenient because 0 0 cece cece cece eect eee eeeenee 24 How often would you like the pop up screen to appear Never Once a week Once a day Repeatedly during the day e g in the early morning and after the lunch Every time I start up my computer 25 When would you like the screen to disappear Independent of voting behaviour De
71. ld regulate their comfort level by means of the Smart system and by opening and closing the ventilation valves personal communication with build ing operator Further they could set the room thermostats though it is unclear whether these were functioning during the Smart field test Finally test users could regulate their level of thermal comfort by adjusting their clothing How and how often the various regulation possi bilities were actually used was measured by means of various questions in the second question naire First we checked how frequently the various regulation possibilities were used during the Smart field test Secondly we asked questions about the preferences of users regulating their comfort level It appeared that most respondents made little use Smart for regulating their comfort level When they did use Smart they mostly used it individually that is according to the majority of respondents 64 no consultation of colleagues in their thermal zone took place In Figure 8 4 it is depicted how often the respondents made use of Smart Respondents who skipped the ques tion are included in the figure For precise data about voting behaviour see the table in Chapter 10 7 This percentage leaves room for the conjecture that a considerable part of the test users did consult colleagues before voting At least during the group interview consultation was said to be common practice 32 ECN C 02 094 Times mentioned 12
72. mart However the present study confirms earlier ones in showing that such total delegation may violate the user s preference for exerting control over his own comfort User involvement in design may also deliver a user interface with optimal clarity The transparency the system should have is a topic that should be discussed with users Early communication with users may lessen or prevent negative effects of certain possibly in comprehensible aspects of the system as showed up too late in the present field test Such pre ventive measures should be complemented with a careful introduction of the system to the user To enhance the visibility of the Smart system the use of a pop up screen should be considered In this as well as in another field test see Oseland et al 1997 test users were found to be in fa vour of such a reminder 52 ECN C 02 094 REFERENCES Beyer H K Holtzblatt 1998 Contextual design defining consumer centered research Morgan Kaufmann Publishers San Francisco Boertjes E H Akkermans R Gustavsson R Kamphuis 2000 Agents to achieve customer satisfaction The COMFY comfort management system paper submitted to PAAM 2000 1 12 April Manchester Fanger P O 1970 Thermal comfort Analysis and applications in environmental engineering McGraw Hill Book Company New York Fanger P O 1996 Moderate thermal environments Determination of the PMV and PPD indices and specification of conditions for t
73. mation about energy saving possibilities Smart Times mentioned 78 ECN C 02 094 No no interest Yes Not answered Question 40 Table C 36 Second questionnaire answer question 40 Interest in energy saving effect of own saving behaviour Times mentioned Not interested Interested in amount of energy saved Interested in information on decrease comfort Not answered WA NN CO Question 41 Table C 37 Second questionnaire answer question 41 Assessment Smart Times mentioned Useful Not useful Useless No opinion Not answered ROW hNW Question 42 Table C 38 Second questionnaire answer question 42 Opinion on change in comfort level since the introduction of Smart Times mentioned Decrease too cold Decrease too warm No change Increase Not answered 0 1 17 0 4 Question 43 Table C 39 Second questionnaire answer question 43 Opinion on subdivision group according to comfort preferences Times mentioned Good idea No opinion Not a good idea Not answered 1 3 15 3 Question 44 Table C 40 Second questionnaire answer question 44 Interest in more information on functioning of Smart Times mentioned Interested Not interested Interest depending on the kind of information offered Not answered 2 6 11 3 ECN C 02 094 79 Question 45 Suggestions offered e It will be nice if the building will be cooled when ev
74. me please tick one Energy saving doesn t concern employees The responsibility for energy savings lies with ECN Energy savings by employees is appropriate for ECN 58 ECN C 02 094 22 The following proposition on the area of energy savings is the most appropriate for me please tick one I want to determine myself how much comfort I surrender in order to save energy The amount of comfort surrendered in order to save energy may be determined by building operation 23 Can you indicate if you agree with the following statements please circle the right answer Energy savings at work have to be voluntary I agree I don t agree Energy savings at work may have to mean that my comfort is decreased I agree I don t agree I m only interested in saving energy if I get feedback on the amount of energy I save I agree I don t agree I consider achieving a pre set energy saving goal as an interesting challenge I agree I don t agree I m only interested in energy savings if I get rewarded for my efforts e g extra leave or a surprise present I agree I don t agree I m willing to wear warmer clothes at my work such that the central heating can be turned down and energy is saved I agree I don t agree I may be willing to join discussions with colleagues in order to save more energy as a group I agree I don t agree B 2 Questionnaire Part 2 Explanation In order to be able to choose participants to be invit
75. mfort management existing beside Smart ii the invisibility of Smart and iii the influence votes by others were supposed to have on the comfort change i e the effect of voting was unclear The following opinions on improvements of Smart that might counteract these weaknesses were derived from the group interview and the questionnaire Test users participating in the group interview hailed changing the system such that Smart in stead of the users would control the ventilation valves as an improvement The questionnaire did not confirm this outcome however A majority of the respondents 52 agreed with the cur rent situation i e the users should continue to operate the valves In their opinion Smart should only regulate the ambient air temperature 31 of the respondents disagreed and found that the small windows could best be operated by Smart In that case the Smart system should decide when to open or close the valves in order to reach the preferred comfort level To increase the visibility of Smart we tested opinions on installing a pop up screen for Smart About half of the respondents indicated that they preferred this solution As preferred by most such a screen should appear repeatedly during the day and should disappear by clicking it away i e instead of disappearing only if a vote had been entered On the office floor the employees desks were clustered according to the kind of work activities they perform In one of the i
