RAF Keevil Aerodrome Glint & Glare Assessment for the Proposed

RAF Keevil Aerodrome

Glint & Glare Assessment

for the proposed

Poulshot Lodge Solar Farm

Photovoltaic Array

Prepared for Hive Energy Ltd by Charles Morelli BEng

27 January 2014

AARDVaRC Ltd, PO Box 10785, Sudbury, Suffolk, CO10 3AY T: +44 (0) 1787 468539 E: [email protected] W: www.aardvarc.ltd.uk

1 Executive Summary 1.1 Background The Ministry of Defence (MoD) has objected to the proposed Poulshot Lodge solar farm on the south side of the A361 road, 4km west of Devizes (planning reference: 13/05244/FUL) on the grounds that it has insufficient information regarding the phenomenon of glint and glare effects on RAF Keevil aerodrome, approximately 6km to the southwest of the site (measured from the site centre to the mid-point of the main runway). This report follows on from AARDVaRC’s earlier general glint and glare assessment (Glint and Glare Assessment for the Poulshot Lodge Farm Photovoltaic Array, dated 26 June 2013); the scope of this earlier report did not include receptor specific analyses, in the report it was observed in Section 6.5.2.5 Keevil Airfield that: ‘…Keevil Airfield is 6km southwest of the site. It is well outside the glint zone for near-ground level receptors, so will be unaffected by glint from the proposed solar farm.’ The MoD noted in its letter of objection (reference: D/DIO/43/2/44 (2013/1105), dated 21 November 2013) that: ‘The application is supported by an assessment of potential impacts of glint and glare. This assessment identifies RAF Keevil in the assessment but only considers it as an area of land. The assessment does not evaluate the potential affects [sic] of glint and glare emissions upon military aviation activities conducted at this airfield. ‘Therefore, at this stage, the MOD considers that the application does not adequately demonstrate that the proposed solar array will not cause a hazard to air traffic safety as a potential source of glare in proximity to an aerodrome approaches. As such, there is insufficient information provided in the application to enable us to determine our statutory safeguarding position in relation to this application.’ ‘The Civil Aviation Authority (CAA) has published interim guidance on the development of solar arrays in the vicinity of aerodromes in the UK. It is recommended that the applicant considers the technical guidance and assessment methodologies referred to in this.’ The CAA document to which the MoD letter refers to is shown at Annex A to this document. This itself notes the requirements of the Air Navigation Order as may be pertinent to solar farms, and refers to a United States Federal Aviation Administration (FAA) document – shown at Annex B to this report – which by the FAA’s own advice is unreliable (as discussed in detail Section 3 of this report) and conclusions should not be drawn from that document on its own. Otherwise the CAA document gives virtually no guidance on assessment methodologies for glint and glare as they may be pertinent to aircraft. This report considers the dates and times and other pertinent information when solar reflections from the proposed solar farm may affect military aircraft operating at RAF Keevil aerodrome. It is noted that there is little formal information published about the aerial activities practiced there by military units, and in the time available, no information has been received. Therefore, assumptions have had to be made as to the military activities conducted at RAF Keevil. These are noted in this report.

Poulshot Lodge PV Array 2 RAF Keevil Glint & Glare Assessment

1.2 Assessment Method 1.2.1 Poulshot Lodge Solar Array It was assumed that solar reflections could occur from any point within the solar farm boundary. For detailed analysis, the solar farm was modelled by 20 points on the boundary and across its area and their 3 dimensional coordinates noted. The solar farm boundary is shown in the following chart showing the solar farm points used in the analyses; 2 solar reflections zones identified in initial analysis are also shown: the ‘near-horizontal solar reflections zone’ marking where solar reflections will be observable when directed at 5° below, to 10° above, horizontal (the zone bounded by orange lines); and the solar reflections zone for all reflections directed a upwards at less than 45° above horizontal (the brown lines) which may be significant for aerial receptors of solar reflection when near the solar farm.

Limits of significant aerial reflections zone

Limits of near-horizontal reflection zone

Prominent points shown in detailed analyses

Poulshot Lodge PV Array site boundary

Ordnance Survey © Crown Copyright 2013. All rights reserved. Licence number 100020449 Whilst solar reflections at angles greater than 45° above horizontal may affect aircraft – and through the day and year it is assumed that they may be directed in any azimuth from 0° to 360° – they are of less significance as they will only be visible from steeply below an aircraft – a direction of limited interest to a pilot – and it will often be blocked by the aircraft’s structure (particularly tactical transport aeroplanes which generally have very limited visibility steeply below the aircraft). Furthermore, as such reflections are directed upwards at a 100% gradient (1:1) or more, they can only be seen when

Poulshot Lodge PV Array 3 RAF Keevil Glint & Glare Assessment

an aircraft is as close to a panel horizontally as it is above it: at 3,000ft above airfield elevation, such reflections will never be seen when 915m or more from the solar farm boundary; at 1,000ft above airfield elevation, this reduces to within 305m from any solar panel. 1.2.2 Receptors 1.2.2.1 RAF Keevil Overview RAF Keevil aerodrome is 8km west-southwest of Devizes, Wiltshire. It’s military use is understood to be primarily by RAF tactical transport aircraft – i.e., C17, C130, and in the future A400M aircraft – to train pilots in techniques that may be used when operating from overseas airstrips that may be only semi-prepared or may not be totally secure from enemy action. There are proposals to construct a new, gravel runway to further facilitate this training. However, it is understood that MoD concerns relate to the airfield’s current use. In any case, the proposed new runway is further to the south of – and almost parallel to – the current runway, i.e., away from the zone of solar reflections and so the analysis of the current runway is a ‘worse case’ and will therefore be valid for future developments. 1.2.2.2 RAF Keevil Runways RAF Keevil has a single 1.8km long runway strip, Rwy 24/06, aligned from approximately northeast to southwest allowing take-off or landing in either direction (i.e., from Rwy 24 or Rwy 06). There are no known radio navigation aids at RAF Keevil, and so only visual (i.e., non-instrument) approaches are understood to be conducted there. No other runways are understood to be in use by military aircraft.

Rwy 06 inset touchdown point Rwy 24 Rwy 06 Threshold Threshold Rwy 24 inset touchdown point

Poulshot Lodge PV Array 4 RAF Keevil Glint & Glare Assessment

1.2.2.3 Receptor Points on Approach Paths to RAF Keevil Runways Approach paths were considered to each runway at RAF Keevil, assuming touchdown at the threshold and at inset touchdown points one third of the runway length in from the thresholds. Three approach paths to each touchdown point were considered: one along the runway’s extended centreline, and one each side from10° to right and left of the centreline (measured from the respective touchdown point). Glideslopes along each approach path were considered at 3°, 6°, and 9° above horizontal to consider the possibility of steep approaches by tactical airlifters. Receptor points on each approach path and at each glideslope were considered at 400m intervals (measured horizontally) from the touchdown point to a distance of 6km. Points in close proximity to, and to the north and east of, the solar farm on approach paths to Rwy 24 were not included for geometrical analysis: a different analytical method was used in this area. 1.2.2.4 Other Significant Receptors at RAF Keevil Other significant aviation receptors at RAF Keevil are: take-off paths, aerodrome circuit patterns, transiting traffic, and possible extreme flight profiles for tactical training. There are many variations for these phases of flight, especially when considering the versatility of helicopters operating at RAF Keevil. The general analysis considered in AARDVaRC’s previous report, combined with the analysis of approach path receptor points in this report provide enough data to consider other phases of flight. However, 20 additional points were chosen for each runway strip to support the discussion of effects of solar reflections on circuit traffic. These points were 0.5nm (926m) and 1nm (1,852m) either side of the runway centreline, 1,000ft above the mean threshold elevation, and are indicative of possible downwind leg flight paths in the aerodrome circuit. Circuit traffic may or may not pass through these precise points but they provide additional solar reflection data to consider. No specific analysis on other flight profiles was conducted; however, solar reflections effects on all these fight profiles are discussed based on the results of other analyses.

Poulshot Lodge PV Array 5 RAF Keevil Glint & Glare Assessment

1.2.2.5 Charts of RAF Keevil Receptors Points The chart below shows the receptor points close to RAF Keevil without a background map (for clarity), but marks the British National Grid eastings and northings at 1km intervals. This chart covers the same area as the next chart below (which has an OS 1:250,000 scale road map underlay). RAF Keevil : Aerial Receptor Points (with runway, solar farm and solar reflection zones marked) N

Limits of zone of Limits of near- solar reflections horizontal solar below 45° (P5) reflections zone (P2)

(P4) Poulshot Lodge solar farm

(P3)

Rwy 24/06 circuit receptor points (green diamonds, (P1) purple squares)

Approach path receptor points (blue dots – 2 dots outlined in red are used as examples)

Poulshot Lodge PV Array 6 RAF Keevil Glint & Glare Assessment

The next chart shows all receptor points extending further from the airfield, overlaid on an OS 1:500,000 scale background map. N RAF Keevil Receptor Points (with runway, solar farm and reflection zones marked)

(P5) (P2)

(P4)

(P3)

(P1)

Contains Ordnance Survey data © Crown copyright and database right 2013

Poulshot Lodge PV Array 7 RAF Keevil Glint & Glare Assessment

1.2.3 General Analysis 1.2.3.1 Overview In AARDVaRC’s previous report, it was shown that solar reflections would only affect surface-based, and most aerial receptors (unless high and/ or close to the solar farm) when directed at vertical angles from 5° below horizontal to 10° above horizontal. In some instances, aircraft operating at RAF Keevil may pass close enough to receive solar reflections directed upwards more steeply than this. Based on the author’s experience as a pilot and common experience of bright light sources, the most significant aerial reflections are assumed to be limited to those directed upwards at less than 45° above the horizontal. Those directed more steeply upwards will only be visible from very steeply below an aircraft – a direction generally of little interest to a pilot, even when an aircraft is banking in a turn, and one which will often be blocked by aircraft structure. This assumption is returned to later, but does not limit the analyses conducted, and the results of analysis are presented fully, irrespective of whether reflections are directed steeply upwards or not. 1.2.3.2 Sun Positions The position of the sun at the solar farm was predicted at 1-minute intervals for the whole of 2014 (a typical year in the life of the solar farm). These are shown graphically on the following chart (equivalent times are shown in Section 1.2.3.3 below).

Poulshot Lodge PV Array 8 RAF Keevil Glint & Glare Assessment

1.2.3.3 Times of Reflections In the chart below dates and times (rather than directions) of all reflections are shown, the vertical orange lines mark the dates when the change from Greenwich Mean Time (GMT) to British Summer Time (BST) and back can take place (the last Sunday in March and October respectively). It also indicates when the sun is up but illuminating the backs of the solar panels so no reflections are possible (the yellow areas). The colour coding allows some comparison of dates and times of reflections to the above chart showing their direction. Since solar reflections only occur when the sun is above the horizon, this chart indicates those times and so may be compared to sun positions in the chart at Section 1.2.3.2 above as well as the directions of reflections chart in Section 1.2.3.4 below.

