Power to the workers

Many studies show wide variations between peoples' preferences for room temperature, ventilation rates and lighting levels. A setting that suits one person may be unacceptable to another working adjacent. Both may be dissatisfied if the environment is centrally controlled.

Of these temperature is probably the most important. Setting the room temperature has often relied on facilities managers responding to workers requests. In large buildings this can place considerable demands on facilities managers.

In such cases, a controls system that enables workers to influence the temperature locally would reduce the demands on facilities managers, and potentially improve worker satisfaction thereby encouraging higher productivity. Such a system could also reduce energy consumption by decreasing the demand for heating, ventilation and air conditioning. The DUCOZT (Democratic User Control of Zonal Temperature) system is intended to meet these requirements.

DUCOZT explained

DUCOZT is primarily for use in open plan, air conditioned office spaces with a full bems-controlled hvac system that allows the setpoint temperature of small zones to be adjusted (see Figure 1).

DUCOZT polls users' temperature preferences and automatically adjusts the setpoint according to the voting pattern within each zone. All occupants can vote on whether they wish to feel warmer or cooler, using an application on their computer terminals. Votes from occupants in the same zone are collated and the percentage of staff requesting a change in temperature is calculated.

The result of this is communicated to the bems which adjusts the zone setpoint accordingly. The system therefore causes little disruption to the staff and the facilities management and is not dependent upon an engineer being present.

Figure 2 shows the dialogue box on the DUCOZT user interface. Once opened, a single click on one of the three response buttons is required to make it disappear. Users may vote for their environment to be warmer, cooler or have no change.

The dialogue box also provides feedback on the current zonal temperature and an indication of whether the area is heating or cooling. This information is obtained from the bems. The building services engineer can also broadcast system faults via the occupants' computer screens.

Users are also told how many people voted recently and the percentage that selected each option. So if the system doesn't respond the way an individual requested, they are aware that it was because they were outvoted and their colleagues required a different change.

To prevent users from constantly voting, and enable the hvac system time to alter the conditions, a limited number of votes will be implemented, for example one vote count per half-hour. The user interface states the time lapse before the next vote count.

The system trialed

A trial study between November 1998 and March 1999 was carried out to assess the application of the DUCOZT control system in an air conditioned office building. The study also identified the actions necessary to install and run the system effectively in a complex, networked environment. The test site was the BT Trinity Park building, a recently completed, three-storey commercial office adjacent to the NEC.

The DUCOZT software was integrated into the building's bems and computer network. The first, second and third floors represented three experimental groups. The ground floor was not included in the trial as there was deemed to be insufficient office-based staff in this area. Staff on floors two and three were engaged in similar types of activity (mainly call-centre work), while staff on the first floor had a variety of office-based jobs.

Floor one was used as a control group and DUCOZT was unavailable in this area. Floors two and three were experimental groups. The order of operating conditions in the two groups differed to enable an analysis of the effect of voting zone size, and of smaller or larger increments in temperature.

The increment size was restricted to avoid draughts from the high airflow needed to make such rapid adjustments. Period one (week one) was a baseline period, with which to compare the effects of the introduction of DUCOZT over the next time periods.

The experimental design allowed researchers to look separately at two possible mechanisms for any reported benefits: changes in perceived control and changes in thermal environment. It was possible to examine the relative importance of perceived control and actual ability to achieve comfortable conditions since the conditions on the second and third floors were presented the same, with control being greater on floor three.

The experimental conditions were reversed at the end of the second experimental period ie in the seventh week, with floor two having full control and floor three dummy control.

Does DUCOZT work?

There was sufficient dissatisfaction among workers at the start of the study to give DUCOZT the potential to demonstrate improvements. Reports from end-users, the building's facilities management and building services teams all indicate that the system is perceived as a net benefit.

The majority of end-users did not use the system daily. Pockets of high activity voting were scattered throughout the three floors. These were caused by environmental hot spots, for example, areas directly in front of a window and subject to high solar gain, or below a ventilation duct with cold air flowing out. There was a clear perceived benefit to these end-users, in being able to control the cause of their discomfort.

Despite the large proportion of end-users not voting, there was an overall perceived benefit. End-users reported a willingness to use the system again, but changes to the existing DUCOZT system would be needed to improve its suitability to their requirements.

For the facilities management team the perceived value of DUCOZT was as a diagnostic tool to allow more pro-active management of the space. It also allowed monitoring of end-user activity. Unusual levels of activity could indicate an emerging problem, requiring investigation. This information would supplement rather than displace the information available from the bems.

The statistical analysis supported the anecdotal feedback in that the perceived benefits were greater on floor three. Here end-users' responses tracked changes in the DUCOZT set-up, but these changes did not follow in the environmental data. This perhaps indicates that changes in the thermal environment caused by votes were not sufficiently large to influence the average floor temperature.

A number of other issues may have masked the statistical analysis. The voting algorithm clearly has a major impact on the effectiveness of DUCOZT, and that used in the trial was decided before the activities of the building's occupants were known. It should be possible to derive a more effective algorithm, to increase the use and impact of the system.

The computer implementation of DUCOZT also caused some difficulties due to the design of the screen interface for the facilities managers. Sometimes the system was shut down when the intention was purely to minimise the dialogue box. A redesigned interface would solve this problem.

The DUCOZT system has the prospect of providing facilities management and services engineers with a means of better managing open plan, air conditioned offices. This raises the prospect of higher productivity, increased user comfort, and greater energy efficiency.