76. n which the Smart interface could be found By means of this interface see below indi vidual test users could enter their thermal comfort preferences into the system Users could do so only once during a certain time period This time lasted for about one hour and was called a voting period Individual comfort preferences entered into the system were represented by soft ware agents which negotiated on the basis of an algorithm for each voting period the preferred temperature to be set within each thermal zone for technical details see Chapter 10 Only at the end of the test it appeared that a few workers at the test site had not succeeded to get access to the Smart system about 5 of the 22 respondents of the second questionnaire Some of these people had not received the initial email while others who did reply to the email still had no access to Smart One employee was located by the Smart system in the wrong thermal zone Consequently about 18 workers could participate in the Smart test as initially intended 8 2 User interface The user interface of the Smart system is shown in Figure 8 1 below It functioned as a dialogue box for communication between the user and the Smart system At the user side preferences could be entered into the Smart system From the system s side the interface supplied the user with information about voting and comfort trends in the zone of interest ECN C 02 094 29 v one Ey pasi r P Ste x c X
77. ncy opening closing windows in relation to the situation before the introduction of Smart Times mentioned Higher frequency Same frequency Lower frequency Not answered ECN C 02 094 75 Question 22 Periods of time when it is not possible to regulate the office temperature sufficiently The following time periods were mentioned afternoons too warm when the weather is sunny mornings too cold dependent on the weather mornings too warm after the lunch too warm Question 23 Table C 23 Second questionnaire answer question 23 Assessment pop up screen Times mentioned Convenient reminder to vote 6 Convenient when regulation of the time the pop up screen 7 appears can be regulated Inconvenient concentration 3 Inconvenient other reason 0 Not answered 6 Question 24 Table C 24 Second questionnaire answer question 24 Preferred frequency pop up screen Times mentioned Never 2 Once a week Once a day Repeatedly during the day During start up computer Not answered aownnNnoe Question 25 Table C 25 Second questionnaire answer question 25 Preferred disappearing behaviour pop up screen Times mentioned Independent of voting behaviour 12 Dependent on voting behaviour 4 Not answered 6 Question 26 Table C 26 Second questionnaire answer question 26 Preferred regulation mechanism Smart Times mentioned Smart by opening Smart page by yourself Smart by pop
78. ndix the second questionnaire presented to the users of building 42 is presented The questionnaire was originally drafted in Dutch First the questionnaire is presented After the questionnaire the answers given to the questions in the questionnaire are presented C 1 Second questionnaire Questions Personal information 1 How old are you 18 25 25 35 35 50 50 2 What is your gender Male Female 3 In the next picture the office you work in is sketched Can you give an indication of the place where you work in this office I work in part of the office Hall 4 Did you take part in the group interview on Smart yes no 5 How would you mark your comfort situation during the last six weeks You can use marks from 1 and 10 where 1 stands for awful and 10 for perfect ECN C 02 094 65 Inner climate building 42 6 What do you normally think of the temperature on the floor you work on Can you please tick the appropriate box below for the situation after the introduction of Smart Table C 1 Possible assessments temperature within building 42 Situation Too Nice and Not too warm but Nice and Too cold cool also nottoocold warm warm When I enter the building in the morning I normally consider the temperature as Round noon I normally consider the temperature as At the end of the day
79. negative A number of Smart independent conditions characteristic for the test situation seem to be re sponsible for these findings 1 Smart was designed for usage in an individual environment but it was applied in one large room in which all the users were located Though this discrepancy was at least partly solved by establishing five comfort zones within the room the comfort delivered could not be fine tuned to individual preferences of users Moreover the equipment used to control the tem perature within each zone small heating units in the ceiling was not designed for this pur pose and was not very effective ECN C 02 094 7 2 Only during the test it became clear that because of the high energy efficiency of the building active heating was needed only during a short period in the early morning before the office workers arrived During the rest of the day the building was passively heating up by heat accumulation due to presence of people and operating appliances There were no provisions for cooling that could be activated by using Smart Thus because of the inertia of the building the practical possibilities to control comfort by Smart were very limited Therefore it is understandable that the test users perceived the effectiveness of Smart as low 3 Test users could adjust comfort by means that were independent of the Smart system i e by manually operating ventilation valves and by adjusting clothing In the logic of use
80. ng 42 while further indicating that energy efficient caused the largest trouble to be un derstood by far Functioning of Smart The group interview showed further that test users were puzzled not only about the way the op tion energy efficiency might work but also were wondering what happened when several us ers within the same segment enter different comfort preferences into Smart within a certain voting period By taking up questions in the questionnaire we checked both observations Thus respondents were asked first whether they understood what happened when they ticked the box energy efficient A large majority of the respondents 76 gave a negative answer to this question Second the question was asked whether it was clear to respondents what happened when more than one person voted during a certain voting period in a certain sector 27 of the respondents who answered this question indicated to be quite sure about what happened when more votes were entered into the Smart system The share of the test population stating either to be unsure or to have no idea about what happened under this circumstance was 38 and 33 respectively Effects of incomprehensibilities Anticipating aspects of Smart that might be unclear to the users we incorporated in the ques tionnaire questions about possible effects of this incomprehensibility When asked how they re acted to such incomprehensibilities most respondents answered that they
