INDIVIDUAL CONTROL OVER LIGHT SOURCE SPECTRUM: EFFECTS ON PERCEPTION AND COGNITION Veitch, J.A., Dikel, E.E., Burns, G.J., Mancini, S. NRC Institute for Research in Construction, Ottawa, Canada jennifer.veitch@nrc-cnrc.gc.ca Abstract For solid-state lighting to make a major dent in the lighting energy budget will require light sources and luminaires that are more efficacious than the existing technology, and a reason for building owners and tenants to change to an initially expensive, new and unfamiliar technology. New luminaire designs and operation modes using the unique characteristics of SSL will aid this transition. The ability to individually tune the light source colour is one such characteristic. Building on prior research concerning the benefits of individual control over illuminance, this experiment tests the effects of tunable light source colour and initial LED spectrum on office workers’ cognitive performance, mood, and satisfaction over the working day. Keywords: solid-state lighting (SSL); light-emitting diodes (LEDs); tunable colour; light source spectrum; cognitive performance; mood; satisfaction; offices; work. 1 Introduction Solid-state lighting (SSL) (light-emitting diodes [LEDs] and organic LEDs [OLEDs]) is a new technology undergoing rapid development. The efficacy of commercial white SSL sources is expected to double that of fluorescent in the long term, although commercial products are not yet as efficacious as linear fluorescents for general lighting. For SSL to make a major dent in the lighting energy budget requires addressing general white light applications, particularly in offices. However, as yet the industry is not conducting very much research and development on new luminaire designs and operation modes using the unique properties of SSL, but remains focussed on incremental source efficacy improvements and socket-for-socket replacements in existing luminaires. New luminaire designs and operation modes promise greater energy savings and value to the user; the latter will encourage earlier adoption of SSL systems. Put colloquially, few people will adopt an unfamiliar, complex, and (initially) more expensive technology unless it has the potential to offer features and functionality that is different from, and perceived as superior to, the existing technology. Previous research at the NRC Institute for Research in Construction and elsewhere has demonstrated that personal control over light level in an office environment has benefits for occupant satisfaction, comfort, and mood, and also contributes to energy savings [e.g., Boyce et al. 2006; Newsham et al. 2004]. LEDs now offer the possibility of some degree of control over delivered light spectrum. This experiment is a test of the viability of individually tunable light source colour for general office lighting. We hypothesized that individual preference for light spectrum does vary substantially across the population of office workers, and that having personal control over light source spectrum will enhance occupant well-being. Others have suggested that varying spectrum systematically by time of day might have direct physiological benefits, via linkages to circadian rhythms, or by other mechanisms [Bommel 2006]. Existing colour quality indicators (e.g., R a ) are poor predictors of colour quality for LED lighting, and the search for better metrics is a major part of international LED research [Davis and Ohno 2009]. The results of this experiment will also address LED colour quality effects as a contribution to this international debate. The experiment is a 2 x 2 (initial spectrum X afternoon tunability) factorial design. The initial spectra are ~3000 K and ~ 6500 K; half of the participants were given the option to change the light source spectrum after lunch in the full-day session. This experiment is designed to test five hypotheses: H1: When given the opportunity, individuals will self-select different spectral power distributions as their preferred lighting. H2: Being given control will stop or reverse the daily, fatigue-driven trend in reduced performance and mood (trends that have been observed previously).