of the load。

Reducing both the peak and overall energy consumption is a key outcome, but designing comfortable buildings demands con- sideration of more than just a purely theoretical estimate of energy consumption。 For example as discussed by Kennedy et al。 [9] and Miller et al。 [10], occupant comfort is affected by many factors such as the ability to use natural ventilation which may be influenced by security, privacy or noise concerns, while the requirement for conditioning is strongly dependent on how the building is actually used and ‘operated’ together with the surrounding environment。 The critical importance of occupant behaviour has been highlighted by Berry et al。 [11], who measured energy consumption in theoreti- cally high efficiency buildings and found that occupant behaviour is the overriding factor in determining overall energy consumption。 In particular, the need or desire for personal comfort has been strongly correlated with predictions of household energy consumption as discussed in the review by Frederiks et al。 [12]。

Yet while the link between energy consumption and occupant behaviour is clear, so too is the fact that for given occupants, the degree to which different buildings succeed in creating comfortable conditions is highly variable, and the financial benefit of various

energy efficiency retrofits can be large in proportion to their cost [13]。 The analysis of the factors that lead to a building that performs well, (being one that results in minimum energy consumption and high levels of comfort), or the quantification of the investment return for a particular energy efficiency measure, is made com- plicated by the fact that how the occupants  take  advantage (or not) of energy efficiency features or devices, the particular local cli- mate and environmental factors, and indeed whether the building is ‘operated’ in a way that maximises its performance, are critical factors。 For example a building with good performance with active mechanical cooling may perform poorly by comparison when used with natural ventilation, or vice versa  [14]。

The logical question then is; “What role can building thermal design play in reducing energy consumption whilst improving occupant comfort?” The obvious answer is that buildings must be designed to work well for the occupants who will use them; under- stand the occupant first, then design the building。 However, this is clearly not particularly useful for the development of ‘one-size-fits’ all guidelines or recommendations, and few occupants will ever have the opportunity to have a building thermal design tailored to their specific requirements。

One approach to manage energy consumption, and in particu- lar air-conditioner peak consumption, is to implement a demand response capability in which consumer’s may elect to have their air- conditioner overridden by a network utility during peak demand events in return for a financial benefit; presumably with some impact on comfort。 The technical implementation of this control at the level of a single air-conditioner is described in a recent Australian Standard [15] and current research is investigating the aggregate control of populations of air-conditioners [16]。 Such a scheme has a number of benefits with price signals potentially used to control the level of uptake as required。 Although large scale trials are currently being undertaken [17], the impact on comfort, and on consumer willingness to sacrifice control of comfort during peak events, or even just the perception of control, is yet to be proven over the longer term。

Another approach is to remove the issue of energy consump- tion and in particular, peak consumption via use of a solar driven air-conditioner。 A considerable body of  work  (see  for  example the recent review by Allouhi et。 al [18]) exists concerning solar thermally driven cooling systems designed to operate with high electrical efficiency using heat as the primary energy source。 Recently a number of authors (for example [19–21]) have also considered cooling systems driven by photovoltaic panels; most notably Otanicar et al。 [22] who compared the economics of several different types solar cooling systems in the USA now and into the future, including both capital costs, lifetimes and embodied emis- sions。 However, in their study a backup energy source to guarantee operation was used and the impact of different buildings on the comfort conditions in a totally off-grid system was not the   focus。