81. nology of smart systems for comfort management in office buildings works in principle and holds if properly implemented a promise for considerable energy savings In contrast to initial low expectations of experts about opportunities for further improving the inherent energy effi ciency of the building in which the Smart test was carried out the test showed that in the dem onstrated discrepancy between real and preferred temperatures a considerable potential for low ering energy consumption is looming If included in building design from the beginning Smart systems can help reaping this fruit by realising a more sophisticated building management over day and night on the basis of a larger data set than was available in this limited field experiment For further experiments the following recommendations apply Smart can best be applied in a building with individual or almost individual compartments This way the future users of Smart can adapt their comfort level individually as originally intended by the designers of the Smart system If possible the users should be involved earlier in the design of the Smart field test and better be informed during the test Early involvement of users can help designers of Smart in making more informed decisions about the division of control options between Smart and the user From a functional perspective it seems most desirable to bring all building provisions for com fort management under the authority of S
82. ntroduced because 0ceeeees 43 Smart could be used to group people according to their comfort preferences What do you think of this idea I think it is a good 1dea Viz nenea o e ene ene cence ene cede eee e eae e ease setae ennnenaes I don t have an opinion on this issue I don t like the idea DeCaUse 0 ccc ccc cece eee eee nn nen e enn Eee eee e EEE EEE Eee EEE ES 44 Would you like to receive some more information on Smart Yes No Maybe it depends on the kind of information ECN C 02 094 71 45 Do you have suggestions that can be used to improve the functioning of Smart C 2 Second questionnaire Answers The questionnaire was answered by 22 persons It was striking that not all questions were an swered by every person The reason for this was mostly that not all persons had access to Smart in the preceding period Question 1 Table C 4 Second questionnaire answer question 1 Age Times mentioned 18 25 1 25 35 3 35 50 8 50 10 Question 2 Table C 5 Second questionnaire answer question 2 Gender Times mentioned Male 11 Female 11 Question 3 Table C 6 Second questionnaire answer question 3 Working place Times mentioned Part 1 7 Part 2 8 Part 3 1 Part 4 5 Between different parts 1 Question 4 Table C 7 Second questionnaire answers question 4 Participation in the Times mentioned group interview Yes 8 No 14
83. o kinds of reasons not to use the thermostat could be dis 24 ECN C 02 094 tinguished The one reason was ignorance about half of the users did not know the thermostat existed The other half deemed adjusting the temperature unnecessary or did not take time to set the thermostat About a third of the respondents 8 reported that they operated the ventilation valves regularly About half of the respondents indicated that consultation about opening or closing took place beforehand while about a third of the respondents indicated that no consultation took place Seven people indicated that temperature control was dependent on time Especially after a cold night the temperature was judged to be too low and too high during the afternoon An individual possibility to improve one s comfort level is adjustment of clothing to the envi ronmental temperature Fanger s clothing factor Whether the office workers practised this op tion was explored by asking which kind of clothing they normally wore at work compared with clothing worn at home A majority 77 of the respondents indicated they wore the same clothes at work as they did at home However some respondents did adjust clothing in relation with their work situation Two people wore thinner clothes at work than they did at home and four people took off part of their clothing e g a jacket or cardigan during the day This out come corresponds with the outcomes presented above indicating tha
84. of user and manager operator coincide In an office building these roles are separated Therefore the Smart system had to deal with two types of users e Building managers operating the building management system e End users instructing the building system through Smart Within the class of end users further heterogeneity is to be expected A relevant discrepancy that may appear is a differ ence in interest for experimenting with the system Gender effects could occur here The distribution of authority and competence between both user types had to be inscribed in the software of the Smart system so it needed attention to be given in the design process The dis tribution inscribed had to reckon with the authorities responsibilities and interests of both user groups In principle by using Smart end users can influence i aspects of comfort such as tem perature humidity and ventilation ii energy used for this comfort to be supplied and iii the share of green energy in the total energy consumption by managing the solar energy generated by the roof covering the hall of the building However the design team decided that building managers should control all cost and energy trading aspects of the building management Saving costs by trading energy should not be a burden of the end user in the office The Smart system was rather to be integrated in the existing management logic of building operators i e setting bounds to the preference
85. omfort perceptions among respondents and to calculate correlation between categories of answers if the group of respondents is large enough Less fo cused data such as insights ideas and considerations and especially the reasoning behind are hard to obtain by means of a questionnaire In this study questionnaires were used to collect and The alternative working with a control group was crossed out because the group of test users was too small to be split and the other floors in the new building were still empty at the time we planned the study 14 ECN C 02 094 to compare concrete data among users before and after the introduction of the Smart system No correlation of data was practised because of the small group size 4 2 2 Personal interviews Personal interviews deliver qualitative data resulting from sessions in which an interviewer in terrogates an interviewee This technique is suitable to acquire detailed and unique data on indi vidual perceptions considerations and experiences but because of workload only from a small group of persons In our study the personal interview was used to acquire information in only one case 4 2 3 Group interviews The group interviews we applied were a weak version of focus groups Using focus groups is a qualitative research method based on discussions in small groups seven to ten persons initiated and led by a moderator to collect data on the topic in which the researcher is interested Foc
