Rethinking Light Levels
Rethinking Light Levels
NOTE: This article was updated Jan 15, 1999. Read more. The recommended illuminance level is but one of many parameters that the lighting designer has to take into account when designing a lighting installation. However, unlike many other parameters, this one is always considered. An international comparison shows that there is no consensus when it comes to defining the "right light level".
Table 1
Figure 1, Figure 2
Lighting design trends are highly dynamic. Our survey of the situations in various countries revealed that recommended illuminance levels have been changing rapidly since the 1930s. In most countries, levels increased and then decreased. The variation among countries is tremendous. In this article we compare recommended illuminance levels for selected tasks in non-residential buildings using data compiled from 19 countries. These dramatic shifts in recommendations have clear implications for lighting design and have had a substantial impact on lighting energy use.
Current recommended levels vary 10- to 15-fold for various office building activities across the 19 countries: Further, we found a 6- to 10-fold variation for schools, a 15- to 30-fold variation for retail stores, a 6- to 10-fold variation for hospitals, and 25- to 40-fold variation for factories. Among the activities/building types showing dramatic variation are reading tasks (75 to 1 000 lux), detailed drafting (200 to 3 000 lux), patient rooms in hospitals (30 to 300 lux), testing and assembly of electronic components (200 to 5 000 lux), and fine knitting and sewing (50 to 2 000 lux).
Belgium, Brazil, and Japan have among the highest levels for the tasks and building types that we examined. Australia, China, Mexico, the former Soviet Union/Russia, and Sweden have among the lowest levels. The North American recommended illuminance levels are about average in most cases, but in a few cases the bottom level in the three-level recommendation is lowest for North America.
RISE AND FALL
Almost without exception, there was a steady increase in levels from the 1930s to the early 1970s. Among the more dramatic cases, the UK's retail lighting level recommendations increased from ~100 lux in 1936 to ~500 lux in 1972. In the former Soviet Union, the recommendation for general office lighting was ~25 lux in 1930, but had risen to 300 lux by 1979. In North America, recommendations for chalkboard lighting rose from ~150 lux in 1938 to 1 400 lux in 1972. Limited evidence indicates that levels were even lower prior to 1930, e.g. 35 lux for detailed drafting in 1915.
Since the early 1970s, however, the trends have either leveled out or changed direction. The general office lighting recommendation in Finland fell from 450 lux in 1974 to 225 lux in 1985. Dutch recommendations for reading fell from 750 lux in 1970 to 400 lux in 1991. Even very demanding tasks, such as detailed drafting work and reading a chalkboard, showed reductions of 50 percent or more. The most dramatic reduction was from 1500 lux for vdt tasks in the 1972 North American IES recommendations to about 300 lux in the 1993 recommendations. Australian VDT task recommendations also dropped precipitously, from 600 lux in 1976 to 160 lux in 1990.
Sometimes such changes can be partly explained by changing definitions that accompany new concepts in lighting design. For example, Swedish office lighting (on the desk) plunged from 1 000 lux in 1970 to 300 lux in 1992. In 1970, a single value was given for reading and general desk lighting, whereas in the 1992 recommendations desk lighting and reading lighting were treated separately. The difference between the 1970 recommendation of 1 000 lux (including the reading field) and the 1992 recommendation of 500 lux for reading is still great, however.
The dynamic nature of illuminance level recommendations is the result of a variety of factors, including changing views concerning the amount of light needed to perform a given task. This dynamic development also reflects a trend towards using more comprehensive lighting recommendations when made within the context of other quality-related attributes such as a glare index and color rendering. In addition, economic considerations play a role. For example, the commercialization of the fluorescent lamp in the 1930s made it possible to dramatically increase light levels without paying a corresponding penalty in energy costs or excessive heat.
THE ENERGY DIMENSION
The assignment and application of illuminance levels represents an important area where lighting design and energy analysis meet. In both fields, illuminance levels are only one of the many relevant parameters describing lighting systems and their performance. Yet in both cases, illuminance levels serve a useful function in helping to quantify the lighting- or energy-related service delivered.
Illuminance standards can also be used to encourage energy-saving daylighting, as is the case in Austria, Germany, the Netherlands, Russia, and Switzerland, all of which require that most types of workspaces have direct access to daylight.