Poulshot Lodge PV Array 9 RAF Keevil Glint & Glare Assessment

1.2.3.4 Directions of Reflections The following chart shows the directions in which solar reflections may be directed.

Times of reflections are not obvious from the above chart itself, so the approximate dates and times at 10 significant data points are shown (only 8 are marked because the equinox data points are virtually coincident). Note that in reality the top of the chart represents a point in space vertically above an observer (the zenith) but is stretched here to the full width of the chart, hence gaps appear between the minute-by-minute data points. 1.2.3.5 Near-Horizontal Solar Reflections ‘Near-horizontal’ solar reflections – taken to be from 5° below, to 10° above, horizontal – were analysed in AARDVaRC’s previous report. Most aerial receptor points – unless close to the solar farm (relative to height) – will only be subject to such near-horizontal reflections. Horizontal (azimuthal) arcs containing all such near-horizontal reflections were calculated and the dates to establish when such reflections could be seen by general ground-based receptors. Dates and times of near-horizontal reflections that affect surface receptors and most aerial receptors are marked as the light green and light brown areas on the chart in Section 1.2.3.3 above. The directions of these reflections can be read from the chart in Section 1.2.3.4 above: they are restricted to arcs to the east and west (shown later on the chart in Section 1.2.3.8 below). 1.2.3.6 Solar Reflections up to 45° Above Horizontal For aerial receptors in closer proximity to the solar farm (i.e., within 1.8km for aircraft 1,000ft above the solar farm’s level, within 5.2km for aircraft 3,000ft above it) a deeper vertical band of solar reflections was considered: reflections between the horizontal plane and 45° above it. Again, horizontal (azimuthal) arcs containing all such

Poulshot Lodge PV Array 10 RAF Keevil Glint & Glare Assessment

reflections within that altitude band were calculated and the dates and times that solar reflections could be observed by receptors within that band established. Dates and times of reflections up to 45° above horizontal that may affect aerial receptors that are either very high or in close proximity to the solar farm are marked as the light and dark brown areas on the chart in Section 1.2.3.3 above. The directions of these reflections can be read from the chart in Section 1.2.3.4 above: they are again restricted to arcs to the east and west (shown later on the chart in Section 1.2.3.8 below). This zone may be described as the ‘zone of significant aerial reflections, as solar reflections at steeper angles than this will generally be insignificant as discussed next. 1.2.3.7 Solar reflections More Than 45° Above Horizontal Aircraft could only subjected to solar reflections directed more steeply upwards than 45° in the immediate vicinity of solar panels, i.e.: at 3,000ft above the solar farm, only within 915m of solar panels; at 1,000ft above it, within 305m of solar panels; at 500ft, within 152m of panels; and at 200ft, within 61m of solar panels. Furthermore, reflections directed upwards more steeply than 45° could only be viewed from well below an aircraft (i.e., from more than 45° below horizontal), are likely to be blocked by aircraft structure and/ or is a region of low interest to a pilot, so he/ she will spend little time looking there. Therefore such reflections are generally not significant. 1.2.3.8 Solar Reflection Zones Chart The solar reflection zones are shown in the chart below with the runway layout, the solar farm boundary and receptor points (marked as the many blue dots for approach paths, 10 mauve squares for the northern circuit points and 10 green diamond shapes for the southern circuit points) overlaid on an OS 1:50,000 scale Landranger map for comparison to real features on the ground.

Limits of significant aerial reflections zone

Limits of near-horizontal reflections zone

Poulshot Lodge PV Array site boundary

1.2.4 Detailed Analysis 1.2.4.1 Geometrical Analysis The geometries between all solar farm points and each aerial receptor points were calculated as altitude and azimuth angles. These geometries were then combined to give a maximum and minimum azimuth and altitude angle for reflections affecting that point, and a 1° buffer in azimuth and altitude applied. This gave a range of azimuths and altitudes to compare to the calculated solar reflection geometries to establish dates and times when reflections may occur at each receptor point (and associated properties, e.g., the associated sun position) so that their effect on the operation of the airfield could be considered. Receptor points were grouped so that approaches to each runway could be considered together (although it was convenient to break the approach to Rwy 24 into 2 parts to be considered separately). Receptor points for circuits to were also considered together; no surface-based receptors (e.g., a control tower, if there is one) were considered as these were confirmed to be unaffected by solar reflections in AARDVaRC’s earlier analysis. The grouping of receptor points for each approach is such that the results for each approach are worse than would be encountered on any single approach, since they cover multiple approach paths and glideslopes. 1.2.4.2 Other Analyses Where low overflight of the solar farm would occur, there is a very large variation in solar-panel-to-aircraft geometries (so solar reflections could occur in almost any possible direction – see the chart in Section 1.2.3.4 above). In such cases (i.e., close to, and to the north and east of the solar farm, on approaches to Rwy 24) geometrical analysis was not conducted. Instead, it was assumed that all possible solar reflection geometries may coincide with solar-panel-to-aircraft geometries as an aircraft crosses over the solar farm, necessitating another measure to process the data in a meaningful manner (since most solar reflection geometries will be irrelevant, e.g., reflections from behind an aircraft). Based on the author’s experience as a pilot and his understanding of the glint and glare phenomenon, as well as from common experience of bright sunlight, it was assumed that solar reflections directed steeply upwards (i.e., at more than 45° above the horizontal plane) or more than 45° away from the aircraft’s flight path (i.e., the ‘approach vector’: for these purposes this is equivalent to 45° angular distance from the aiming point on the runway as viewed by the pilot) are of low significance. This is pertinent for the approach paths Rwy 24 when aircraft are in close proximity to the solar farm, and this is considered when an approaching aircraft is approximately 4km or more from the runway (when closer to the runway, such an aircraft would be flying away from the solar farm and any reflections). 1.2.5 Results 1.2.5.1 Overview Times of solar reflections for the various receptor groups (i.e., runway approaches and circuits), are shown. Detailed analysis of phases of flight other than these was not undertaken since sufficient analysis had been conducted to support thorough discussion of the impact of solar reflections on them.

Poulshot Lodge PV Array 12 RAF Keevil Glint & Glare Assessment

1.2.5.2 Rwy 24 Approach At or within 3.6km of the Rwy 24 Threshold. When an aircraft is at or within 3.6km of the runway on approaches to Rwy 24, there is no possibility of solar reflections reaching it, even from behind the aircraft. At slightly greater distances, any reflections affecting an aircraft will be from behind it and hence insignificant (although this analysis is applied later to aircraft taking-off from Rwy 06, i.e., in the opposite direction). Beyond 3.6km of the Rwy 24 Threshold. Beyond 3.6km from touchdown, aircraft may pass through solar reflections briefly (in a similar manner to en-route or transiting traffic), potentially from one of many directions (depending on flight path, and of course the date and time of day). The most significant reflections will be those directed upwards at less than 45° above horizontal. However, for all Rwy 24 approaches the much brighter sun will always be shinning from closer to the aircraft’s approach vector (i.e., the runway aiming point ahead which is the pilot’s focus of vision). This will render effects of the much dimmer reflections insignificant. 1.2.5.3 Rwy 06 Approach Aircraft operating within the approach geometries considered will never receive solar reflections when approaching to land on Rwy 06. 1.2.5.4 Other Phases of Flight Take-Off Any reflections will be behind aircraft taking off on Rwy 24. No reflections will be visible to aircraft taking off from Rwy 06 until nearly 4km from the airfield (assuming the departure is almost straight ahead). At this point, effects will be similar to those for en-route or transiting traffic, well below the aircraft and the pilot’s focus of view will be the sky above the aircraft into which it is climbing. Circuit Traffic The only circuit that may be affected by solar reflections is the circuit on the north side of the airfield, and then only if wide and/ or high circuits are flown. It is not known, but is considered unlikely that circuits are flown on this side of the airfield, as it is relatively densely populated compared to the south side of the airfield. If northern circuits are flown, effects will be insignificant for aircraft using Rwy 06, as they would be flying away from the solar farm when reflections may occur. Northern circuits flown to Rwy 24 may be affected by solar reflections very early in the morning as shown in this chart – and slightly later if circuits are flown at higher levels. Any effects would be similar to those for en-route or transiting traffic. It would be exceptional for such flying training to take place so early in the morning.

Poulshot Lodge PV Array 13 RAF Keevil Glint & Glare Assessment

Transiting Traffic This author has never known the MoD to raise concerns regarding solar farms due to effects on en-route or transiting traffic, so it is appears that the MoD accepts that effects of solar reflections form solar farms on such traffic are insignificant. The following points are pertinent.  Since reflections from all panels are parallel to each other, an aircraft flying straight and level could be subjected to solar reflections only for the time it takes to fly across the length of the solar farm. As the solar farm is approximately 1km across in any direction, it would take an aircraft flying at approximately 120kts (62m per second) less than 20 seconds to cross through any reflections. Aircraft flying faster, or climbing or descending will generally take less time to pass through any reflections.  Transiting traffic will normally be relatively high with any reflections from well below and often blocked by the aircraft structure (or come from behind the pilot).  Solar reflections will not come from a direction of specific and immediate interest to the pilot (there is no reason why a pilot would normally need to stare in the direction of the solar farm whilst in in transit, reducing any nuisance from solar reflections).  Such aircraft will not be descending towards the solar farm so effects of solar reflections will be transient.  For aircraft transiting at 3,000ft above the solar farm (or less), solar reflections directed upwards at more than 45° can only affect an aircraft within 1km of the solar farm boundary, perhaps in any direction. Otherwise the zones for solar reflections shown in Section 1.2.3.8 apply, i.e.: less than 10° (‘near horizontal zone’) beyond 5.7km – about 3nm – from the solar farm (for aircraft at 3,000ft); or the ‘less than 45°’ zone within this distance (for aircraft at 3,000ft). Extreme Flight Profiles Extreme flight profiles may be used in training for tactical situations, e.g., a very steep, rapid descent from, say, 20,000ft at an angle of perhaps 45° below the horizontal. It is unlikely that solar reflections would occur in in front of a pilot conducting such a manoeuvre towards Rwy 24: if so, they would be from well below the aircraft. Towards Rwy 06, it is unlikely that the aircraft would be high enough and far enough north of the runway’s extended centreline to receive solar reflections. If they did occur, they would be in the far distance (and hence dim), and well ‘above’ the runway) as viewed by the pilot). If they did occur, they would last a very short time as the aircraft rapidly descended through them (i.e., less than 5 seconds), so would probably not be noticed.

Poulshot Lodge PV Array 14 RAF Keevil Glint & Glare Assessment

1.2.6 Impact Assessment 1.2.6.1 Overview The dates and times and other significant characteristics of solar reflections from the solar farm as they affect each runway approach are discussed here. Impact was assessed against a 10-point numerical scale (11 point in theory but in practice only 10 points since 0, for ‘zero impact’, is hard to justify for any nature of development and would not normally be used). The definitions of each point on the scale are as shown below. Level of Impact Definition Impact Description Absolutely zero impact. In practice, such a level of impact from any nature of development 0 is virtually impossible to justify, so this level is normally unused. Nil/ Effects not observable unless significant effort is taken to notice them, so there is 1 no impact practically no impact. Effects may be observable with little effort but with no practical impact on the task in hand 2 and will quickly be dismissed from an observer’s awareness. Observable effects that may persist in an observer’s consciousness but have no effect on 3 his/ her execution of the task in hand. Negligible Effects that are readily and continuously noticeable that do not affect the execution of the 4 task in hand. Effects that may cause an observer to take some almost unconscious action with no 5 noticeable effect on the task in hand. Small Effects that may cause an observer to take some conscious action that does not interfere 6 with the execution of the task in hand. 7 Effects that require a noticeable change in the execution of the task in hand to manage. Medium Effects that require a deliberate and noticeable change in the execution of the task in hand 8 to manage. 9 Effects that require a considerable effort to manage in the execution of the task in hand. Large 10 Effects that require an excessive effort to manage in the execution of the task in hand.

The baseline used for this impact assessment is consideration of landing directly towards the sun. Based on the above table, the sun’s impact on the approach and landing would be Level 6 (small) and there is no question that safety is affected: it is a relatively common occurrence and at RAF Keevil will be occasionally necessary. The impact of solar reflections from dark panels must be less than this, and where the much brighter sun is close (in angular terms) either to any reflecting panels or to the runway ahead – or other point where a pilot’s attention is focussed, it’s brightness will reduce the impact of reflections further. The assessed levels of impact are given below. 1.2.6.2 Approach and Landing Rwy 24 At more than approximately 4km from touchdown, the sun will be close to any reflecting panels, or to the runway ahead (and invariably closer to the runway ahead than any significant solar reflections) so impact is assessed as Level 3 (negligible). Within 4km from touchdown, the solar farm (and any reflections) will be from below or behind the aircraft and impact is assessed as Level 1 (nil) applies for the remainder of the approach. Rwy 06 Solar reflections could only affect aircraft approaching this runway if they are excessively displaced to the north of the runway’s extended centreline. Such