86. on value may be undermined by use practice i e by interaction processes between building managers and their clients resulting in tempera tures going up The main recommendations derived from the study are 1 Smart like systems should be tested and applied in situations where they can make a differ ence In the first place this implies that at the test site all parameters needed to control com fort can be effectively manipulated and monitored by flexible equipment Second if Smart should be an obligatory passage point for users a building provisions relevant for comfort management should be under the control of Smart and only accessible through Smart A pop up screen reminding users of occasions for voting can further support the enrolment of users However the logic of Smart should not overrule the demonstrated desire of users to keep control over their own comfort On the other hand this desire is expected to weaken as Smart better satisfies individual comfort preferences and succeeds in building trust among its users 2 The spatial layout of the test site should be such that deliverance of individual comfort can be guaranteed In an open office space these conditions are hard to realise When we include also the previous recommendation this implies that the design logic of the building has to converge with the design logic of the mode of climate control cf outcomes of WP 4 1 3 Early user involvement in the design of the Smart fiel
87. ot understood 9 Did vote although the consequences of voting behaviour were unclear 0 Not answered 9 ECN C 02 094 77 Question 32 Table C 31 Second questionnaire answer question 32 Answers given on understanding what happens Times mentioned when more people vote within one segment No 6 Yes but I m not sure 7 Yes I m quite sure 5 No answer given 4 Question 33 Ideas on what happens when several comfort preferences are entered within one segment Most people indicated that the average comfort preference will be calculated Question 34 Table C 32 Second questionnaire answer question 34 Effect of not completely understanding when more Times mentioned comfort preferences are entered No effect Smart is used less Comfort preferences are entered more extremely Other effect Not answered SR RAND Question 35 e Aclearer background will be appreciated Question 36 Table C 33 Second questionnaire answer question 36 Use of energy saving option Times mentioned Yes 3 No 15 It depends 0 Not answered 4 Question 37 Table C 34 Second questionnaire answer question 37 Idea what happens when the option energy saving is applied Times mentioned Yes 4 No 13 No answer given 5 Question 38 Most people didn t have an idea what happened when the option energy saving was marked Question 39 Table C 35 Second questionnaire answer question 39 Interest in more infor
88. our for individual days Most votes especially in the afternoon hours indicate that the user s perception is too warm The Monday pattern only indicated a strong tendency in the voting behaviour toward an increase in temperature in the morning The afternoon peak in votes for colder appears to be delayed on Monday Comfort vote preferences north segment 0 2 23 0 1 22 8 0 r 22 6 17 18 c 0 1 22 4 Ss 5 g g 5 0 2 22 2 5 2 S 2 a 22 amp g lt 0 4 21 8 3 0 5 7 21 6 1 ra p 0 6 21 4 0 7 21 2 Time Thu Fri Mon Tue Wed Average T Average Figure 10 9 Voting behaviour north segment and average temperature For the east segment with fewer voters a comparable trend can be observed see Figure 10 10 The Monday pattern with its characteristic morning too cold peak and its shift of the too warm peak in the afternoon can be seen as well The overall voting behaviour is consistent with the responses of users to the questionnaires i e the ambient air temperature in the afternoon is perceived as too high 48 ECN C 02 094 Comfort preferences East segment 23 5 23 22 5 Average Vote 22 21 5 0 4 21 Time Thue Fri Mon Tue Wed Average Temperature Figure 10 10 East segment comfort votes and average temperature Conclusions From t
89. parts The first part was anonymous and was meant to obtain information on among other things comfort perceptions regulation possibilities and atti tudes in relation to energy saving measures Users were asked to fill in their name and some personal questions in the second part of the questionnaire this part was separated from the first part of the questionnaire The answers given in this second part were used for formulation of the group that was asked to participate in the group interview after the introduction of Smart The outcome of group interviews is strongly dependent on the composition of the group inter viewed according to among others Morgan and Krueger 1997 B 1 1 Questionnaire Part 1 Personal information 1 How old are you 18 25 25 35 35 50 50 2 What is your gender male female 3 In the next picture the office you work in is sketched Can you give an indication of the place where you work in this office I work in part of the office Hall ECN C 02 094 55 4 How would you mark your present comfort situation You can use marks from 1 and 10 where 1 stands for awful and 10 for perfect 5 Which aspects would you consider if you have to indicate if you find your working place comfortable my perception of the temperature of the air in the building the heat or cold radiated by the r
90. pendent of voting behaviour with other words you have to vote to make the screen disap pear 26 If you would have to choose how would you like to use Smart I would like to evoke the Smart page myself Would like Smart to appear as a pop up screen I don t care how Smart can be used Regulation possibilities 2 On this moment Smart can only be used to regulate the temperature in your sector of the office Which aspects of the inner climate would you like to be regulated by Smart I would like Smart to regulate the temperature I want to be able to open or close the small windows myself I would like Smart to regulate the temperature and the small windows I would like that peo ple can tell Smart if they want the windows to be opened or closed so that Smart can decide to open or close these windows I would like Smart to regulate the temperature and the small windows I would like people to be able to enter temperature settings into Smart so that Smart can decide to open or close the windows in order to reach the preferred temperature 68 ECN C 02 094 28 What is the reason you chose one of the possibilities while answering the last question amp ae f A 2 gi i i Figure C 1 The smart interface 29 How user friendly is this screen to you User friendly it is completely clear what you have to do to enter your preferred settings Quite user friendly Not user f
