Chapter Daylighting

The provision of daylight in a building is strongly linked to the spatial and architectural design. Unlike other environ mental services, the elements of daylighting-windows and surfaces—are surely the most visual and expressive. For this reason windows have a long history of attention from architects; indeed, architectural style has been defined by patterns of fenestration probably more than by any other single characteristic. It should be of interest then, to the architect, that daylighting design also has a strong influence on both the energy use of a building and the general comfort and well-being of the occupants (Figure 6.1).

Daylight as energy

Daylighting can save energy by displacing the electrical energy that would otherwise be used to provide artificial lighting. In most non-domestic buildings this is potentially the most significant energy-saving measure. Reference to Figure 4.3 shows that for air-conditioned and predominantly artificially lit buildings, lighting is the single largest user of energy, but can be reduced by more than half in daylit, shallow-plan buildings. In fact a typical shallow-plan office building, with a plan depth not greater than 15m, occupied for normal working hours, can obtain 70% of the working illumination by daylight.

In addition to this benefit of saving lighting energy the heat generated for a given amount of light is less for daylight than for artificial light. Typical luminous efficacies are given in Table 6.1.

Thus in principle, for a given level of illumination tungsten lighting would produce between 5 and 14 times more heat than daylight.

Table 6.1 Luminous efficacies of common light sources

Lamp

Luminous efficacy 1m/W

Tungsten GLS

8-20

Tungsten halogen

12-25

High-pressure mercury

40-60

Compact fluorescent

50-60

Tubular fluorescent

60-90

Metal halide

70-80

High-pressure sodium

60-120

Daylight

115

If daylight were evenly distributed throughout the room and held at constant intensity by controls, the full benefit of the increased thermal efficiency would be realised. However, in reality the variation of daylight level within a space and the variation with time will lead to over-illumination in some parts of the room and at some times of the day. The resulting increase in thermal gain will probably nullify the benefit of the higher luminous efficiency. Thus in practice, without resorting to a sophisticated variable transmission control and distribution system, the use of daylight cannot be expected to lower thermal gains.

Although this over-illumination carries the penalty of creating a cooling load in summer, it is not in other respects a disadvantage. It is the variation of lighting levels in time and space which gives daylight its essential quality. Too great a spatial variation, however, can lead to glare and inefficient use of artificial lighting. Furthermore, in special buildings such as art galleries and museums, over-illumination must be prevented for the protection of the contents. For other buildings in the UK climate, by simply restricting shading controls to the function of cutting out direct radiation and allowing only diffuse light to enter a room at ambient level, cooling loads and glare can be reduced sufficiently.

6.1 A well-daylit working environment not only saves energy but is also preferred by the occupants. (Housing 21 Beaconsfield, Architects Jestico and Whiles.)

6.1 A well-daylit working environment not only saves energy but is also preferred by the occupants. (Housing 21 Beaconsfield, Architects Jestico and Whiles.)

Well Daylit Working Environment

Daylight factor

Daylight factor

The effect of daylight in a room reduces as one moves farther away from the daylight source. This variation of light intensity is not linear and shows a wide range across the occupied part of a room. The daylight intensity at a given point is related directly to the size of the windows or rooflights. The reflectance of the ground and room surfaces, the shape of the room and the detailed design of the window opening are all factors which influence the intensity and distribution of daylight.

This important topic of daylighting design is outside the scope of this volume and the reader is referred to other sources listed in the Select Bibliography [p. 227]. Here we are concerned with how the initial design decisions on building shape and façade design affect the potential for energy saving by daylighting.

Due to the variation of the illumination from the sky, it is not useful to describe the daylighting in a building in units of illuminance. Rather, the daylighting performance of a building is described using a ratio of the illuminance inside the room to that from the unobstructed sky outside. This is called the Daylight Factor.1

6.2 Variation of Daylight Factor (DF) in a side-lit room for glazing ratios (glazing to external wall area) of 30% and 65%. DF averaged across breadth of room.

6.2 Variation of Daylight Factor (DF) in a side-lit room for glazing ratios (glazing to external wall area) of 30% and 65%. DF averaged across breadth of room.

