Daylight Harvesting and Historic Buildings

This article explores how energy can be saved by utilising and controlling daylight and reducing our reliance on artificial lighting.

Why is daylight so important?

Humans are diurnal animals; we are awake during the day and daylight has an enormous impact on our health and wellbeing. Sunlight supports our biological requirements by triggering our circadian rhythms (a natural, internal process that regulates the sleep–wake cycle and repeats roughly every 24 hours, sometimes called our ‘body clock’) via non-visual receptors in the eye. By increasing daylight within our working environments, we can create better, healthier and more productive spaces as well as saving energy.

The Chartered Institution of Building Services Engineers (CIBSE) has produced Technical Memorandum 40: 2020 'Health and wellbeing in building services' to demonstrate how this, along with other performance parameters, can impact our health.

A pilot research project to investigate lighting conditions in Historic England offices

With the drive towards ‘net-zero’ we were interested to see what energy saving outcomes some of our office buildings could achieve with historic glazing designs. We modelled the principles of ‘daylight harvesting’ for parts of two of Historic England’s offices. Daylight harvesting uses daylight to offset the amount of electric lighting needed to light a space, thereby reducing the energy consumed. It also uses lighting control systems that are able to dim or switch electric lighting in response to changing daylight availability.

For the studies Bessie Surtees House (grade I) in Newcastle and the Engine House (grade II) in Swindon were chosen.

The first office comprises two linked merchant’s houses, one (Bessie Surtees House) timber-framed and sixteenth century in date, the other (Millbank House) a brick building of the seventeenth century. The rooms picked for the analysis are architecturally typical of the periods and are both on the main south-east facing façade.

The room and window proportions vary substantially, with the room in Bessie Surtees House having a higher ceiling but deeper footprint than the office in Millbank House. In addition, the window designs are greatly different.

The rooms chosen for study in the Engine House were very different. The building was originally the offices of the Great Western Railway Company. It was built in the 1840’s, extended in the 1870’s and again in 1903-08. Today it provides extensive open-plan offices on four levels. We chose two rooms, one on each of the second and third floors.

Setting up the parameters and identifying targets: the Newcastle office

Using specialist software, a computer model was created for each of the rooms. The model took account of any overshadowing from neighbouring buildings and to give accurate results 3D camera scans, listing records, details of the existing lighting, the use of the spaces and current desk layouts were incorporated.
The first part of the analysis looked at the internal daylight in relation to current best practice standards and the second looked at the potential for energy savings with daylight dimming incorporated into the rooms’ lighting controls. The controls would utilise a light level sensor or photo-cell that measures the light level against a pre-set value and then to achieve that the control system either dims, switches off or lowers blinds, to adjust the light levels within the space to the desired level.

For the purpose of the analysis, it was assumed that both rooms in the Newcastle office had the same light reflectance values (the percentage of light a surface reflects) and the same Climate Based Daylight Modelling (CBDM) data (daylight conditions derived from standard meteorological datasets).

We wanted to discover whether the rooms could achieve a given illuminance level, as measured in ‘lux’. We took 300 lux at desk height by daylight alone as an adequate standard, and studied for what percentage of the occupied hours the rooms could do this for. This value is known as Spatial Daylight Autonomy (sDA). A good standard is if 55% of a space achieves 300 lux for 50% of the occupied hours and an excellent standard is if 75% of the space achieves the same.

An analysis of the annual lighting energy consumption was carried out based on an energy consumption density figure of 13 watts per square metre (W/m2) for fluorescent lamps which is the predominant form of lighting used in these offices.

The results for Bessie Surtees House

This room achieved a sDA of 37.9% which means it falls below the 55% target for a good standard. Although this room is south-east facing and has ample glazing, it is challenged by the depth of the room relative to the size of the window, its lower floor-to-ceiling height and its location on the second floor, which makes it susceptible to overshadowing from nearby buildings.

The total annual energy consumption with no daylight dimming was calculated at 1184 kilo-watt hours (kWh) per year. However, with daylight-linked controls this reduces to 602 kWh – a saving of 582kWh. This equates to a total saving across the year of £95.00 and 136 kg.CO2 (1kg of CO2 = 0.27kg of carbon) based on an average cost of electricity of 16.3p/kWh and a carbon factor of 0.233kg.CO2/kWh.

The results for Millbank House

This room achieved a sDA of 62.5% which exceeds the minimum for a good standard but does not reach that for excellent. The height of the room in relation to its depth is favourable for daylighting and its elevated position on the third floor makes it less susceptible to overshadowing.

The results for this room indicate that the total annual energy consumption with no daylight-linked controls is 619kWh per year. With daylight linking it reduces to 316kWh, a saving of 303kWh. This equates to around £50 saving per year for this room and would reduce carbon emissions by 71kg.CO2.

These results demonstrate that even with a low sDA result, such as that for Bessie Surtees House, significant savings can still be made by incorporating daylight-linked controls.

Setting up the parameters and identifying targets: the Swindon office

This second study involved two typical office areas on the second and third floors of this mid-19th century office.

This was more detailed analysis than the first and included the use of additional metrics, including Useful Daylight Illuminance (UDI), which is the percentage of the occupied time that a space can achieve useful daylight illuminances within a given range.

The range chosen was 300 – 3000 lux on the horizontal plane 0.85 metres above floor level. Any lux level above that was considered excessive and under that range as needing supplementary lighting.

Other metrics included Temporal Analysis, which was used to identify times of the day and year when glare (difficulty seeing in the presence of bright light) would likely be a problem and Annual Daylight Glare Probability (DPG) analysis which helped to understand the proportion of the occupied hours that different desk locations in the space would experience glare. The offices were chosen for their orientation and the skylights, both of which give a greater potential for glare.

Again, we were able to utilise information such as desk layouts and plans to ensure an accurate result. All outcomes were calculated based on illuminance values with and without blinds. As expected, without blinds the third floor, where the skylights are located, had a consistent level of excessive UDI, that is, illuminances above 3000 lux on the task plane. Predictably when simulating the use of sensor-controlled blinds on the third floor there was an increase in the levels of acceptable UDI. The reverse was seen on the second floor, where if the blinds are used there is a decrease in acceptable UDI, which equates to the lights being switched on.

The areas of the office likely to experience glare were established by the working plane being adjusted to 1.2 metres from finished floor level, that is, to a height where most people’s eye level would be when seated at their desk. Both levels 2 and 3 receive high levels of direct sunlight along the south-west façade throughout the year and it is here that occupants are likely to experience glare. Employing Temporal Analysis, it was established when during the day and year this would occur.

The results for the Engine House were calculated using a lower lighting energy consumption density of 3 W/m2 because the existing installation has been retrofitted with LED lamps. This indicated that without daylight linking the lighting would consume 4,211 kWh per year, but with it the total energy consumption would fall to 1941 kWh, a saving of 2270 kWh. Using the same unit price of electricity as before, this equates to financial savings of £370 per year just for the level 2 area: for both floors in the study this increased to around £640 per annum, with a carbon emissions reduction of 930 kg.CO2 per year.

Further information

Download as a PDF format magazine