91. ply that the design logic of the building made opportuni ties for the test users to save energy through the Smart system questionable i e it might make the implementation of the energy saving option in Smart more difficult One could say that in this respect the design logic of Smart was at odds with the design logic of the building see the outcome of Smart WP 4 1 Jelsma 2001 The fact that Smart a tool for individual comfort management had to prove itself in a shared environment an open plan office floor added to ECN C 02 094 11 this mismatch of logics This mismatch had implications for the design of the Smart system as well as for the study of the user aspects see below 12 ECN C 02 094 3 RESEARCH QUESTIONS Taking into account the foregoing approach the following questions were leading the study of the user aspects of the Smart test e How were users represented by designers which assumptions did the designers make about the logic preferences views attitudes values of users and how were these translated into the design of the Smart system e How did interactions develop between the test users and the provisions for comfort man agement Smart system and building provisions in building 42 1 Did any new practice or routine develop and how was it being shaped How did the use of Smart relate to the use of other provisions for comfort management e Which forms of user logic underlie relevant user preferences attitu
92. r r 1 0 6 12 18 24 Time hrs e Thu Fri a Sat x Sun x Mon e Tue Wed Average Figure 10 7 Average central heating water temperature winter period Temperature of inblown air C 23 22 5 22 21 5 21 20 5 20 19 5 r 1 1 1 Time hrs Thu Fri a Sat x Sun x Mon e Tue Wed Average Figure 10 8 Temperature of in blown air Increasing the water temperature of the central heating and switching on the ventilation system later in order to preheat the climate zones with accumulated air from the mechanical ventilation would be possible steps to increase the possibilities for more instantaneous comfort control In addition the temperature of the air blown in could be decreased during the afternoon An ex tended Smart system in combination with a leading edge BMS can optimise these parameters in a more sophisticated way than is possible with the current configuration ECN C 02 094 47 10 2 2 Voting behaviour The voting behaviour of the population in the different climate zones is shown in the following table Table 10 1 Distribution of votes per climate zone during the test period Segment Possible Voters Actual Voters Votes TooWarm Equal Too Cold West 2 2 2 0 1 1 Oost 7 5 18 13 1 4 North West 2 2 10 3 2 5 North East 5 2 5 3 0 2 North 7 7 63 38 10 15 The trend in the votes collected is depicted in Figure 10 9 The dashed lines denote the voting behavi
93. riendly at all I still have problems trying to understand how best to enter my pre ferred comfort settings ECN C 02 094 69 30 Can you indicate if you understand the concepts used on the Smart screen Can you encircle your choice in the table below You can choose the following possibilities e Yes I understand the concept e partially Ihave an idea what is meant by this concept e No I don t understand the concept Table C 3 Understanding concepts used in Smart Concept I understood the concept right I now understand after the introduction of Smart the concept segment Yes partially no Yes partially no warmer Yes partially no Yes partially no unaltered Yes partially no Yes partially no colder Yes partially no Yes partially no Energy saving Yes partially no Yes partially no Requests in the present voting period Yes partially no Yes partially no Tendency Yes partially no Yes partially no Annul Yes partially no Yes partially no OK Yes partially no Yes partially no 31 If you didn t completely understand the concepts used on the Smart screen which effect does that have on the way you enter your comfort settings into Smart I didn t vote I don t vote I voted I vote but I don t use the possibilities I don t understand I voted I vote although I m not sure about the possible consequences of my voting behaviour 32 Do you think you know what happens when more people w
94. rs these alternatives were to be preferred over using Smart because i they were perceived as being more effective than Smart and ii they were already included in user routines before Smart arrived Consequently Smart had to confront an adverse building logic Its resulting lack of effective ness then reinforced a user logic that turned users away from Smart It is quite possible that if a building with more favourable characteristics had been chosen as the test location the Smart system would have been assessed more positively The field test revealed some shortcomings of the prototype used Not all aspect of the Smart system notably the use of the option for saving energy and some terms used by the Smart inter face were fully understood by the users This incomprehensibility of some of its parts was shown to lead to decreasing the use of Smart and to negligence by test users of those options they did not understand The scant information about the system given to the test users aggra vated these shortcomings The physical measurements taught that the real temperatures in the building were considerably higher than the average temperature test users indicated as the preferred one In this discrepancy looms an opportunity for the saving of energy in a new building that was estimated to be highly energy efficient This outcome demonstrates that the delegation of energy efficiency to inherent building characteristics large mass high isolati
95. rt e Expectations and opinions on Smart and the appreciation of the Smart system 9 1 Comfort assessment In the preliminary investigation as well as at the end of the Smart field test the test users were asked to assess certain aspects of comfort in the office thermal comfort ventilation ambient air quality as they perceived it For comparative reasons the questions used to this end were the same in both questionnaires The assessment of thermal comfort and ventilation appeared to be almost equal in both situa tions Only the ventilation level at the end of the day was valued somewhat lower during the field test than in the measurement carried out in February Markings by users of the thermal comfort and the ambient air quality in both measurements did not differ much either In Table 9 1 the figures are compared in this table also the standard deviation sd is included Table 9 1 Comparisons with respect to thermal comfort and ambient air quality Preliminary investigation Smart field test Mark thermal comfort scale 1 10 where 10 is perfect 6 9 sd 0 9 6 5 sd 1 2 Mark air quality scale 1 5 where 5 is fresh 2 9 sd 0 9 2 6 sd 0 8 If the standard deviation is not taken into account a weak trend towards a decrease in ambient air quality and thermal comfort can be discerned However the differences between the two measurements fall entirely within the standard deviation if this is taken into account So it c