Poulshot Lodge PV Array 15 RAF Keevil Glint & Glare Assessment

reflections (if they ever occur) would be in the far distance, hence well attenuated, so impact is assessed as Level 2 (nil). 1.2.6.3 Other Phases of Flight In no instance will they be greater than Level 4 (negligible). The worst case impacts will be for transiting aircraft (i.e., Level 4, negligible) and/ or during the downwind leg of wide and high circuits to Rwy 24 on the north side of the airfield, and will be similar to what may occur to any transiting aircraft. Other impacts will be less than this. 1.2.6.4 Overall Impact on RAF Keevil Given the impact levels of the most significant impacts on aircraft (most notably on those approaching to land on some runways), the overall impact of solar reflections from the proposed solar farm on RAF Keevil is assessed to be no greater than Level 4 (negligible). This impact assessment ignores the limited time in which the solar reflections can occur, only considering the impact level when solar reflections occur. Extending this to consider the overall impact of solar reflections across airfield operations through an entire day (and year), the overall impact must be even less than this due to the limited periods of impact (and the requirement for the sun to be shining). 1.3 Conclusions 1.3.1 General Effects of Solar Reflections 1.3.1.1 General Characteristics Solar reflections are commonplace occurrences for most people either from wet roads, expanses of water, or windows and mirrors of cars and buildings. Solar cells are designed to absorb light to generate electricity, not reflect it, and so are much less reflective than other sources of solar reflection. Solar reflections can only occur when the sun is shining. It has no significance when the sun appears very close to – that is, in almost the same direction as – the reflecting object as seen by an observer (i.e., the observed angle between the sun and its reflection is close to 0°) as the much brighter sun will completely mask any reflections and the observer’s eyes will be attuned to brightness when looking in that direction, thus reducing the observed intensity of any reflections. Conversely, solar reflections are at their worst when an observer is facing the reflecting object, is in shade from the bright sun so that his/ her eyes aren’t attuned to brightness, and the sun is behind the observer (i.e., the angle between observed reflections and the sun is close to 180°). Solar reflections from PV panels may typically have intensities of 20 W/m² (about 2,000 lux). Ambient light levels on a sunny day in shade, but illuminated by the entire blue sky, are typically 20,000 lux so the worst problem that may be caused by reflections from solar panels is a nuisance from looking at or near them. 1.3.1.2 Characteristics of Solar Reflections from Poulshot Lodge Solar Farm Significant Aerial Reflections Solar reflections from the horizontal plane up to 45° above it may be significant for aircraft operating at RAF Keevil near the proposed solar farm. Reflections within this vertical band are described as ‘significant aerial reflections’. Such reflections can occur all year round:

Poulshot Lodge PV Array 16 RAF Keevil Glint & Glare Assessment

 in the mornings at receptors to the west of the solar farm from 05:59 to 09:38 Greenwich Mean Time (GMT), or 06:59 to 10:38 BST (after the last Sunday in March until the last Sunday in October);  in the evenings at receptors to the east of the solar farm from 14:36 to 18:17 GMT, or 15:36 to 19:17 BST (after the last Sunday in March until the last Sunday in October). Aircraft would need to be in extremely close proximity to the solar farm to receive reflections at steeper angles than 45°, although in some cases such reflections may occur in any azimuth at certain times. Aircraft 1000ft above the solar farm could only observe such reflections within 305m of its boundary; this increases to within 915m for aircraft 3,000ft above it. The first and last times of reflections on any day are later and earlier, respectively, than for ‘near-horizontal reflections because in this case, reflections below the horizontal plane are not considered. 1.3.2 Effects of Solar Reflections on RAF Keevil Solar reflections will never affect aircraft as they land or take-off at RAF Keevil. Aircraft operating in the wider vicinity, e.g., in a wide and high circuit to the north of the airfield or on approach at some distance (more than 4km) from touchdown at RAF Keevil. Such aircraft may receive solar reflections at various times. Overall solar reflection impacts for aircraft operating at RAF Keevil are assessed as negligible; no specific impact on a specific operation was assessed to be greater than negligible. This does not consider the very limited times when impacts are possible, and the requirement for the sun to be shining: true impacts on the operations at RAF Keevil will therefore be even less than those assessed here. All Civil Aviation Authority guidance on solar farms near aviation sites is met by the Poulshot Lodge proposal.

Poulshot Lodge PV Array 17 RAF Keevil Glint & Glare Assessment

2 Contents 1 Executive Summary ...... 2 1.1 Background ...... 2 1.2 Assessment Method ...... 3 1.3 Conclusions ...... 16 2 Contents ...... 18 3 Glossary of Terms and Definitions...... 19 4 Introduction ...... 22 4.1 Background ...... 22 4.2 Glint and Glare Effects on Aviation ...... 23 4.3 Map Datums and Time Zones ...... 26 5 Poulshot Lodge Photovoltaic Array Development ...... 27 5.1 The Development...... 27 5.2 Site Information ...... 27 5.3 Site Boundary ...... 27 5.4 Initial Analysis ...... 29 6 Receptors at RAF Keevil Aerodrome ...... 30 6.1 Background ...... 30 6.2 Data Sources ...... 30 6.3 Runway Nomenclature ...... 30 6.4 Runway Data ...... 30 6.5 Surface Receptors ...... 31 6.6 Airfield Chart ...... 31 6.7 Runway Approach Paths ...... 32 6.8 Other Phases of Flight ...... 32 6.9 Identification of Solar Reflection Receptors ...... 32 7 Glint and Glare Assessment ...... 38 7.1 Outline ...... 38 7.2 Sun Data ...... 38 7.3 Solar Reflections ...... 40 7.4 General Effects of Solar Reflections ...... 41 7.5 Outline Assessment Method ...... 41 7.6 General Analysis ...... 42 7.7 Specific Receptor Analysis ...... 42 8 Conclusions ...... 57 8.1 General Effects of Solar Reflections ...... 57 8.2 Effects of Solar Reflections on RAF Keevil ...... 57 Appendix 1 – Solar Farm, Reflection Zones and RAF Keevil Receptor Points Chart ...... A1-1

Poulshot Lodge PV Array 18 RAF Keevil Glint & Glare Assessment

3 Glossary of Terms and Definitions Altitude. The astronomical term for the vertical angle from an observer to a celestial object, measured from the horizontal plane. In this report, ‘altitude’ is also used to describe the vertical direction of reflected light (0° is horizontal, 90° is vertically upwards). Note that there is scope for confusion between the astronomical and aviation terms, ‘altitude’. In aviation terms it is the vertical distance of an aircraft above mean sea level, normally measured in feet using a barometric altimeter. Unless otherwise specified the astronomical definition is used: the aviation term is identified by referring to ‘aircraft altitude’ or ‘flight altitude’. Where there remains scope for confusion, the term ‘altitude angle’ or ‘astronomical altitude’ may also be used. amsl. Above Mean Sea Level, a datum for elevations or aircraft altitudes. For the purposes of this assessment, it is synonymous with ‘AOD’. AOD. Above Ordnance Survey Datum, a datum for elevations. For the purposes of this assessment, it is synonymous with ‘amsl’. Azimuth. The term for the horizontal angle from an observer to an object, measured clockwise (as viewed from above) relative to True North. BNG. British National Grid, a Cartesian coordinate system of Eastings and Northings for Great Britain based on the Ordnance Survey Great Britain 1936 (OSGB36) datum. BST. British Summer Time, one hour ahead of GMT (see below) and used as local time in the UK after the last Sunday in March and before the last Sunday in October. Circuit. An Airport’s traffic pattern in which an aircraft takes-off, turns to fly ‘downwind’ (i.e., parallel to the runway in the reverse direction for take-offs and landings) to position itself for final approach and landing. Aircraft may join or leave the circuit at any point, or may remain in the circuit for take-off and landing practice. Equinox. At the equinoxes, the sun is on a line through the centre of the Earth perpendicular to its rotational axis, i.e., it is the point in time that it crosses the extended plane of the Equator (the ‘equatorial plane’). There are 2 equinoxes each year, in Spring – the ‘Vernal Equinox’ (occurring about March 21st in the northern hemisphere), and in Autumn – the ‘Autumnal Equinox’ (occurring about September 23rd in the northern hemisphere). An observer at the Equator (0° latitude) – and at an appropriate longitude – will see the sun’s geometric centre rise due east and set due west of him/ her, or pass directly overhead, at either equinox. At other latitudes, this is not the case due to atmospheric refraction normally causing the sun to appear slightly higher than its true position. Also, actual sunrise and sunset times are for the highest point of the sun, not its centre. The equinox dates and times for 2014 are March 20th at 16:57 (UTC), and September 23rd at 02:29 (UTC); note that the dates may vary by 1-2 days from year to year. Note that although an equinox is defined specifically as an instant in time, the date on which the equinox occurs is commonly referred to as ‘the equinox’. Glare. Reflected diffuse lighting, e.g., from bright sky around the sun (glare may describe a range of brightness of reflections). Glare will generally be much dimmer than glint. See Glint below for an alternative definition of glare that may also be used elsewhere. Glint. Specular (direct) reflection of the sun. This is the principal issue regarding the potential for nuisance to an observer. Under certain conditions, e.g., a thin layer of high cloud dimming the sun, it may be difficult to distinguish between glint and glare (although the peak intensity of the glint would be reduced).

Poulshot Lodge PV Array 19 RAF Keevil Glint & Glare Assessment

Whilst ‘glint and glare’ are often referred to together, the main issue with solar farms is normally only ‘glint’. This assessment concerns itself specifically with ‘glint’ although the term ‘glint and glare’ may be used occasionally. Note that other sources use different definitions of these terms: they define ‘glint’ as short-lived periods of brightness, and ‘glare’ as prolonged periods of brightness. Both definitions may be found in different places on the internet, so care should be taken to avoid confusion. GMT. Greenwich Mean Time, for the purposes of this report it may be considered equivalent to UTC (see below). GMT is local time in the UK before the last Sunday in March and after the last Sunday in October (see BST above). Grid North. Grid North is the [northerly] direction of the easting grid lines (i.e., that run north to south) of the grid system in use: for this report, the British National Grid (BNG) system is used, and at all points on the British National Grid, Grid North is parallel to True North at the 2° West meridian of longitude. Note that the difference between Grid North and True North varies from place to place. The local difference is stated in this report. Hand span at arm’s length. An easily visualised reference for observed angular arcs: an observer’s hand held at arm’s length with the fingers comfortably spread is assumed to subtend (typically) an arc of 15° to 20° between the tips of the small finger and thumb. Although a useful indicative measure, for the avoidance of doubt, the reader should ‘calibrate’ his/ her own hand span at arm’s length against known angles before using this angular reference. Local time. In the UK, local time is GMT from January 1st until early on the last Sunday in March and from early on the last Sunday in October until December 31st each year; otherwise local time is BST. Normal. A line perpendicular to a planar surface (such as a solar cell). PV Farm. A large-scale installation to convert sunlight into electricity using an array of photovoltaic (PV) cells. In this report the term may be used synonymously with ‘Solar Farm’ or ‘Solar Park’. Solar Farm. A large-scale installation to convert sunlight into usable energy, normally electricity. In this report the term may be used synonymously with ‘PV Farm’ or ‘Solar Park’. Solar Park. This term may be used synonymously with ‘Solar Farm’ or ‘PV Farm’. Solstice. The point in time when the sun is furthest from the equatorial plane, either to its north (Summer Solstice in the northern hemisphere) or south (Winter Solstice in the northern hemisphere). The solstices are when the sun appears to stop moving away from the equatorial plane, and begins moving back towards it. There are 2 solstices each year, sometimes referred to as ‘midsummer’ or ‘midwinter’, around June 21st and December respectively. An observer at the Tropic of Cancer or of Capricorn (approximate latitudes 23° North and South respectively) at the appropriate solstice (e.g., the northern hemisphere’s summer solstice at the Tropic of Cancer) – will see the sun’s geometric centre rise due east and set due west of him/ her, or pass directly overhead, at either equinox. At other latitudes, this is not the case due to atmospheric refraction normally causing the sun to appear slightly higher than its true position. Also, actual sunrise and sunset times are for the highest point of the sun, not its centre. The solstice dates and times for 2014 are June 21st at 10:51 (Summer Solstice in the northern hemisphere – when the sun is at its highest), and December 21st at 23:03 (Winter Solstice in the northern hemisphere – when the sun is at its lowest); note that the dates may vary by 1-2 days from year to year. Note that although a solstice is defined specifically as an instant in time, the date on which the solstice occurs is commonly referred to as ‘the solstice’.

Poulshot Lodge PV Array 20 RAF Keevil Glint & Glare Assessment

Specular reflection. Specular, or direct, reflections may be observed from a polished mirror in which the angle of reflection is equal to the angle of incidence but on the opposite side to (and hence, in the same plane as) the normal line, as shown in the diagrams below. ‘Non-specular reflections’ are scattered reflections which may occur from a rough surface which causes the scattering effect. Specular Reflections Normal

Incident light Reflected light

Angle of Angle of reflection incidence (= angle of incidence)

Smooth reflecting surface

Example of specular reflections in which the angle of reflection is the same as (but opposite) the angle of incidence – measured relative to the Normal line.

Non-specular Reflections Normal

Reflected light

Angle of incidence (not equal to angles of reflection)

Treated surface to minimise glint from specular reflections

Example of non-specular reflections in which the angle of reflection is not the same as the angle of incidence: e.g., due to surface treatment of solar cells, such as a ‘roughening’ of the surface to minimise glint effects. True North. The [northerly] direction of the meridian of longitude through a certain point, i.e., the direction of the northern polar axis. Note that the difference between Grid North and True North varies from place to place. The local difference is stated in this report. UTC. Coordinated Universal Time, for the purpose of this report, it is equivalent to GMT (see above). Unless otherwise specified, all times in this report are UTC. Zenith. The point in the sky directly (i.e., vertically) an observer.