Daylight Factor Graph

Figure 6.2 shows the variation of Daylight Factor (DF) for a side-lit room. In a room 3m high, at a distance greater than 6m from the window, the DF on the work plane will fall to typically less than 1%, which will be a lower limit for most uses. Increasing the glazing area above 40% of wall area will increase the minimum DF, but will also lead to an unacceptably low uniformity ratio. Thus if a building is greater than 12m deep the inner central zones, i.e. beyond 6m from either side, will need to be permanently artificially lit. Furthermore, the intermediate zone between 3m and 6m will be daylit for fewer hours than the outer zone. A double height space will allow useful penetration up to 12m (assuming the window height is close to that of the wall), indicating that the penetration of daylight is dependent upon the ratio of room height to depth. If it is possible to use rooflights then there is no constraint on plan depth, although clearly this can only be adopted for top floors or single-storey buildings.

The Daylight Factor is predictable from design parameters by various procedures. Physical scale models may be used under real or artificial skies. It may be evaluated from tables or graphical tools such as the BRE Daylight Factor Protractors, or from computer-based mathematical models, which are becoming increasingly available.

Figure 6.2 shows the variation of Daylight Factor (DF) for a side-lit room. In a room 3m high, at a distance greater than 6m from the window, the DF on the work plane will fall to typically less than 1%, which will be a lower limit for most uses. Increasing the glazing area above 40% of wall area will increase the minimum DF, but will also lead to an unacceptably low uniformity ratio. Thus if a building is greater than 12m deep the inner central zones, i.e. beyond 6m from either side, will need to be permanently artificially lit. Furthermore, the intermediate zone between 3m and 6m will be daylit for fewer hours than the outer zone. A double height space will allow useful penetration up to 12m (assuming the window height is close to that of the wall), indicating that the penetration of daylight is dependent upon the ratio of room height to depth. If it is possible to use rooflights then there is no constraint on plan depth, although clearly this can only be adopted for top floors or single-storey buildings.

The Daylight Factor is predictable from design parameters by various procedures. Physical scale models may be used under real or artificial skies. It may be evaluated from tables or graphical tools such as the BRE Daylight Factor Protractors, or from computer-based mathematical models, which are becoming increasingly available.

The sky as a light source

The illuminance from the sky varies over a wide range on an hourly and seasonal basis. There is a large difference, up to tenfold, between direct sunlight and light from the dif fuse or clear sky. A typical overall variation might be from 2000 lux at midday on a gloomy day in December to 100,000 lux in full sun in June. In the British climate, unfortunately, cloudy skies prevail and daylighting design is gen erally based on this assumption. Daylighting from direct sunlight is rarely considered and generally presents problems due to its strongly directional nature.

Sky illuminances—that is, the intensity of light falling on a horizontal surface from the unobstructed sky—have been measured at various sites. A useful way of recording this data is in the form of cumulative frequency curves. Figure 6.3 shows the percentage of working hours that the sky illuminance is above specified illuminance values. Working back from a required internal illuminance it is possible to predict the fraction of time that this level could be met by daylight, and thus the duration for which artificial light would be needed. This is the principle of the evaluation of lighting energy in the LT Method described in Part Two.

Interaction of shading with daylighting

Shading is an almost essential part of passive building design. Its use has three purposes:

1 to reduce the solar heat gain to the room;

2 to prevent sunlight from falling onto occupants;

1 The daylight in a room is described using a relative value, the Daylight Factor (DF). The daylight factor is defined as: DF=I/Iox100%

where Ii is the illuminance in the room at the point of interest and Io is the illuminance from the unobstructed sky, i.e. nominally the outdoor ambient illuminance.

6.3 Availability of daylight for southern UK. Example shows how from a required minimum illuminance and DF, the fraction of daylight sufficiency over the working year can be evaluated.

ofe percentage of year forwhkjh a given diffuse illuminance is exceedea

6.3 Availability of daylight for southern UK. Example shows how from a required minimum illuminance and DF, the fraction of daylight sufficiency over the working year can be evaluated.

ofe percentage of year forwhkjh a given diffuse illuminance is exceedea

diffuse illuminance 1 (XX) s lux Slandarfl Year 09.00- !7.30h BSTApr- Oct inc.

diffuse illuminance 1 (XX) s lux Slandarfl Year 09.00- !7.30h BSTApr- Oct inc.

3 to reduce glare.

However, the presence of most shading devices will also reduce the useful daylight. All too often the 'blinds down lights on' strategy will be adopted and ironically, extra lighting energy will be used when in fact there is a surplus of luminous energy. In other cases, where poorly designed fixed shading devices are used, their obstruction to the diffuse sky is such that artificial lighting is needed at all times except perhaps when there is bright sunshine.