96. s measured by Building Operation amp Maintenance Main findings and conclusions 8 RESULTS OF THE SMART FIELD TEST 8 1 Introduction of the field test 8 2 User interface 8 3 Results of the field test 8 3 1 Comfort assessment 8 3 2 Use of options for comfort regulation 8 3 3 Saving of energy 8 3 4 Comprehensibility of Smart 8 3 5 Possible improvements of the Smart system 8 3 6 Overall assessment of Smart 9 COMPARISON OF THE TWO MEASUREMENTS ol Comfort assessment 9 2 Saving energy attitude of users and use of the energy saving option T3 Expectations and opinions on Smart and the assessment of the Smart system ECN C 02 094 10 TECHNICAL ASPECTS AND PHYSICAL MEASUREMENTS 41 10 1 Technical aspects 41 10 1 1 Connection of Smart to the existing building management system 41 10 1 2 Measurements and control settings 42 10 1 3 Comfort index calculation as a function of time over the day 43 10 2 Measurements during the field experiment 44 10 2 1 Temperatures and settings 44 10 2 2 Voting behaviour 48 11 CONCLUSIONS AND RECOMMENDATIONS 51 REFERENCES 33 4 ECN C 02 094 LIST OF FIGURES Figure 7 1 Assessment of thermal comfort during the day 24 Figure 7 2 Assessment of ventilation during the day 24 Figure 7 3 Assessment user interest in Smart 25 Figure 7 4 Average daily temperature at the test sit
97. s on the actions of users Relevant guidelines in this respect were considered to be as follows Avoid settings entered by users from force of habit or decency it is perceived comfort that is at stake To break routines e g routines of setting thermostats as being done at home interface settings to be made by users should preferably not refer to familiar values such as degrees Celsius It is better to present choices in differential terms more less or In asking users for comfort preferences it might even be better to refer to direct perceptions of feeling comfortable see interface box below How would you like to feel Warmer No change Cooler L Report status and trends information about what the system is about to realise i e trend might be of higher value to users than only information about the condition s at a certain moment i e status since conditions change quicker than trends last That is information on status is obsolete the next moment whereas information on trends is not An example of giving combined information on status and trend is shown below cf Oseland et al 1997 Current temperature in your zone is 19 C Your zone is now warming Set clear and motivating targets This could be helpful in supporting end users to save en ergy cf McCalley and Midden 1998 6 4 Enrolling end users The first thing the Smart system must achieve is recruiting users The force o
98. s that end users are able to enter into the system Further the technical design characteristics the scripts of the building introduced further constraints on the condi tions users were able to influence see Jelsma 2001 Consequently the authority of Smart users was restricted to management of thermal comfort within the limits set by the building operators and to saving energy in the supplying of this comfort Since energy efficiency is an aspect of sustainability that users might want to promote but also one of cost saving for the employer the team s general opinion was that saving energy should be an option that users could choose to use or to neglect Further it was clear that end users should have only competence to control comfort in the primary rooms i e where they do their daily work Secondary rooms such as cor ridors and hall were to stay under the exclusive authority of the building managers A second aspect of use was the challenge to find solutions for serving individuals within the constraints of the present script of the building i e in an environment of shared technical equipment for creating comfort in an open plan office environment The technical solution sought was to define five thermal or comfort zones in which the temperature could be regulated more or less by bringing the settings of the ceiling heat convection units in each zone under the control of Smart In the kind of hybrid environment present three options
99. small windows to regulate the level of ventilation on the office floor Yes quite often Yes sometimes No almost never 17 What happens when someone opens or closes the windows No consultation takes place Before the windows are opened or closed consultation takes place Consultation takes place after the windows are opened or closed are changed The consulta tion process ultimately leads to a compromise 18 If you never do so why don t you adapt the ventilation level in the room by means of the small windows I don t find it is necessary I don t have time to care about the ventilation level Unknown possibility Because of the necessary consultation processes When I open or close the windows the windows are often closed or opened again by some one else It is thus not very useful to adapt the ventilation level by means of the small win dows ao i a 19 Are there periods of time when it is not possible to regulate the temperature on the office floor sufficiently If applicable please detail your answer which time of the day which kind of weather is it too cold or too warm etc 20 How do you judge the facts that Smart makes participation in the setting of comfort levels possible JT lik th eidea becalsenirnoie nni e E e a oldies a oP i OE SOE NE I don t know Energy savings 21 The following proposition on the area of energy savings is the most appropriate for
100. t the temperature on the floor was generally perceived as high Section 7 2 7 4 Interest in Smart In the questionnaire we took up one question to test the interest of the future test users in having a stake in their own comfort management by operating Smart We could not exclude the possi bility that office workers would see such management as a burden preferably to be delegated to others building operators and systems The basic attitude on this matter was positive though opinions of most users were conditional i e dependent on how the system would function see Figure 7 3 Times mentioned fo O NO A HK DN CO CO T 1 positive negative conditional no assessment assessment assessment assessment Assessment participation possibilities Smart Figure 7 3 Assessment user interest in Smart 7 5 Attitude on energy saving To complement the expert meeting see Section 6 5 we probed test users attitude on their willingness to save energy in the office In order to prevent predictable answers as much as pos sible we asked users response to statements on the topic Per item users were asked to indicate ECN C 02 094 25 which one of pairs of contradicting statements were true see Appendix B The statements of fered were either directed towards the user s attitude to the saving of energy as such or toward the conditions under which users w