Field measurements of existing light levels followed by the adjustment of actual levels to comply with recommendations will, in many cases, yield energy savings. This is evidenced by one large survey of a major national laboratory operated by the US Department of Energy. Of the nearly 86 000 fluorescent lamps included in the survey, 32 000 (or 62 percent) could be delamped without dropping below 500 lux. Another field study conducted by the French Lighting Association illustrated that improving sub-standard light levels neednÕt result in increased energy usage. The French study involved retrofitting luminaires in schools in order to achieve improved energy efficiency and to meet existing standards with regards to illuminance levels for desks, walls, and blackboards.
For some countries, reductions in recommended illuminance levels over time may, to some extent, have offset the past growth in lighting electricity demand due to increased floor area. The proposed European-wide cen recommendations could have a significant effect on future lighting energy use. For retail (ambient) lighting, the proposed levels are 45 percent lower than current recommendations in France, 20 percent lower than BelgiumÕs average recommendations, and about 30 percent lower than those for Finland. On the other hand, the levels proposed for VDT tasks are higher than most current national recommendations.
The presence or absence of task lighting is central to the overall energy efficiency of illumination. The Japanese illuminating standard provides very extensive guidance when it comes to illumination levels required for specific tasks, encouraging designers to reduce ambient lighting to as little as one-tenth that of the task illuminance. The Japanese approach to articulating their recommendations facilitates the specification of task lighting by designers.
ACTIVITIES AND TECHNOLOGY CHANGE
The changing nature of certain activities suggests another potential linkage between illuminance levels and energy use. One clear illustration is the increasing importance of computers and VDTs in the workplace. VDTs are replacing the drafting table and many paper-based reading and writing activities. These tasks require less illumination when the medium is the computer than when it is ink and paper.
An assumed lamp lumen depreciation on the order of 20-30% has traditionally been embedded in standards and in how designers interpret those standards. Today, new light sources, such as 26 mm (T8) triphosphor fluorescent lamps and the recently introduced 16 mm (T5) lamps, show minimal depreciation. Similarly, dimming ballasts allow for automated lumen maintenance over time. Such technologies reduce the need for illuminance standards to anticipate reductions in delivered illuminance over time, and, as a result, initial lighting power densities can be lowered without compromising the level of illuminance.
The recent shift in the UK from average to maintained illuminance will increase energy use where higher wattage lamps are installed as a means of achieving the recommended light level wherever nominal values in the 1994 cibse recommendations remain unchanged. The proposed cen recommendations are also defined in terms of maintained levels, even where they replace current recommended average levels.
WHAT IS RIGHT LIGHT?
There is certainly more to lighting design than specifying illuminance levels. The quality of illumination is a function of many other factors, including spatial orientation (horizontal vs. vertical illuminance), scotopic/photopic content, glare, contrast, color rendition, color temperature, and flicker. This further complicates the problem of defining meaningful measures of lighting services (both in terms of energy use and illumination quality).
Recently, there has been a trend towards formulating illuminance and associated lighting-design recommendations in a more sophisticated and sometimes flexible manner. In many European countries, glare criteria and recommendations concerning color rendering have been available since the sixties. In other countries, as typified by the most recent Japanese standard for office lighting, glare criteria have only recently been introduced alongside illuminance levels. The current proposed Russian standard specifies disability glare indices, recommended rates of luminaire cleaning, and indices for all lighting applications, as well as color rendering indices and color temperature ranges for industrial and residential settings. There is also a trend towards specifying illuminances for specific types of activities rather than for types of architectural settings, and it is becoming increasingly common to take a more holistic approach that combines task and ambient lighting.
There is still no consensus among countries as to the "right" light level for a specific task and building type - even within a given country over time. Currently recommended illuminance levels in the various countries are tending to converge at levels significantly lower than those existing in recent decades. In all probability, the future will see a more sophisticated integration of energy and non-energy considerations in lighting design. Although it has been a convenient measure of lighting energy services, the "lux" is only an approximate and incomplete indicator. In the end, this trend will probably help to further raise the energy efficiency of lighting services.
Evan Mills
Nils Borg
This article summarizes a more in-depth report forthcoming in the Journal of the Illuminating Engineering Society of North America.