How can this conflict be resolved? First let us consider movable shading devices. Figure 6.4 indicates that in sunny conditions the critical illuminance level is easily achieved at the back of the room, whilst in overcast conditions the shading device is removed from the window aperture. Since the shading device does not affect the illuminance at critical positions and times, i.e. when artificial light has to be switched on, a shading device such as this, provided it is operated correctly, carries no energy penalty. This also presumes that its transmission can be adjusted to allow sufficient daylight when it is deployed. Louvre blinds can be adjusted to do this.

Figure 6.5 shows a light shelf, which partly performs a shading function and partly a redistribution function. Although it completely shades the lower part of the window (greatly reducing the illuminance gradient in the room), reflections from the upper surface redirect light to the back of the room compensating for the obstructing effect of the shelf, even in overcast conditions. Thus, due to its redistribution function, this type of device also has no negative effect on energy consumption. Even a simple overhang can perform nearly as well as this provided there is good reflection from the ground outside to redirect light to the back of the room.

We will call these shading types A1 and A2 respectively—neither increase the use of artificial light.

Now consider a fixed shading device as in Figure 6.6. This reduces daylight by the same amount as it reduces solar gain, and because it is not geometrically selective, it reduces diffuse light by the same amount as direct sunlight. Thus the room is now under-illuminated on diffuse days, demanding more artificial light and hence leading to extra energy demand. If the shading effect is compensated by extra glazing area, this is self-defeating since the window will now allow more solar gain from direct sunlight. We will call this type B shading.

A summary classification of shading elements is given in Table 6.2.

In practice, many shading devices do not fall exactly into these two categories. But the list below indicates broad categories. Furthermore, some devices, often referred to as advanced daylighting elements, primarily function as redirecting devices rather than shading devices (Figure 6.7). These may increase the daylight level at the back of the room in both direct and diffuse conditions.

It is beyond the scope of this volume to give detailed information on shading design. However, the important message is that not all shading devices are equal and they have a widely differing impact on lighting energy use, as well as ventilation and view. This classification of shading type is used in the LT Method.

Lighting control systems

It is obvious that the energy saving of daylight will only be realised if the artificial lighting is turned off when there is sufficient daylight. Because artificial lighting rarely contributes significantly to over-illumination, i.e. a level which threatens

6.4 A movable shading device is only deployed when there is a surplus of light and thus does not have a detrimental effect on the daylighting at times of poor daylight availability. These devices are classified type A1.

6.4 A movable shading device is only deployed when there is a surplus of light and thus does not have a detrimental effect on the daylighting at times of poor daylight availability. These devices are classified type A1.

500 lux 700 lux 10%
300 lux 600 lux 100%

visual comfort, there is not a comfort incentive to turn lights off. The incentive has to be an 'intellectual' one, i.e. a moral attitude to energy use or at least a reasoned concern for energy cost.

This results in light-switching performance in non-domestic buildings being very poor. This is often further exacerbated by light switches not being easily accessible, or luminaires wired in large groups not corresponding to daylight distribution. For example, if a deep room has one lighting circuit, the lights will be switched on, or left on, to provide artificial light to the back of the room, although the front half of the room may be adequately daylit.

Automatic lighting systems which detect daylight illuminance levels can replace the need for switching by the occupants. Field studies have shown that annual lighting

Table 6.2 Classification of shading elements

Type A1

Type A2

Type B

Movable blinds and louvres with variable transmission

Light shelves, fixed reflective louvres, overhangs (with ground reflection), prismatic glass, holographic film

Fixed grids and fixed non-reflective louvres, fritted, tinted and reflective glass

energy is typically reduced by 30-40%. However, fully automatic systems are unsatisfactory, causing annoyance to the occupants mainly due to the frequency and apparent randomness of their operation, often responding to short-term changes in the sky. Modern systems now adopt a compromise approach:

1 the lights are switched off at pre-set times, e.g. on the hour, if the daylight illuminance is above the critical datum;

2 lights may be manually switched on again at any time;

3 no automatic switch-on is provided;

4 lights may also be switched off by a null occupancy detection from an occupancy detector leading to further savings.

Further improvements can be made by top-up or dimming controls. Suppose the daylight illuminance is 200 lux in a space requiring 300 lux. Instead of the artificial lights being switched fully on, which would then provide an illuminance of 500 lux, the artificial lights are dimmed to provide only a further 100 lux, saving power and making the control less obtrusive. Dimming controls save more energy than on/off controls since they make use of daylight even when there is insufficient to meet the illuminance datum, and can reduce annual lighting energy by as much as 60%.