Poulshot Lodge PV Array 21 RAF Keevil Glint & Glare Assessment

4 Introduction 4.1 Background The Ministry of Defence (MoD) has objected to the proposed Poulshot Lodge solar farm on the south side of the A361 road, 4km west of Devizes (planning reference: 13/05244/FUL) on the grounds that it has insufficient information regarding the phenomenon of glint and glare effects on RAF Keevil aerodrome, approximately 6km to the southwest of the site (measured from the site centre to the mid-point of the main runway). This report follows on from – and should be read in conjunction with – AARDVaRC’s earlier general glint and glare assessment (Glint and Glare Assessment for the Poulshot Lodge Farm Photovoltaic Array, dated 26 June 2013. The scope of this earlier report did not include receptor specific analyses, in the report it was observed in Section 6.5.2.5 Keevil Airfield that: ‘…Keevil Airfield is 6km southwest of the site. It is well outside the glint zone for near- ground level receptors, so will be unaffected by glint from the proposed solar farm.’ The MoD noted in its letter of objection (reference: D/DIO/43/2/44 (2013/1105), dated 21 November 2013) that: ‘The application is supported by an assessment of potential impacts of glint and glare. This assessment identifies RAF Keevil in the assessment but only considers it as an area of land. The assessment does not evaluate the potential affects [sic] of glint and glare emissions upon military aviation activities conducted at this airfield. ‘Therefore, at this stage, the MOD considers that the application does not adequately demonstrate that the proposed solar array will not cause a hazard to air traffic safety as a potential source of glare in proximity to an aerodrome approaches. As such, there is insufficient information provided in the application to enable us to determine our statutory safeguarding position in relation to this application.’ ‘The Civil Aviation Authority (CAA) has published interim guidance on the development of solar arrays in the vicinity of aerodromes in the UK. It is recommended that the applicant considers the technical guidance and assessment methodologies referred to in this.’ The CAA document to which the MoD letter refers to is shown at Annex A to this document. This itself notes the requirements of the Air Navigation Order as may be pertinent to solar farms, and refers to a United States Federal Aviation Administration (FAA) document – shown at Annex B to this report – which by the FAA’s own advice is unreliable (as discussed in detail Section 4.2.2 below) and conclusions should not be drawn from that document on its own. Otherwise the CAA document gives virtually no guidance on assessment methodologies for glint and glare as they may be pertinent to aircraft. This report expands on the work conducted for the earlier study and considers new receptor-specific analysis of flight paths around RAF Keevil aerodrome to establish the dates and times and other pertinent information when solar reflections from the proposed solar farm may affect military aircraft operating at RAF Keevil. It is noted that there is little published information ion the aerial activities practiced there by military units, and in the time available, no information has been received. Therefore, assumptions have had to be made as to the military activities conducted at RAF Keevil. These are noted in this report.

Poulshot Lodge PV Array 22 RAF Keevil Glint & Glare Assessment

4.2 Glint and Glare Effects on Aviation 4.2.1 Overview There is little formal published guidance on the assessment of glint and glare from solar farms, either in the UK or overseas. It is noted that the definitions given by some aviation bodies for ‘glint’ and ‘glare’ differ from those often used in the solar power industry. ‘Glint’ may be defined either as the specular reflection of light, or as short duration bright flashes of light (or similar), from an object; ‘glare’ may be defined either as non-specular (diffuse) reflection of light, or as longer periods of bright light (or similar), from an object. Hence it is considered better to avoid these terms in a technical discussion. Here, the limited published guidance from the USA and the UK is considered. 4.2.2 United States: Federal Aviation Administration In the United States, the Federal Aviation Administration (FAA) has published some guidance in the document Technical Guidance for Evaluating Selected Solar Technologies on Airports, dated November 2010, although this document is caveated with the following text: ‘NOTE: As of October 23, 2013, the FAA is reviewing multiple sections of the "Technical Guidance for Evaluating Selected Solar Technologies on Airports" based on new information and field experience, particularly with respect to compatibility and glare. All users of this guidance are hereby notified that significant content in this document may be subject to change, and the FAA cautions users against relying solely on this document at this time. Users should refer instead to the Interim Policy (http://federalregister.gov/a/2013-24729).’ [A previous version of the disclaimer read: ‘NOTE: As of June 26, 2012, the FAA is reviewing Section 3.1.2 ("Reflectivity") of the "Technical Guidance for Evaluating Selected Solar Technologies on Airports" based on new information and field experience. All users of this guidance are hereby notified that significant content in this section may be subject to change, and the FAA cautions users against relying solely on this section at this time.’ Section 3.1.2 of this document considers the technical aspects of solar reflections and is discussed here.] The webpage the current disclaimer references does not provide much useful guidance, but offers a free online glint analysis tool and recommends analysis of solar reflections on a minute-by-minute basis for a whole year (as AARDVaRC has done). A reader may wish to use that tool, but AARDVaRC is unable to recommend the use of its results as evidence of the acceptability or otherwise of solar farm developments near airports as it has no knowledge of the algorithms used to generate its results. The tool also seems to be related to the advice in the above document which the FAA cautions against using. In the FAA document, there appear to be serious flaws in the discussion of the effects of solar radiation, discussed briefly here. The FAA document states: ‘Often 1000W/m2 is used in calculations as an estimate of the solar energy interacting with a panel when no other information is available.’ [Note: W/m2 (watts per square metre) is a unit of intensity, i.e., the power passing through a unit of area]

Poulshot Lodge PV Array 23 RAF Keevil Glint & Glare Assessment

And: ‘According to researchers at Sandia National Lab, flash blindness for a period of 4-12 seconds (i.e., time to recovery of vision) occurs when 7-11 W/m2 (or 650- 1,100 lumens/m2) reaches the eye.’ [Note: 1 lumen/m2 (lumen per square metre) ≡ 1 lux] And: ‘Today’s panels reflect as little as 2% of the incoming sunlight depending on the angle of the sun and assuming use of anti-reflective coatings. Using the previously mentioned value for solar irradiance, this would mean roughly 20 W/m2 are reflected off of a typical PV panel.’ Apparently reflections from solar panels are bright enough to cause flash blindness lasting up to 12 seconds. However, statements in Section 3.1.2 Reflectivity of the FAA document give reference to: ‘Ho, Clifford, Cheryl Ghanbari, and Richard Diver. 2009. Hazard Analysis of Glint and Glare From Concentrating Solar Power Plants. SolarPACES 2009, Berlin Germany. Sandia National Laboratories’. This reference states that the information about ‘flash blindness’ comes from research into how a pilots’ eyesight might be affected by bright flashes from nuclear detonations which is rather divorced from reflections from dimly reflecting solar panels. The values of light levels quoted are meaningless to most people so it is useful to put the data provided into context with a more commonplace example: a cricket match can be halted due to ‘bad light’ when the light levels fall to below 1,000 lux 1 (typically) at the wicket (with 650 lux acceptable at the boundary). Yet, the FAA document states that these [bad light] levels can cause ‘flash blindness lasting up to 12 seconds’. It seems reasonable to construe that those light levels can only cause flash blindness in the worst case: i.e., a flash occurring directly in front of eyes that have been in total darkness for some time beforehand. This does not correspond in any way to reflected sunlight from dark solar panels which can only occur with high ambient light levels. The realisation of this may be one reason why the FAA is unwilling to be held accountable for any conclusions drawn from its document. When in shade but otherwise illuminated by an entire blue sky, the light level is typically 20,000 lux; the midday light levels with an overcast sky are typically 10,000 to 25,000 lux (data from the Wikipedia website 2), i.e., 215 W/m² (using a standard factor of 93 lux per W/m2 for sunlight 3 that is equal – or at least very close – to the factor used in the FAA document). This is much higher than the 20W/m² that is quoted as the level of reflected light from solar panels. It is virtually impossible to conceive how ‘flash blindness’ could be experienced from solar panel reflections given such ambient light levels. Considering the intensity of reflections from solar panels and the high ambient light levels that must be present when they may be observed, the worst case effects will be

1 http://www.opticianonline.net/Articles/2003/09/12/6563/Lighting+the+sports+arena.htm, last accessed 27 January 2014. 2 http://en.wikipedia.org/wiki/Daylight, last accessed 27 January 2014. 3 http://en.wikipedia.org/wiki/Luminous_efficiency, last accessed 27 January 2014.

Poulshot Lodge PV Array 24 RAF Keevil Glint & Glare Assessment

the nuisance caused by looking at, or near, reflecting panels. This will always be much less significant than looking into the bright sun (perhaps when driving a car or landing an aircraft towards it, neither of which are considered to be safety issues). The disclaimer also states that new information and field experience has led to the review of the document, suggesting that it is perhaps the solar energy industry that is leading the development of appropriate ways to assess reflections from solar panels. There also seems to be little urgency given to understanding the issue of reflectance from solar panels for aviation by the FAA (since the document has been ‘under review’ for some time) suggesting that there is little concern regarding it. In summary, this information appears to be based on extracts from academic research for entirely different purposes without consideration of its applicability to solar reflections from dark solar panels. Furthermore, this document is meant for facilities actually on airfields (whereas the proposed solar farm is not on an airfield), and conclusions from the information in this document are not supported by the FAA itself. 4.2.3 United Kingdom: Civil Aviation Authority In the UK, the Civil Aviation Authority (CAA) has published the document Interim CAA Guidance - Solar Photovoltaic Systems, dated 17 December 2010. It seems that, along with the FAA in the United States, the CAA does not consider the addressing of this issue to be particularly urgent. This document refers to the FAA document, but offers limited guidance and is principally for airports considering on-site solar farms. It refers to the legal requirements of the Air Navigation Order (ANO) and specifically notes that Local Planning Authorities should be cognisant of the following articles of the ANO with respect to any solar photo-voltaic development regardless of location.  Article 137 – Endangering safety of an aircraft: A person must not recklessly or negligently act in a manner likely to endanger an aircraft, or any person in an aircraft.  Article 221 – Lights liable to endanger: (1) A person must not exhibit in the United Kingdom any light which: (a) by reason of its glare is liable to endanger aircraft taking off from or landing at an aerodrome; or (b) by reason of its liability to be mistaken for an aeronautical ground light is liable to endanger aircraft. (2) If any light which appears to the CAA to be a light described in paragraph (1) is exhibited, the CAA may direct the person who is the occupier of the place where the light is exhibited or who has charge of the light, to take such steps within a reasonable time as are specified in the direction: (a) to extinguish or screen the light; and (b) to prevent in the future the exhibition of any other light which may similarly endanger aircraft. (3) The direction may be served either personally or by post, or by affixing it in some conspicuous place near to the light to which it relates. (4) In the case of a light which is or may be visible from any waters within the area of a general lighthouse authority, the power of the CAA under this article must not be exercised except with the consent of that authority.

Poulshot Lodge PV Array 25 RAF Keevil Glint & Glare Assessment

 Article 222 – Lights which dazzle or distract: A person must not in the United Kingdom direct or shine any light at any aircraft in flight so as to dazzle or distract the pilot of the aircraft. These are all considered in this report. 4.2.4 Examples of Solar Farms near Airports There are a number of known solar farms at or near airports. A review 4 of such instances from around the world written for the Caddington solar farm with many examples from the United States, dated 22 December 2010 (since when there have been a number of further developments, particularly in the UK, e.g., on Gatwick Airport, near Newquay Airport, and near Exeter Airport) shows various examples demonstrating the general compatibility of solar farms and aviation. 4.3 Map Datums and Time Zones Unless otherwise stated, locations are given as British National Grid OSGB36 Eastings/ Northings coordinates and azimuths/ bearings are given as Grid bearings. The entire UK is in the Greenwich Mean Time (GMT) time zone, but daylight saving time (i.e., British Summer Time, BST) is currently used from the last Sunday of March to the last Sunday of October. Unless otherwise specified, all times in this report are given in coordinated universal time (UTC) – equivalent to GMT for the purposes of this assessment. The term ‘local time’ refers to BST when daylight saving time is in use, otherwise it refers to GMT. Where specific examples are used, the dates for daylight saving time are for 2014, planned to be the first full year of operation of the proposed PV farm.

4 http://www.emsraynerenewableenergy.co.uk/downloads/GlintGlareStudyReview.pdf, last accessed 27 January 2014.