101. tems of the questionnaire it was proposed to cluster workers ac cording to comfort preferences as recorded by Smart In that case a more optimal comfort could be realised because the bandwidth within which Smart calculates compromises between indi vidual preferences entered would be smaller Most respondents 78 resisted this idea strongly explaining that office work would suffer too much from such re allocation ECN C 02 094 37 8 3 6 Overall assessment of Smart At the end of the questionnaire respondents were asked to make a final assessment of the use fulness of Smart The outcome is shown in Figure 8 8 The high amount of respondents 50 that apparently felt to be unable to make any assessment in this respect no opinion is remark able Times mentioned k N OUO A aT DN O OO oO Useful Not useful Useless No opinion Assessment Figure 8 8 Assessment of the usefulness of Smart 38 ECN C 02 094 9 COMPARISON OF THE TWO MEASUREMENTS In this chapter the outcomes of the preliminary investigation are compared with those of the Smart field test Apart from additional discrepancies reported in Chapter 8 the only difference between the two situations was the opportunity for users to use Smart during the field test The comparison includes the following aspects e Comfort assessment e Attitude with respect to saving energy and the actual use of the energy saving option of Sma
102. the measured temperature of the five internal climate zones the outside temperature and wind velocity A number of parameters in the central installation necessary for the air and heat flow calculation into the segments was measured as well Furthermore the setpoints of the controllers of the local ceiling heating units were measured These were under control of SMART in the field test period they were modifi able by the user in other periods by setting wall thermostats For the field test the number of lo cal controllers was increased from two to five in order to reflect the segmentation of the office floor Apart from the field test period some data were collected in the winter period to get some insight in the building management operational strategy In the SEBOS model the local con troller for the ceiling heat convection unit and the central air treatment box was modelled as de picted in Figure 10 2 Centrally heated air partly exchanged with the outside is mixed with am bient convected ambient air in a heating unit that obtains it s heat from the centrally heated water 12 Insite View is a more advanced system for building management than the one currently in use at the test site at ECN indicated as BMS InstiteView is developed and appplied by the technical partner in the Smart project Kropman BV Studying opportunities for future extensions of systems such as InsiteView by smart systems as developed and tested here was a major o
103. ticking the option energy efficient were unclear This led to almost complete neglecting of the option for saving energy A better explanation beforehand could have made a difference Linking physical measurements and discussions with experts to data gathered from users we conclude that ambient air temperatures to which employees had become accustomed were considerably higher than the ones they indicated to prefer by means of the question naire This practice of maintaining high office temperature seems to have a history of building operators raising the ambient air temperature in reaction to repeated complaints by office workers who feel too cold while others who then start feeling too warm open win dows to realise the preferred comfort of their own In this way all workers are more or less satisfied but at high costs Just in this situation a well operating Smart system could bring improvement by supplying individual comfort and thus safe cost and energy In sum the outcomes of the field test confirm those of the study on the design stage of Smart WP 4 1 Jelsma 2001 The limited functioning of Smart as demonstrated in this report has not ECN C 02 094 51 much to do with the Smart system as such but results mainly from clashing design logics The building in which Smart was tested had its own logic of handling comfort and saving energy that was at odds with the principles of Smart Nevertheless the field test proved that the tech
104. uch choices is partly a histori cal product of the work situation itself partly imported to that situation from other settings such as the home where also comfort management is taking place On the other hand comfort man agement at work may be governed by a different logic just because one is not at home leading to different practices in terms of needs values assessments and actions Taking user logic into account as a context of interpretation may thus make actions and reactions of users within the system more comprehensible and may save the investigator from making de contextualised in terpretations of user behaviour that may be too individualistic and therefore less informative Designers may anticipate user logic and adapt the scripts of designs to achieve a better match between design logic and use logic Greater congruence between both may ameliorate the im plementation of new technology and prevent unintended forms of use For such anticipation to be possible forms of user representation have to be practised during design The context of the Smart field test was further complicated by the fact that the prototype of the system tested was implemented in a new building which already had a very high energy effi ciency by design That is the saving of energy had been taken out of the hands and minds of the users and delegated to a large extent to actions of the building provisions itself in collabora tion with the operators This could im
105. uilding managers e Brainstorm session with experts e Questionnaire addressing the test users in building 42 1 During the second phase after the introduction of Smart the following methods or instruments were applied e Contextual interviews e Group interview e Questionnaire addressing the test users of building 42 1 16 ECN C 02 094 5 TEST SITE AND TEST USERS 5 1 Test site Smart was tested on the second floor of building 42 1 a new and innovative multi purpose building on the ECN site Designed as a showcase of sustainability the building has leading edge energy technology well integrated in its architecture For instance to avoid mechanical cooling the principle of summer night ventilation has been applied This implies that the build ing mass which is larger than usual to act as a buffer is cooled after a hot sunny day by passive ventilation during the night For this purpose various ventilation shafts have been included in the building design Cool night air enters the building through ventilation valves above the win dows These valves can also be used for additional ventilation during the day the windows can not be opened Blowing in air which is being heated on cold days centrally ventilates the building in the day time Due to the fact that the building is very well isolated the central heating in building 42 is only in operation a few hours a day during autumn winter and springtime The temperature of the
