This study looks at the daylighting implications of several options fo
r the refurbishment of an existing office building in Brussels, with a
perimeter ceiling height of 2.5 m and a width of 16 m. Each side has
a double-skinned facade, with a 1.4 m wide maintenance walkway in the
space between the internal and external glazing. Scale models in an ar
tificial sky, and computer simulations, were used to examine the effec
ts of changes to the walkway. Two changes in the building's form were
also examined, re-entrant slots in the facade, and lowering of the cen
tral area of floor. The results from these experiments are generally a
pplicable to buildings designed with double-skinned facades, buildings
using horizontal solar shading devices, light shelves or buildings wi
th low Boor to ceiling heights. If an area of the floor space is consi
dered to be daylit when it receives at least 300 Lux for over 50% of t
he working year, it was found that using the walkway options alone, th
e daylit area can be increased by up to 23%. Re-entrant facade slots p
roduced no increase in the daylit area. Lowering the central floor are
a produced an increase of up to 14%. None of the walkway options were
able to produce a daylit area of greater than 53% of the total floor s
pace. Thus, until redirecting glazing becomes commercially viable, it
is clear that shallow-plan designs are the best option for new buildin
gs. The RADIANCE computer modelling package was used in addition to sc
ale-model measurements. The use of both of these methods of daylightin
g analysis is compared. The results of both methods show good agreemen
t. (C) 1997 Elsevier Science Ltd.