Poulshot Lodge PV Array 26 RAF Keevil Glint & Glare Assessment

5 Poulshot Lodge Photovoltaic Array Development 5.1 The Development The proposed Poulshot Lodge photovoltaic array is discussed in AARDVaRC’s previous report. Some pertinent information is repeated here for simplicity, to support additional information provided in this report. 5.2 Site Information 5.2.1 Grid and True North Grid North (for the British National Grid – BNG, based on the OSGB36 datum) at the Poulshot solar farm site is 0.04° west of True North, i.e., True bearings are 0.04° less than Grid bearings and the PV panels are aligned on a Grid bearing of 180.04°. Grid bearings are used throughout this report unless otherwise specified. Charts and other pictorial representations in this report are typically aligned with Grid North at the top for consistency with OS mapping used: associated graphical representations of the directions of solar reflections, etc., are adjusted accordingly. Due to the small difference between True North and Grid North at this location, it will normally be adequate for the reader to assume – for simplicity – that they are coincident for the purposes of this report. 5.2.2 Local Sun Data Sun position data used in this report was predicted to a precision of 0.1° in altitude and azimuth at 1 minute time intervals for all of 2014 – equivalent to a typical year in the operation of the solar farm – at WGS84 longitude: 002° 03’ West, latitude: 51° 21’ North (the nearest whole arc-minute of longitude/ latitude to the site). The sun is modelled as a point in the sky, but in reality has an angular diameter as viewed from the Earth of approximately 0.5° (i.e., 0.25° around the central point). Atmospheric effects – specifically refraction – typically cause the sun to appear approximately 0.5° higher in the sky than its true position: the predicted sun data accounts for this using ‘standard refraction’, although there may be small variations due to weather effects. The actual angular size of the sun and possible variations in its position due to weather effects are accommodated in this report. The path of the sun across the sky changes gradually from year to year. Significant changes over the life of the solar farm are corrected each leap year so the calculated sun data for 2014 is valid for the project’s whole period of operation. 5.3 Site Boundary The site boundary is marked on the site chart in Section 5.3.1 below. It was assumed that solar reflections can come from anywhere within the site boundary (although this is worse than the actual case: no panels will be sited right on the boundary). For detailed geometrical analysis, the azimuth and altitude angles between 20 significant points in the solar farm and each receptor point were considered. These significant points are described next. 5.3.1 Site Chart The site boundary is marked as the red line overlaid on a background 1:25,000-scale OS Explorer map in the following chart. Points at the 20 numbered dots were used for detailed geometrical analyses; analyses of those with yellow circles are shown as examples in this report (these are typical of the most significant points on the solar

Poulshot Lodge PV Array 27 RAF Keevil Glint & Glare Assessment

farm; coordinates for these 5 points are given in Section 5.3.2 below). Also shown in the chart are the ‘near-horizontal’ solar reflection limits (orange lines) and the limits where reflections may be directed at less than 45° above horizontal (brown lines) as calculated later.

Limits of reflections <45° above horizontal

Limits of near-horizontal reflection zone

Prominent points shown in detailed analyses

Poulshot Lodge PV Array site boundary

5.3.2 Poulshot Lodge Significant Points 20 significant points were chosen on and within the site boundary at Poulshot Lodge for detailed geometrical analyses. A number of points were used so that – combined with a number of closely spaced receptor points and appropriate buffers – there were no gaps when the geometries of the solar farm points and receptor points were considered. Of the 20 points, 5 were chosen as prominent points for which the geometrical analyses are shown in this report. Coordinates and panel elevations (amsl, equivalent to Above OS Datum – AOD – for the purposes of this report) are given for these in the following table (Point IDs are as in the chart above). Terrain elevations are from OS Digital Terrain Model (DTM) data, and solar panels were assumed to be 2.5m high.

Poulshot Lodge PV Array 28 RAF Keevil Glint & Glare Assessment

Terrain Modelled Panel Point Easting Northing elevation Panel elevation Notes ID /m (AOD) height /m /m (AOD) 1 396333 161423 57 59.5 Northernmost point 4 396808 161249 67 69.5 Northeastern point 11 396024 161038 54 2.5 56.5 Westernmost point 15 396825 160899 72 74.5 Easternmost point 20 396341 160558 67 69.5 Southernmost point Although only a selection of solar farm points are shown, all data were used in the analyses to generate the results given in this report. 5.4 Initial Analysis 5.4.1 Background To aid the selection of the most significant receptors, an initial general assessment of solar reflections was conducted in AARDVaRC’s earlier report. This identified that ‘near-horizontal’ or ‘near-ground level’ reflections (taken to be no higher than 10° above horizontal and no lower than 5° below horizontal occur in 2 zones: to the east and to the west. Reflections within the +10°/ -5° altitude band will never occur outside these azimuthal zones. Due to the proximity of RAF Keevil, further early analysis was conducted as detailed in this report to establish a zone within which solar reflections may occur from horizontal up to an altitude angle of 45° above horizontal, the ‘significant aerial reflections zone’. Generally solar reflections at steeper altitude angles than this will be insignificant to pilots (discussed later). Although this initial analysis is helpful, it does not eliminate the need to conduct detailed analyses on all receptor points considered using the entire data set for solar reflections, as described later (the detailed analyses consider all possible reflections, and were not limited to those that may occur within the reflection zones described here). The solar reflection zones in relation to all the aerial receptor points identified are shown in the charts at Section 6.9.4 below. 5.4.2 Solar Reflection Zones The ‘near-horizontal reflections zone’ is marked by orange lines on the chart at Appendix 1. The ‘significant aerial reflections zone’ (which bounds all reflections directed upwards at less than 45° above horizontal) is marked by brown lines on the chart at Appendix 1. It is assumed for simplicity that solar reflections may be directed more steeply upwards than 45° above horizontal in any azimuthal direction 5. However, since slopes of greater than 45° represent gradients of 100% (1:1) or more, such reflections could only be observed when aircraft are as close as (measured horizontally), or closer than, their height is above (measured vertically) a solar panel; i.e., an aircraft 1,000ft above the level of a solar panel must be within 1,000ft (305m) of that panel to have any chance of observing solar reflections directed so steeply upwards.

5 Although this assumption is made, it is not strictly true: reflections are confined in 2 arcs up to approximately 65° altitude angle above which reflections may occur in any azimuth at certain times; this may be seen from the chart in Section 7.3.2 below.

Poulshot Lodge PV Array 29 RAF Keevil Glint & Glare Assessment

6 Receptors at RAF Keevil Aerodrome 6.1 Background RAF Keevil aerodrome is 8km west-southwest of Devizes, Wiltshire. It’s military used is understood to be primarily by RAF tactical transport aircraft – i.e., C17, C130, and in the future A400M aircraft – to train pilots in techniques that may be used when operating from overseas airstrips that may be only semi-prepared or may not be totally secure from enemy action. There are proposals to construct a new, gravel runway to further facilitate this training. However, it is understood that MoD concerns relate to the airfield’s current use. In any case, the proposed new runway is further to the south of – and almost parallel to – the current runway, i.e., away from the zone of solar reflections and so the analysis of the current runway is a ‘worse case’ and will therefore be valid for future developments. Its runway, Rwy 24/06, is 1.8km long and aligned from approximately northeast to southwest. There are no known radio navigation aids at RAF Keevil, and so only visual (i.e., non-instrument) approaches are understood to be conducted there. There are no other runways understood to be in use by military aircraft. 6.2 Data Sources There is very little published data regarding the military use of RAF Keevil. The locations of the ends of the runway are extracted from OS map data (with elevation data from interpolated OS digital terrain model data), and from these, all further receptor points were calculated. 6.3 Runway Nomenclature Runways are referenced with numbers depending on the runway direction as follows.  The runway number is the approximate heading (rounded to the nearest 10°) divided by 10, e.g., Rwy 24 is aligned on a heading of 240° (strictly, a magnetic heading, but that is not relevant for the purposes of this report, other than to note that Grid or True headings given later are not the source data for runway identification), so the runway in the opposite direction (Rwy 24) is aligned on a heading of 060° (i.e., Rwy 06).  A physical runway surface is referred to by both runway directions, e.g., Rwy 24/06 refers to the strip that serves Rwy 24 and Rwy 06 6.4 Runway Data Two runway touchdown points were considered for each runway in this assessment: the runway thresholds, and inset touchdown points 1/3 of the runway length in from the thresholds (most aircraft would be expected to touchdown between these points). These locations are given in Section 6.9.2 below. Here, runway headings and lengths are summarised, and shown graphically in Section 6.6 below. The lengths of the runways and their headings (relative to True North) are shown in the following table as calculated from map study and GPS survey data. Runway Heading Runway Strip Runway Length Runway /° (True) Rwy 24 238° Rwy 24/06 1800m Rwy 06 058°

Poulshot Lodge PV Array 30 RAF Keevil Glint & Glare Assessment

6.5 Surface Receptors No surface receptors – e.g., a Visual Control Room (VCR, or ‘control tower’) if there is one – will be affected by solar reflections as the entire airfield is outside the zone for solar reflections as discussed in AARDVaRC’s previous report, in Section 6.5.2.5 Keevil Airfield. Therefore no further analysis or discussion of surface receptors is given in this report. 6.6 Airfield Chart The following chart shows the runway layout superimposed on an OS Explorer 1:25,000 scale background map. The runway thresholds and inset touchdown points are shown. Being surface receptors, the touchdown points have no significance other than as anchor points for calculation of aerial receptor point locations.

Rwy 06 inset touchdown point

Rwy 24 Rwy 06 Threshold Threshold Rwy 24 inset touchdown point

Poulshot Lodge PV Array 31 RAF Keevil Glint & Glare Assessment

6.7 Runway Approach Paths For this analysis, it was assumed that:  aircraft approaching to land may touchdown between the runway threshold and a third of the runway length beyond the it;  approaches may be made from up to 10° either side of the runway extended centreline, measured from the respective touchdown point.  aircraft may make approaches from 3° up to very steep 9° glideslopes (to accommodate tactical approaches used to minimise an aircraft’s exposure to small arms fire – it may be that aircraft would descend at even steeper angles than this before intercepting a less steep glideslope for ‘short finals’, the last stage of the approach to land, this possibility is not analysed but is discussed later). Therefore, multiple approach paths for each runway direction were modelled for geometric analysis in relation to the solar panels, as follows.  Touchdown points were taken as the runway threshold, and a point one third of the runway length beyond the landing runway threshold (i.e., 2 possible touchdown points on each runway).  Approach paths were taken to be along each runway’s extended centreline, and also 10° to each side of the extended centreline (i.e., 3 possible approach directions), measured from each possible touchdown point.  Glideslopes were assumed to be at 3°, 6° and 9° for each approach path to allow for possible variations. This gives a total of 18 approach variations modelled for each runway. 6.8 Other Phases of Flight Other phases of flight (i.e., take-off, the airfield circuit pattern, and transit) are less sensitive than approach and landing as aircraft are further from the ground and not descending towards it at low level, and they will not be flying towards any solar reflections for an extended period of time so that effects will be transient. Therefore only limited analysis was conducted for them which was sufficient, with the other analyses conducted, to support discussion of effects. In the absence of definite information, it seems likely that circuits to RAF Keevil are flown to the south of the runway, as there is above a wide area with few dwellings, whereas, the north side of the runway has many villages. However, since there is no information, it is assumed that both sides of runway may be used for the circuit. Discussion of solar reflection effects on other phases of flight is conducted later in Section 7.7.3 based on the results of the various analyses. 6.9 Identification of Solar Reflection Receptors 6.9.1 Outline For dim solar reflections from dark solar panels to be a significant nuisance, panels must be close to the direction of a viewer’s main interest for extended periods. There is no conceivable possibility of ‘flash blindness’ from brief exposure to solar reflections form solar panels given the high ambient light levels that must be present for them to occur (as discussed in Section 4.2.2 above) so brief exposure to reflected sunlight from solar panels may be less significant than prolonged exposure.