106. us groups lend themselves particularly to explore a new topic and to generate a rich understanding of participants experiences and beliefs see Morgan 1997 An advantage of focus groups above personal interviews is that during the focus group participants can react to each other s opinions and experiences and so uncover valuable additional information In this way it is pos sible in a short time span to disclose unexpected aspects of or reasoning behind user behaviour and about shared or contested opinions The focus group participants must be interested or in volved with the subject of the focus group It is advised to have different rounds per group until saturation of information occurs and to run parallel groups in which homogeneity is sought in terms of sex age education etceteras In this investigation the latter conditions could not be fulfilled because of the small size and the limited availability of the user group To save at least some of the advantages of a group approach especially its explorative and discursive function we applied a weak form of focus groups in the second phase of the study That is we undertook a group interview with a small sample of the test users seven persons recruited via the ques tionnaire in the first phase 4 2 4 Contextual interviews A contextual interview is a combination between observation and an informal interview Beyer and Holzblatt 1998 In a contextual interview a user of a system
107. ut the inner climate of your working environment No yes I did once Yes I did complain to FAC more than once I don t know what FAC is 4 What did you complain about please specify 5 What is your name cee eee eee e ee ence cece eee e teas eee eeaeenaeees 6 What is your gender male female 7 Are you possibly interested in participating in a group interview this group interview will probably take place in May June and will take about one hour yes I would like that I don t know maybe I m interested maybe not no I m not interested 8 When have you planned to go on holidays this year 9 Are there specific days or periods during the day that you re not at work fso please Specily oneee orig aire dos vines bea roca T A E eben eee B 3 First questionnaire Answers The questionnaire was answered by 26 persons Only the answers to questions to the first part of the questionnaire will be given Personal information Question 1 Table B 4 First questionnaire answer question 1 Age Times mentioned 18 25 0 25 35 5 35 50 6 50 15 Question 2 Table B 5 First questionnaire answer question 2 Gender Times mentioned Male 15 Female 11 60 ECN C 02 094 Question 3 Table B 6 First questionnaire answer question 3 Working place Times mentioned Part 1 8 Part 2 9 Part 3 3 Part 4 6 Question 4 The mark given w
108. ws and participant observation were practised The field test made clear that the Smart technology worked in principle and that the test users took a positive attitude toward the system However a comparison of the situation before and after the introduction of Smart taught that the type of building in which the system is implemented can have important consequences for the user assessment of novel comfort regulation systems such as Smart 2 ECN C 02 094 CONTENTS LIST OF FIGURES SUMMARY 1 INTRODUCTION APPROACH Z 3 RESEARCH QUESTIONS 4 PLAN METHODS AND INSTRUMENTS 4 1 4 2 4 3 Plan Survey of research methods and instruments 4 2 1 Questionnaires 4 2 2 Personal interviews 4 2 3 Group interviews 4 2 4 Contextual interviews 4 2 5 Participant observation 4 2 6 Brainstorm session Research scheme 5 TEST SITE AND TEST USERS al Test site 5 2 Test users 6 USER REPRESENTATION IN DESIGN 6 1 User heterogeneity 6 2 Control trust and access 6 3 Communication with end users 6 4 Enrolling end users 6 5 Energy efficiency 7 RESULTS OF THE PRELIMINARY INVESTIGATION aul Interview with building operator iz Comfort assessment and control La Use of options for comfort regulation 7 4 Interest in Smart t2 Attitude on energy saving 7 6 Lal Temperature preferred by users and temperature a
109. y be considered too that is whether the system should inform the user beforehand on the effects of instruc tions entered e g in the case of voting showing before voting what others have voted al ready and of saving energy On the other hand the system should not bother users with in formation they do not want and it should not be a burden by interfering with normal office work too much e Reliability to inspire confidence of users the system should give reliable information and not conceal information that users think to be important For that reason option 2 of user representation see above was found to be problematic Further the system should never interfere with normal office work as carried out on the computer under Windows If users suspect it to do so they will abandon the system 20 ECN C 02 094 Equality and access all end users should have equal chances to manipulate the system for controlling comfort and saving energy This means that the interaction with the system through the user interface should be as simple as possible otherwise differences in compe tence to handle computers might introduce unwelcome heterogeneity between users in real ising benefits through the system 6 3 Communication with end users The formats and forms in which information is communicated by the system to its users were thought to be very important Information contains a script i e the layout of the information ex erts specific force
110. y from the measured temperature on the office floor 22 5 degree showing a slight increase during the day This difference is consistent with i respondents reported perception of the building s climate being rather too hot than too cold ii this per ception becoming stronger during the day and iii the fact that about a quarter of the re spondents wore less clothes than at home ECN C 02 094 27 e Respondents communicated that they prefered to control their own comfort Again control of lighting was a salient parameter here Respondents had poor knowledge about provisions for comfort management already present the thermostats or used them only modestly the ventilation valves e The attitude of the future test users with respect to Smart was found to be rather positive though mixed with a certain reservation concerning its effectiveness e Respondents attitude towards voluntary saving of energy in the office was positive on the condition that such a mission would not lead to decreasing comfort e g compelling workers to wear more clothes On the basis of this findings we conclude that Smart faced a challenge to prove itself at the test site chosen Comfort parameters perceived as important such as ventilation and light fell outside its control The only parameter it could hope to influence ambient temperature appeared to be subordinated to the inertia of a building with a high mass meant to save energy on its own 28 ECN

Download Pdf Manuals

image

Related Search

Related Contents

MT Service Manual  R&S FSH Analizador de espectros  conjunto de reparación tubeless cera para marcar neumáticos    User`s Manual  OL32 User Manual  lait pour sol  Motores Marinos  Tristar KZ-2216 coffee maker  xb31 sb21  

Copyright © All rights reserved.
DMCA: DMCA_mwitty#outlook.com.