Recent advances in control systems by lighting manufacturers have been to provide 'local intelligence' not linked to a central control system. This has reduced costs, and provided the occupant with much more personal control. In some systems operation can be carried out by TV-style infra-red controllers, greatly simplifying switching layout and reducing hard wiring. Details of these systems can be obtained from the leading lighting manufacturers.

A decision chart for choosing a lighting control strategy, developed by P.Littlefair, is reproduced in Figure 6.8 by kind permission of the BRE.

6.5 A fixed light shelf has a dual function, to shade and reject direct radiation from the front of the room and to redirect light to the back of the room. In spite of being a fixed device, its redistribution function ensures that the illuminance at the back of the room (i.e. the limiting condition) is not affected. These devices are classified type A2.

Bre Sunlight Availability Protractor

6.6 A fixed type B device reduces the diffuse light by the same fraction as the direct sunlight, i.e. it is not geometrically selective. This will create an increased demand for artificial lighting.

Nick Baker Koen Steemers Method
Daylighting and thermal function of glazing

The parameters of the glazing of a building envelope cause a complex energy balance to occur as illustrated in Figure 6.9. Radiation passes through the glazing where the visible part may displace the need for artificial lighting. When absorbed, this same radiation will create heat, and possibly displace auxiliary heating. Both of these functions represent an energy gain. In the summer, however, the heat gain may not be useful and may lead to a cooling load. Also, since glazing usually has a much

6.7 Advanced daylighting devices, such as reflecting louvres and prismatic glass, redirect light to the back of the room, thereby reducing the demand for supplementary artificial lighting.

6.7 Advanced daylighting devices, such as reflecting louvres and prismatic glass, redirect light to the back of the room, thereby reducing the demand for supplementary artificial lighting.

6.8 Decision chart for choosing a lighting control strategy for the most cost-effective energy savings. Source: Littlefair, BRE.
Decision chart for choosing a lighting control strategy

Pattern of occupancy

Lighting control strategy

Number of people

Time

Time switching

j Localised , switching

Occupancy linking

Photoelectric daylight Imkinc

It daylight is avallabte

Many people

Variable

• ••

••

*

HI

Intermittent but scheduled

• ••

M

Continuous

• ••

••

• ••

One or two people

Variable

• ••

• •

» 0

Continuous

• ••

• •

Rarely A occupied 'jj

Intermittent

• ••

It no daylight

All types of occupancy

• ••

-

*••

--

6.9 The energy balance at the glazed envelope of a building.

Building Energy Balance

poorer insulation value than an opaque wall, the inclusion of glazing in the envelope to provide light and useful thermal gain is at the cost of increased conductive losses.

This building energy balance varies hour by hour, month by month, throughout the year. Figure 6.10(a) shows the total monthly primary energy consumed by a double-glazed south-facing office room of 54m2 with 40% glazed external wall in the southern UK. These figures were derived from the LT Model, which has been used to produce the curves in the LT Method. It clearly shows how the demand shifts from energy for heating and lighting in winter to cooling in summer. In reality, both heating and cooling requirements can occur in one month or even one day. Because the model is based upon a monthly energy budget, the results do not show this. However, this limitation is of little significance to the predicted annual energy.

For comparison, Figures 6.10(b) and 6.10(c) show a similar analysis for two rather different European climates. Figure 6.10(b) shows the monthly consumption by a similar south-facing office in Athens. Here, there is a much smaller heating load and a smaller lighting load. The cooling load forms 56% of the total energy demand compared with 15% for the

UK. The lighting load is actually larger in the summer for Athens compared with London due to the need for shading devices and the shorter day.

Figure 6.10(c) shows the results for a similar office in Copenhagen. Cooling energy is now reduced to only 10% of the total. Heating is the main component of the primary energy demand. Note that in all cases it is assumed that full use is made of daylight, i.e. the lights are not on when the daylight is above a datum value of 300 lux. Note too that for the sake of comparison the same U-value for walls is assumed, whereas in practice the insulation standards for Denmark would be much higher than those for Greece.

Figure 6.11 shows two sets of curves generated by the LT Model, giving the annual primary energy demand per m2 for a south-facing office in the UK, as a function of glazing ratio, for single and double glazing. Energy units are primary energy. The glazing ratio is the area of unobstructed glass as a fraction of the total wall area (including glass). The LT Model assumes that full use is made of the available daylight by automatic controls or good occupant control.