Poulshot Lodge PV Array 32 RAF Keevil Glint & Glare Assessment

In the case of a solar farm viewed from an aircraft, such prolonged exposure can normally only occur when a pilot is flying in a straight line in the general direction of the solar farm (i.e., descending). The only likely instance being where the solar farm is close to a runway and the aircraft is descending to land there. At this stage, it is assumed that aircraft approaching to land at RAF Keevil may meet these criteria with respect to the Poulshot Lodge solar array proposal: the possibility is discussed further later. Aircraft taking-off, manoeuvring in the vicinity of the airfield, or transiting through the local airspace do not meet all of these criteria. Therefore, of these, aircraft approaching to land at RAF Keevil are most sensitive and are considered in detailed analyses whereas other air traffic is not considered in such depth (although it is discussed later). Aircraft commencing a ‘straight-in’ approach to land at long distances from a runway at RAF Keevil may in theory be subjected to solar reflections for longer periods of time (due to the slow rate of change of bearing and vertical angle from the solar farm). However, at long distance from solar panels, reflections will be attenuated by scattering effects (e.g., diffraction – the bending of light passing through an aperture – which even applies to specular reflections as the reflecting surface effectively forms an aperture) and atmospheric obscuration (this is apparent in commonplace experience, e.g., solar reflections from car windows and mirrors some distance ahead on a motorway). Also, at such distances on the approach, aircraft will be high so there will be no concern regarding their proximity to terrain. Furthermore, since RAF Keevil is a minor aerodrome, it is unlikely that long, straight-in approaches would be used regularly. Dim (due to atmospheric attenuation with distance – discussed later) reflections may be seen to persist from aircraft approaching to land at shallow approach angles to Rwy 06 at long distances from touchdown. This will not occur on approaches to Rwy 24 due to the distance of the solar farm from the runway: aircraft would fly through any solar reflections in a short time (discussed in Section 7.7.3.4 below) rather than down a beam of solar reflections. 6.9.2 Runway Touchdown Points and Approach Heading Data Runway touchdown points (in each case, the runway threshold and an inset touchdown point, 1/3 of the runway length in from the threshold) assumed for this assessment are given in the table below. Runway threshold data is taken from map inspection as British National Grid (BNG) eastings and northings in the OSGB36 coordinate system. Inset landing point locations and approach headings were calculated from threshold location data. Terrain elevations were taken from interpolated OS DTM elevation data.

Runway Elevation Approach (Grid) Heading Runway Touchdown point Easting Northing strip /m AOD 10° right of On 10° left of centreline centreline centreline Threshold 391548 156825 64 Rwy 24 228° 238° 248° New Rwy Inset (1/3 rwy length) 392067 157144 57 24/06 Inset (1/3 rwy length) 392585 157463 50 Rwy 06 048° 058° 068° Threshold 393104 157782 48 Due to the location of solar panels near to some Rwy 24 approach paths, geometrical analysis between the solar panels and approach paths was not conducted in close proximity to the solar farm (within approximately 1km of its boundary), nor to the north and east of the solar farm at such distances from the airfield, any effects from solar reflections will be similar to those for en-route (transiting) traffic (discussed separately later). Instead, for simplicity, the worst case situation that reflections may reach such

Poulshot Lodge PV Array 33 RAF Keevil Glint & Glare Assessment

approaching aircraft from any possible solar reflection geometry was assumed, with reflections from the most significant directions specifically considered. Pilot eye height was taken to be 2m above ground level at the touchdown points and at all approach points for continuity. 6.9.3 Flight Profiles other than Approaching to Land In general, aircraft on other flight profiles could only ever be subject to transient solar reflections. Coming from the ground, any reflections will be below the aircraft and hence often blocked by the aircraft’s structure; furthermore, they will not be in a direction of particular interest to a pilot for extended periods (unless, for example, the pilot is investigating solar reflections). Therefore other flight profiles are not analysed in such detail in this report. However, to support discussion of other flight profiles, particularly circuit flying, points on either side of the runway threshold at circuit height (1,000ft above mean threshold elevation), 0.5nm (926m) and 1nm (1852m) from the runway extended centreline, were assessed for solar reflections. Other flight profiles, including circuit flying are discussed – based on the findings of all the analyses conducted – in Section 7.7.3 below. 6.9.4 Airborne Receptor Point Locations 6.9.4.1 Runway Approach Paths The runway approach path receptor points used in detailed analyses were taken for approaches to each touchdown point specified in Section 6.9.2 above. Approaches considered were: along each runway’s extended centreline, and displaced 10° to the left and right of centreline (measured from the respective touchdown point). Three glideslopes were considered for each approach path: 3°, 6° and 9°. Receptor points for detailed analysis were placed at 400m intervals to a distance of 6km from each touchdown point and for each approach variation. Beyond approximately 4km from the runway on approach to Rwy 24, an aircraft may come close to the solar farm and may be subject (at the respective times) to solar reflections from many directions (but from only one direction at any instant). Here, geometric analysis is of limited use, and other analytical methods were used (as described later) 6.9.4.2 Circuit Receptor Points To provide additional data to support the discussion of effects of solar reflections on circuit traffic in particular, 5 points along the downwind leg of each of 4 possible circuit patterns were analysed (20 points in total). These were located ½nm (926m) and 1nm (1,852m) from the extended centreline on either side of the runway, abeam each threshold, the runway midpoint, and at an angle of 45° beyond each threshold relative to the centreline. It was assumed that circuits are flown at 1,000ft above the mean threshold elevation. Although circuit traffic may or may not pass through these precise points, they provide additional information to support the discussion of effects of solar reflections on circuit traffic, whether fixed- or rotary-winged aircraft.

Poulshot Lodge PV Array 34 RAF Keevil Glint & Glare Assessment

6.9.4.3 Chart of RAF Keevil Receptor Points The chart below – shown with OS British National Grid lines at 1km intervals but with no background map for clarity – marks all receptor points considered for RAF Keevil. The blue dots mark the approach path receptor points with circuit receptor points (green diamonds and mauve squares) runway outlines, two significant reflection zones (orange and brown lines), and the solar farm boundary also shown.

RAF Keevil: Aerial Receptor Points (with runway, solar farm and solar reflection zones marked) N

Limits of zone of Limits of near- solar reflections horizontal solar below 45° (P5) reflections zone (P2)

(P4) Poulshot Lodge solar farm

(P3)

Rwy 24/06 circuit receptor points (green diamonds, (P1) purple squares)

Approach path receptor points (blue dots – 2 dots outlined in red are used as examples)

Poulshot Lodge PV Array 35 RAF Keevil Glint & Glare Assessment

The chart below is a repeat of the above chart overlaid on an OS 1:250,000 scale roadmap without the labels for geographic reference purposes. N RAF Keevil Receptor Points (with runway, solar farm and reflection zones marked)

(P5) (P2)

(P4)

(P3)

(P1)

Contains Ordnance Survey data © Crown copyright and database right 2013

Poulshot Lodge PV Array 36 RAF Keevil Glint & Glare Assessment

6.9.5 Selected Receptor Point Locations BNG coordinates and descriptions of 5 aerial receptor points are given in the following table. These are used later as examples in this report to illustrate the analyses of all receptor points. It is impractical to show coordinates and analyses for each of the many points considered in this report. Those listed are shown later with solar panel/ receptor point geometry to support the detailed analyses. It is emphasised that the detailed receptor-specific analyses were not limited to these points, nor to solar reflections that occur only in the zones described earlier: the complete data sets were used throughout. It is only due to the practicalities of space that a selection of data is presented here. Point ID Runway/ Circuit Receptor Description Easting Northing Altitude 10° left of centreline, 9° glideslope, 6km from touchdown at 1000m / P1 Rwy 06 approach point 385969 154617 the inset touchdown point 3282ft 10° right of centreline, 9° glideslope, 3.6km from touchdown at 636m / P2 Rwy 24 approach point 395796 160172 the runway threshold 2087ft South Circuit Point 5 P3 Each point is 45° off the runway’s extended centreline 394378 157478 (0.5nm from centreline) measured from Rwy 24’s threshold (to the east or north of the North Circuit Point 5 threshold, respectively, i.e., on the final approach side, not the 361m / P4 393408 159056 (0.5nm from centreline) runway side of the threshold), and either ½ or 1nm (measured 1184ft North Circuit Point 10 perpendicularly) from the extended centreline. Circuit height is P5 393711 160330 (1nm from centreline) taken to be 1000ft above the mean threshold elevation.

Poulshot Lodge PV Array 37 RAF Keevil Glint & Glare Assessment

7 Glint and Glare Assessment 7.1 Outline The terms ‘glint’ and ‘glare’ are commonly used together, and different definitions are used by various sources referring either to the specular (direct) and non-specular (diffuse) reflections of sunlight, or to brief and prolonged bright reflections of light, respectively. To avoid confusion, the terms ‘glint’ and ‘glare’ will be avoided where possible in this section. This section considers specular reflections and the duration of those reflections as they affect receptors will be considered where necessary. Specular reflections are much brighter – hence more significant – than non-specular ones: this report focusses on the specular variety. The difference in significance between bright flashes and longer periods of brightness is less clear: where an impact of brightness may be significant, it may be exacerbated by sudden and brief exposure to it with no possibility of becoming accustomed to it (e.g., while driving at night on a dark country lane and suddenly seeing the bright headlights of an oncoming vehicle directly ahead). However, continuous flying or driving directly towards the low sun may be considered a greater nuisance than a fleeting glimpse of the sun as it passes in front, say, of a manoeuvring aircraft or car. These aspects are considered later where appropriate. PV cells are designed to absorb light and are therefore dark in colour and not very reflective – much less reflective than, say, a body of water or standard glass (e.g., windows and car windscreens). To further minimise nuisance from reflections, additional treatment is commonly added to the surface of PV cells to scatter reflected light in a non-specular manner (i.e., in many directions hence attenuating rapidly). Effects of specular solar reflections from the PV panels are considered here. These effects diminish with distance from the solar farm due to scattering effects that are even associated with specular reflections (e.g., diffraction effects) and atmospheric attenuation. This is a commonplace phenomenon: solar reflection off a car windscreen (which is more reflective than solar panels) a few metres away is much more of a nuisance than similar reflections tens or hundreds of metres away. Where appropriate, another everyday) example (to pilots and non-pilots alike is used to gauge likely effects: the nuisance of looking at or very near to the sun. The sun is many, many times brighter than reflections from dark solar panels and there is no attenuating effect with distance similar to that for reflections from solar panels. 7.2 Sun Data 7.2.1 Overview The sun’s position was plotted on a minute-by-minute basis for a whole year. In this section, it’s position in the sky and the times when it is above the horizon through the year are considered. From this data, the directions of solar reflections were calculated, with dates and times of those reflections. 7.2.2 Positions of the Sun The position of the sun in the sky varies with the time of day and seasonally. The chart below shows the various positions the sun occupies in the sky at various times (although date and time information is not depicted in the axes of this chart, 6 date/ time data points are specified on the chart – those at the equinoxes occurring twice) at Poulshot Lodge.

Poulshot Lodge PV Array 38 RAF Keevil Glint & Glare Assessment

The daily movement is westwards from east, starting at 0° altitude (sunrise) in an arc with increasing altitude reaching its highest point about midday, and then continuing the westward arc with reducing altitude until sunset in the west at 0° altitude again; azimuths are given as True bearings. Apart from the innermost and outermost arcs, the sun passes through similar points in the sky twice each year. The inside boundary of the arc is the sun’s path at around the Winter Solstice (about 21 December), the outer boundary is the sun’s path at around the Summer Solstice (about 21 June). At the equinoxes, the sun rises and sets at almost due east and west respectively, i.e., its path is then roughly in the middle of the arc shown.

Poulshot Lodge PV Array 39 RAF Keevil Glint & Glare Assessment

7.2.3 Dates and Times when the Sun is Above the Horizon All dates and times that the sun is above the horizon are shown graphically in the following chart. The chart also shows the altitude bands of reflected sunlight from the proposed solar farm at these times (also where no reflections are possible when the sun is shining on the undersides of the solar panels), which are referred to later in this assessment.

7.3 Solar Reflections 7.3.1 Overview At any instant, solar reflections from all of the panels in the solar farm will be parallel to each other, the direction determined by the position of the sun and the orientation of the panels. In this section, the directions of all reflections, and the dates and times when they occur, are considered. 7.3.2 Directions of Solar Reflections As well as the time of day and year, the direction of solar reflections (in altitude and azimuth) depends on the latitude (from which the position of the sun at any instant can be calculated) and the orientation of the panels. The chart below shows the directions of all solar reflections calculated at 1 minute intervals through the year; azimuths are given as True bearings. Apart from the points on the curved arcs on the outside of the shaded area, reflections are in similar directions twice each year. The gaps at the top of the chart are due to the stretching of the zenith from a single point to the length of the axis and the 60 second intervals between data points.

Poulshot Lodge PV Array 40 RAF Keevil Glint & Glare Assessment

For the purposes of this assessment, solar reflections directed at more than 45° above horizontal are generally considered insignificant for pilots. Earlier (in Section 5.4.2) it was assumed that solar reflections directed upwards at more than 45° above horizontal may occur in any azimuth: the above chart shows that it is possible to specify this even more precisely: only reflections above approximately 70° (the higher of the altitude angles of solar reflections at the summer and winter solstices – see Appendix 2) occurs in all azimuths. Below approximately 60° altitude angle (the lower of the altitude angles of reflections at the summer and winter solstices – again, see Appendix 2), solar reflections are confined to 2 azimuthal arcs. The remaining solar reflections (below the 45° altitude line) are clearly confined in 2 arcs: one approximately westerly (occurring in the morning), the other approximately easterly (occurring in the evening). There is no indication of dates and times of solar reflections in the axes of the chart above, so 10 date/ time data points are shown (8 points are marked but the equinox points occur twice in a year). 7.3.3 Dates and Times of Solar Reflections The dates and times of reflections at various altitude angles are shown in the chart at Section 7.2.3 above. 7.4 General Effects of Solar Reflections These are described in AARDVaRC’s earlier report in Section 6.2. 7.5 Outline Assessment Method 7.5.1 General This was described in AARDVaRC’s earlier report in Section 6.3.

Poulshot Lodge PV Array 41 RAF Keevil Glint & Glare Assessment

7.5.2 Tasks In addition to the tasks described in AARDVaRC’s earlier in Section 6.3, specific receptor points were identified for detailed geometrical analysis with the solar farm using results from the solar reflection calculations and identify dates and times when reflections match the specific geometries identified (with appropriate buffers). 7.5.3 Assumptions Due to the volume of data generated (e.g., the sun is normally above the horizon for more than 250,000 minutes in a year, and solar reflection data was calculated at 1 minute intervals with more than 1,000 receptor points specifically analysed), the following general assumptions were made to reduce the data to a manageable quantity for both analysis and presentation in this report (these allow the data to be broken down into various levels of significance, they were not used to discard results), based on the author’s experience as a professional pilot, common experience, and other factors.  Solar reflections reaching an aircraft are not generally significant if they come from an angle of greater than 45° below horizontal as it is a direction of limited interest to a pilot and is likely to be blocked by the aircraft structure. This is valid for aircraft approaching to land, and also turning (banking in a turn may expose more of the ground to view, but directions steeply below a pilot remains of little concern to the pilot during the turn).  Approach paths are assumed to be straight lines to the point of touchdown to simplify geometrical analysis.  No geometrical analysis was conducted where an aircraft may fly directly over or close to the solar farm on approach to Rwy 27: instead it was assumed that solar reflections may reach the aircraft from any of the possible directions shown in the chart in Section 7.3.2 above; i.e., the solar farm is effectively assumed to have no boundaries (which is obviously worse than the reality).  Solar reflections reaching an aircraft approaching to land on any runway is generally of low significance unless it is within 45° (vertically, horizontally, or obliquely) of the touchdown point, which is taken to be where the pilot’s gaze is focussed. 7.6 General Analysis The general analysis was conducted in the original AARDVaRC report and may be found at Section 6.4 of that report. In addition, significant aerial reflections (considered to be those directed upwards at less than 45° above horizontal) can occur at the times and dates given in the results of the analysis for Rwy 24 in Section 7.7.2.2 below. 7.7 Specific Receptor Analysis The geometries (altitude and azimuth angles, and distances) between each of the RAF Keevil receptor points (specified in Section 6.9 above) and each of the significant solar farm points (specified in Section 5.3 above) were calculated and are tabulated here. The maximum and minimum altitude and azimuth angles for each receptor point was taken with an appropriate buffer of ±1° as bounding values for comparison with the solar reflections data to establish dates and times when reflections are within those limits. This allows for factors such as the angular size of the sun, variations in the

Poulshot Lodge PV Array 42 RAF Keevil Glint & Glare Assessment

sun’s observed position due to weather (hence atmospheric refraction effects) or in the alignment of the panels, and fills any gaps between receptor points (the use of multiple analysis points across the solar farm also serves this function). 7.7.1 Geometries between Solar Farm and Receptor Points 7.7.1.1 Overview The geometries – i.e., azimuths, altitude angles, and horizontal distances – between selected points on the solar farm and receptor points are shown below. Although only selected examples are shown here for brevity, all geometries between all solar farm points and all approach path points described earlier were considered in the analysis. Where all possible geometries were assumed to occur, (i.e., due to overflight of solar panels), geometrical data was not used to ascertain dates and times of reflections – discussed below where appropriate. In the tables, azimuths are given first as grid bearings with true bearings in parentheses. For the geometrical calculations, pilot’s eye height was taken as 2m above an aircraft’s assumed position. 7.7.1.2 Geometries between Selected Solar Farm and Receptor Points Geometries between Points 1, 5, 11, 15 and 20 on the solar farm and the various aerial receptor points are tabulated below.

Receptor Point P1 Receptor Point P2 Receptor Point P3 Receptor Point P4 Receptor Point P5 Point on PV Farm Alt Az Dist Alt Az Dist Alt Az Dist Alt Az Dist Alt Az Dist /° /° /m /° /° /m /° /° /m /° /° /m /° /° /m 236.7 203.2 206.4 231.0 247.4 4.34 12399 22.96 1361 3.94 4403 4.61 3763 6.09 2841 1 (236.7) (203.2) (206.3) (231.0) (247.3) 238.5 213.7 209.5 235.0 252.3 4.19 12707 24.44 1258 3.97 4297 4.53 3769 5.83 2922 5 (238.5) (213.7) (209.5) (235.0) (252.3) 237.4 194.7 204.8 232.9 253.0 4.52 11930 32.91 896 4.47 3922 5.33 3282 7.22 2419 11 (237.4) (194.7) (204.8) (232.8) (252.9) 239.9 234.7 215.6 241.7 259.6 4.22 12543 24.03 1260 3.92 4206 4.25 3883 5.20 3165 15 (239.9) (234.7) (215.5) (241.6) (259.6) 240.2 234.7 212.5 242.9 265.0 4.45 11953 40.32 668 4.59 3652 5.09 3296 6.34 2640 20 (240.2) (234.6) (212.5) (242.8) (265.0) 7.7.2 Analysis of Solar Reflections on RAF Keevil Approach Paths 7.7.2.1 Background The detailed analyses focus on the approach paths to all the runways at RAF Keevil. Where an approach involves flight directly over (or very close to) the solar farm with solar panels all around, an aircraft passes through very many possible solar reflection geometries each of which may affect the aircraft at certain dates and times. This applies to some approaches to Rwy 24. Therefore, where receptor points are above or very close to the solar farm, it was assumed that solar reflections may affect such aircraft from any possible reflection direction (but only from a single, discrete, direction at any time). This was applied only at distances of greater than 3.6km from the Rwy 24 threshold (including to the north and east of the solar farm, as at such distances from the airfield any effects from solar reflections will be similar to those for en-route (transiting) traffic) – the standard geometrical analysis was applied to all other receptor points of Rwy 24 approaches. To focus on the most significant data, it was assumed – based on the author’s own professional piloting experience – that the area directly below an aircraft is of little consequence to a pilot, so that solar reflections from steeply below an aircraft (more

Poulshot Lodge PV Array 43 RAF Keevil Glint & Glare Assessment

than 45° below the horizontal, i.e., reflections directed upwards at more than 45° above horizontal) were considered insignificant. Reflections from well to the side of the pilot’s focus of vision – taken to be the runway aiming point ahead – taken as more than 45° away (in any direction) from the approach vector (that is, the direction – in 3 dimensions – of the approach path, accounting for off-centreline approaches and atypical glideslopes) were also considered insignificant. This technique is valid for other receptor points but was not used to limit the data presented at other points, although it is discussed where appropriate. Where charts are shown to mark dates and times of reflections, they have 2 vertical orange stripes. These mark the dates within which local time changes from GMT to BST (the last Sunday in March, i.e., varying between 25 and 31 March from year to year) and from BST to GMT (the last Sunday in October, i.e., varying between 25 and 31 October). It may be helpful to refer to Appendix 2 of AARDVaRC’s earlier report when considering the direction of the sun and solar reflections at a particular date and time. Although only limited data is shown in that appendix, it should assist in understanding the nature of the phenomenon at any instant. 7.7.2.2 Rwy 24 Approach At or within 3.6km of the Rwy 24 Threshold. Aircraft approaching to land on Rwy 24 will have passed the solar farm when more than 4km from the runway threshold. There is no possibility of significant effects beyond this point, even if solar reflections could reach the aircraft (any reflections would then be from behind the aircraft). However, there is no possibility of solar reflections reaching an aircraft at 3.6km or less from the Rwy 24 threshold – even from behind – at a vertical angle of 9° or less above horizontal measured from any point between that threshold and a third of the runway length beyond it, within 10° of the extended centreline measured from the same point of the runway. Although this is irrelevant to aircraft approaching to land as the reflections would come from behind, it is pertinent – as discussed later – for aircraft departing on Rwy 06 (i.e., in the opposite direction). Beyond 3.6km of the Rwy 24 Threshold. Beyond 3.6km from the runway threshold, any effects from solar reflections will be similar to those for en-route (transiting) traffic (discussed later). However, close to the solar farm, reflections will be most significant when directed upwards at less than 45° above horizontal. Further to the east and north of the solar farm, reflections might only ever be visible when the aircraft is much less than 45° above horizontal (relative to the solar farm), but the results will be contained within the analysis of reflections at up to 45°. Effects in such cases will be similar to those on en-route (transiting) traffic. The chart below shows the dates and times of all reflections directed upwards at less than 45° above horizontal. These may be described as significant aerial reflections.

Poulshot Lodge PV Array 44 RAF Keevil Glint & Glare Assessment

Reflections to the east (from in front of an aircraft approaching Rwy 24)

Reflections to the west (from behind an aircraft approaching Rwy 24)

These dates and times are summarised in the table below.

Solar Reflection Occurrences Morning (to the west) Afternoon (to the east)

Dates of reflections All year

05:59 GMT/ 06:59 BST 14:36 GMT/ 15:36 BST Times of reflections to to 09:38 GMT/ 10:38 BST 18:17 GMT/ 19:17 BST Other Factors. During the summer months, solar reflections at less than 45° above horizontal only occur more than 45° to either side of the approach vector to Rwy 24. During approaches to Rwy 24, when the reflections are at less than 45° above the horizontal, the sun is always closer to the approach vector, i.e., the direction of the runway aiming point ahead (as viewed by a pilot) than the solar farm and any reflections from it. Such reflections will necessarily be insignificant due to the proximity of the much brighter sun to the runway ahead on which the pilot’s gaze will be focussed. Effects will therefore be of low significance.

Poulshot Lodge PV Array 45 RAF Keevil Glint & Glare Assessment

7.7.2.3 Rwy 06 Approach Considering the variations in possible approach path discussed at Section 6.9.4 above, aircraft approaching to land on Rwy 06 at glideslopes of up to 9°, approaching from within 10° either side of the runway extended centreline as measured at touchdown points from the runway threshold to a third of the runway length in from the threshold, will never receive solar reflections from the Poulshot solar farm. Possible more extreme approach paths are discussed later. Effects will therefore be of low significance. 7.7.3 Discussion of Solar Reflections Effect on Phases of Flight Other Than Approach and Landing 7.7.3.1 General Other phases of flight are not as sensitive to solar reflections as landing for the following reasons.  An aircraft will not be flying directly towards the solar panels (as it may do when approaching to land and panels are close to the runway), so the aircraft will quickly pass through any reflections and effects will be transient.  The pilot’s focus of gaze will not be towards, or close to, the solar farm as it may be when approaching to land.  An aircraft will either be on the ground (during the take-off roll – although there is unlikely to be a view of the solar farm from aircraft on the ground); climbing rapidly away from the ground with the pilot’s focus of vision being the sky into which he/ she is climbing (just after take-off), i.e., away from any reflecting panels; or well away from the ground (all other phases of flight). It will not be flying towards the ground at low level. Therefore, limited detailed analysis has been conducted on these phases of flight and it would not be possible to analyse every possibility. However, points 0.5nm (926m) either side of each runway threshold they assessed for solar reflections to support the discussion of effects within the circuit. Other effects are discussed here based on the findings from these and the other analysis already described. 7.7.3.2 Aircraft Taking-Off General. An aircraft’s take-off profile is loosely similar to that for an aircraft approaching to land in the opposite direction (although the initial climb after take-off is normally steeper than typical glideslopes). This fact allows approach path solar reflections analysis to be considered in this discussion. There are 2 phases to take-off considered: the ground roll, followed by the initial climb. If reflections are observable during the ground roll, they cannot affect the initial climb and vice-versa (as the solar panel-aircraft geometries will change, but the reflection geometry won’t change noticeably in such a short timeframe). Occasionally, aircraft are required to take-off directly into the sun. This is accepted and there is no safety impact. Taking off into the sun will always be much more severe than effects of solar reflections from dark panels.