In both cases, as glazing is introduced from zero there is a rapid initial reduction in lighting energy demand due to the sudden benefit of quite modest glazing areas replacing the need for artificial lighting. For a glazing ratio of 20% the double-glazed office consumes slightly more lighting energy than the office with single glazing. This is because double glazing transmits about 15% less light than single glazing. As the glazed area increases, the rate of increase of useful lighting reduces and the thermal benefit of the double glazing becomes more significant. By 80% glazing ratio the single-glazed building consumes nearly twice the heating energy of the double-glazed building, but the same amount of lighting energy.

For single glazing there is an optimum area of 35% glazing for heating and lighting, but rather less than that in buildings with mechanical cooling. For double glazing, energy consumption for heating and lighting is still reducing at 70% glazing ratio and is 17% less than the energy demand for the optimum single-glazing value. However, the risk of overheating is greatly increased for large glazing areas on east, south and particularly west façades.

This increased overheating risk is rather under-represented by the LT curves because the cooling load is shared over a year. This issue is discussed in more detail later [12.6].

The cooling energy shown here is mainly the energy needed for heat rejection (refrigeration) and the extra fan power needed for cooling at higher air-change rates. This assumes a Variable Air Volume (VAV) system where the flow rate responds to the cooling (or heating) demand. In buildings where the fans operate at a constant rate, unnecessarily high recirculation occurs continually and leads to wasteful energy use.

For a successful passive building, the perimeter zones will not need mechanical ventilation nor cooling, and will make maximum use of daylight. This is the basic concept of the passive zone which is discussed in more detail in Chapter 8.

The fraction of glazing, the glazing type of the façade, and the depth of plan and section, all influence the proportion of the floor area to which daylight is available. For typical ceiling heights the depth over which daylight is useful is no more than 6m, but a rooflit space will place no such restrictions on plan depth. Strong variations of daylight factor lead to the use of more supplementary artificial lighting.

The energy balance of the glazed envelope leads to a shallow optimum for glazing ratios from 25% upwards according to glazing type and orientation. However, glazing ratios in excess of 50% should generally be avoided due to increased risk of overheating.

(Note that energy is given in Primary Energy units. The reason for this has been discussed in Chapter 4. All of the values in this chapter are given as examples only and should not be used for calculation.)

6.10 Monthly primary energy consumption (kWh) for south-facing offices in London, Athens and Copenhagen (from LT Model).

LONDON

heal

cooling

lights

total

JAN

969

0

650

1620

FEB

665

0

296

961

MAR

390

0

232

622

APR

232

0

90

497

MAY

17

0

46

322

JUN

0

251

45

296

JUL

0

467

46

513

AUG

0

354

46

400

SEP

0

73

135

208

OCT

132

0

279

411

NOV

458

0

405

863

DEC

819

0

697

1515

ANNUAL

3681

1145

2968

7795

Table 6. Wa: Monthly primary energy consumption (kilo-Watt hours)lor a south-facing office room in London (Irom LT Model)

ATHENS

heat

cooling

lights

total

JAN

240

0

335

565

FEB

164

0

212

375

MAR

193

0

239

332

APR

0

185

93

721

MAY

0

628

46

63

JUN

0

586

90

676

JUL

0

609

93

702

AUG

0

579

139

719

SEP

0

527

135

663

OCT

0

483

232

716

NOV

0

38

270

307

DEC

63

0

325

388

ANNUAL

660

3634

2188

6482

COPENHAGEN

heat

cooling

lights

total

JAN

1589

0

697

2286

FEB

1080

0

381

1461

MAR

640

0

232

872

APR

407

0

90

497

MAY

119

0

46

165

JUN

0

153

45

198

JUL

0

505

46

551

AUG

0

399

46

445

SEP

4

0

180

184

OCT

221

0

279

499

NOV

678

0

674

1352

DEC

1216

0

697

1913

ANNUAL

5953

1056

3414

10423

TableB. 10b: Monthly primary energy consumption (kilo-Wall hours) tor a south-lacing office room in Athens (from LT Model)

Table 6.70c: Monthly primary energy consumption (kilo-Watt hours) for a south-facing office room in Copenhagen (from LT Model)

6.11 LT curves showing annual primary energy consumption (megawatt hours per square metre) for heating, lighting and cooling, for south-facing office in southern UK.

+1 0

Post a comment