Poulshot Lodge PV Array 46 RAF Keevil Glint & Glare Assessment

Rwy 24. Aircraft taking-off on Rwy 24 will be departing to the southwest, i.e., away from the solar panels. Any reflections will be in the early morning from well behind the aircraft and will not be significant. Rwy 06. The analysis for the approach paths to Rwy 24 is pertinent to departures from Rwy 06 as there are similarities in the flight path (but in the opposite direction). Repeating the analysis for the Rwy 24 approach path receptor points, but with points at 250m intervals (instead of 400m as described previously) to a distance of 3.75km from the ‘touchdown point’ on Rwy 24 along the approach paths, it was found that only the point at 3.75km from the Rwy 24 threshold, 10° north of the extended centreline and 9° above horizontal (both measured from the runway threshold), could ever receive solar reflections, and these are at 45.0° or more above horizontal. The dates and times of these reflections are shown in the following chart.

These dates and times are summarised in the table below.

Solar Reflection Occurrences

Dates of reflections May 29th to July 14th (local time is always BST between these dates)

10:28 BST Times of reflections to 10:52 BST

Poulshot Lodge PV Array 47 RAF Keevil Glint & Glare Assessment

Beyond this point from departure, reflections would be at more than 45° above horizontal Although this doesn’t cover every possible departure profile, there is no possibility of solar reflections affecting the early part of the departure from this runway. A departing aircraft would climb quickly through any solar reflections, and the aircraft nose attitude would be high with the pilot’s view focussed on the sky ahead and above which the aircraft is climbing towards. If solar reflections were ever observed by aircraft departing RAF Keevil from Rwy 06, effects would be similar to those for en-route (transiting) air traffic. Effects will therefore be of low significance. 7.7.3.3 Circuit Traffic A circuit pattern for a particular runway typically consists of an aircraft taking-off, making one (180°) or two (90°) ‘upwind’ turns to fly a ‘downwind leg’ in the opposite direction to take-off and landing (so called because aircraft normally take-off and land into wind) until some distance beyond the touchdown point then making 1 or 2 turns to line up on the runway’s extended centreline for final approach to land. When departing from or arriving at an airfield, aircraft may leave or join the circuit at any point. It is not known whether circuits at RAF Keevil are flown to the north or south of the runways, although due to the more sparsely populated area to the south, this direction would seem probable. Circuits Analysis Solar reflections at 20 points in total, 5 points on lines 0.5nm (926m) and 1nm (1,852m) either side of the runway were analysed. The points are referred to as being on the north side or south side of the runway with the points numbered form 1 to 10 on each side of the runway (see the chart is Section 6.9.4.3 above for the point locations and numbers). Circuits were assumed to be flown at 1,000ft above the mean threshold elevation. All 20 points are at vertical angles of less than 10° above horizontal from all points in the solar farm and only those 1nm north of the runway’s extended centreline are within the zone of ‘near-horizontal’ solar reflections, implying that only those points 1nm north of the runway could ever be subject to solar reflections. This was confirmed in the detailed analysis. Therefore, the circuit to the south is not discussed further: it will not be affected by solar reflections, and the following discussion of effects of solar reflections on the northern circuit only considers ‘wide’ circuits, with the downwind leg flown at approximately 1nm from the runway. Rwy 24 Circuit. As an aircraft executes the upwind turn (the right, i.e., north) after take-off, the solar farm will gradually come into view in front of the aircraft, and will be close to directly ahead at the start of the downwind leg. In the very early morning, there is a slim chance that the pilot may see solar reflections from the panels (in which case, reflections would not be visible by the end of the downwind leg). As the aircraft gets closer to the solar farm and the vertical angle between the panels and the aircraft increases, solar reflections may occur progressively later in the morning, until the finals turn is started, after which the aircraft will be turning away from the solar farm (and any reflections).

Poulshot Lodge PV Array 48 RAF Keevil Glint & Glare Assessment

Dates and times that solar reflections may reach any of the 5 receptor points assessed are shown in the following chart.

These dates and times are summarised in the table below.

Solar Reflection Occurrences

Dates of reflections March 31st to September 12th (local time is always BST between these dates)

07:08 BST Times of reflections to 07:47 BST It is likely that any training at RAF Keevil at such times would take place only exceptionally, if ever. Effects will therefore be of low significance. Rwy 06 Circuit. It is virtually inconceivable that solar reflections could reach an aircraft until after it has initiated the upwind turn and is flying away – or at least, turning away – from the solar farm and ay reflections. Effects will therefore be of low significance.

Poulshot Lodge PV Array 49 RAF Keevil Glint & Glare Assessment

7.7.3.4 En-Route (Transiting) Traffic The following points are pertinent to aircraft in transit.  Since reflections from all panels are parallel to each other, an aircraft flying straight and level could be subjected to solar reflections only for the time it takes to fly across the length of the solar farm. As the solar farm is approximately 1km across in any direction, it would take an aircraft flying at approximately 120kts (62m per second) less than 20 seconds to cross through any reflections. Aircraft flying faster, or climbing or descending will generally take less time to pass through any reflections.  As the geometry between the solar farm and the flight path of a transiting aircraft is generally quite different from those on the approach, solar reflections may be experienced by such traffic outside the reflections zones shown in Appendix 1 at various times of day through the year.  Aircraft will normally be relatively high so reflections will often be blocked by the aircraft structure (or come from behind the pilot).  For aircraft transiting at 3,000ft above the solar farm (or less), solar reflections directed upwards at more than 45° can only affect an aircraft within 1km of the solar farm boundary, perhaps in any direction. Otherwise the zones for solar reflections shown in Appendix 1 apply, i.e.: less than 10° (‘near horizontal zone’) beyond 5.7km – about 3nm – from the solar farm (for aircraft at 3,000ft); or the ‘less than 45°’ zone within this distance (for aircraft at 3,000ft).  Solar reflections will not come from a direction of specific and immediate interest to the pilot (there is no reason why a pilot would normally need to stare in the direction of the solar farm whilst in in transit, reducing any nuisance from solar reflections).  Such aircraft will not be descending towards the solar farm so effects of solar reflections will be transient. This author has never known the MoD to raise concerns regarding solar farms due to effects on en-route or transiting traffic, so it is appears that the MoD accepts that effects of solar reflections form solar farms on such traffic are insignificant. Effects will therefore be of low significance. 7.7.3.5 Extreme Flight Profiles Tactical situations may necessitate ‘extreme’ flight profiles, such as very steep, rapid descents from high altitudes to intercept a more standard visual approach glideslope shortly before touchdown. Training in such techniques may take place at RAF Keevil Modern tactical transport aircraft are capable of very rapid descents, say from 20,000ft altitude at an angle of, perhaps, 45° below horizontal. To land safely following such a rapid descent, the aircraft’s descent must be reduced to a more normal rate for final approaches at some distance (perhaps a small distance) from touchdown. If such a descent was conducted for a ‘straight-in’ approach to Rwy 24, the solar farm would be very steeply below (or to the side) of the aircraft when it was at very high altitudes: the solar farm would almost certainly be out of sight from the aircraft, and if not in the initial stages, would move behind the aircraft very quickly. For such a descent towards Rwy 06, the aircraft is unlikely ever to be more than 45° above horizontal measured from the solar farm, and would therefore be outside the solar reflections zone throughout, unless displaced to the north to the runway extended centreline. Any reflections that may be visible would only be so early in the

Poulshot Lodge PV Array 50 RAF Keevil Glint & Glare Assessment

procedure (i.e., at very high altitudes, and/ or well displaced to the north of the extended centreline), and from the pilot’s perspective, would only be visible at long distances (i.e., they will be well attenuated and hence dim), and only well ‘above’ the pilot’s view of the runway ahead on a steep descent for perhaps as long as 5 seconds (the time it would take the aircraft to descend through them), so would probably not be noticed. Effects will therefore be of low significance. 7.7.4 Assessment of Solar Reflection Impacts 7.7.4.1 Background and Comparison to Commonplace Examples At times, pilots landing on various runways at RAF Keevil will have to land directly into the sun, so this fact is used as the basis for impact assessment. It is almost impossible to convey any meaningful quantification of intensity in a report such as this (even video representation would be very limited for this purpose); however, it is reasonable to state that however bright any solar reflections are, they will always be much less bright than the sun (typically about 2% of the sun’s brightness). There are a number of airfields where runways are near (or extend into) areas of water when solar reflections may occur suddenly and late in an approach; it may also occur from a wet tarmac runway just before touchdown. This author is not aware of concerns ever having been raised in such cases. The table below defines the impact levels used in this section: a 10-point scale (actually an 11-point scale in theory but not in practice, as is explained in the table) is considered necessary to accommodate the subtleties involved in assessing the impact of solar reflections, although the impact descriptions given are reduced to a 5-point scale. Only impacts of Level 5 (i.e., ‘small’) and higher are significant.

Level of Impact Definition Impact Description Absolutely zero impact. In practice, such a level of impact from any nature of development 0 is virtually impossible to justify, so this level is normally unused. Nil/ Effects not observable unless significant effort is taken to notice them, so there is practically 1 no impact no impact. Effects may be observable with little effort but with no practical impact on the task in hand 2 and will quickly be dismissed from an observer’s awareness. Observable effects that may persist in an observer’s consciousness but have no effect on 3 his/ her execution of the task in hand. Negligible Effects that are readily and continuously noticeable that do not affect the execution of the 4 task in hand. Effects that may cause an observer to take some almost unconscious action with no 5 noticeable effect on the task in hand. Small Effects that may cause an observer to take some conscious action that does not interfere 6 with the execution of the task in hand.

7 Effects that require a noticeable change in the execution of the task in hand to manage. Medium Effects that require a deliberate and noticeable change in the execution of the task in hand 8 to manage.

9 Effects that require a considerable effort to manage in the execution of the task in hand. Large 10 Effects that require an excessive effort to manage in the execution of the task in hand.

Poulshot Lodge PV Array 51 RAF Keevil Glint & Glare Assessment

Landing directly into the sun is deemed to have a Level 6 (small) impact: it is certainly noticeable, but a pilot just does it and copes by squinting, looking slightly sideways, or perhaps wearing sunglasses or using a visor: this takes little or no effort to manage. There is no concern regarding flight safety: this author knows of no airfields that would close a runway because the sun happens to be in that direction. Effects of reflections from dark solar panels must be less than this, even considering the ‘planned’ nature of approaches (or departures) into the sun. A commonplace non-aviation example of reaction to sudden bright light sources is how car-drivers’ react to the main beam headlights of oncoming traffic suddenly appearing from around a corner on a dark country road. Although some drivers might react with annoyance by flashing their own lights, if this option was not available drivers would squint and maybe try to look sideways to minimise the nuisance (as well as adjusting their driving to accommodate the change in the environmental conditions). Again, there is no question of safety being compromised and most commentators are likely to agree that reflections from dark solar panels on a bright day must be less severe than the sudden dazzling effect of looking into car headlights on a dark night. However, the latter example has some similarities to the possibility of solar reflections occurrences as it would be a sudden change in the environment prompting a reaction, but as it is less severe than flying (or driving) directly towards a low sun, it must only qualify for a Level 5 (small) impact. If the headlights are viewed from even slightly to the side rather than directly in front, it would be reasonable to reduce the impact to Level 4 (negligible). Where an impact is significant (i.e., level 5 and above) it may be reasonable to add 1 to the level of impact when there is a sudden change in the environment (e.g., headlights directly in front of a driver): in this instance, the immediate impact from the headlights suddenly appearing would be Level 6 (small), reducing to Level 5 (small) after the initial change. This wouldn’t normally apply to impacts that are negligible unless there is some other significant factor to make this appropriate. It is worth considering that when an observer’s eyes are attuned to ambient brightness (as will always be the case when solar reflections may be observed by a pilot approaching to land at RAF Keevil), the impact from the example of the car’s headlights would again be substantially reduced. The following standard assumption is used as a reasonable starting point for the assessment of solar reflection impact. In bright daylight where solar reflections come from a dark solar panel directly ahead of an observer with neither the observer nor the observer’s focus of view in deep shade (as will be the case when any reflections are observed by aircraft approaching to land at RAF Keevil – the aiming point on the runway will always be brightly illuminated by the sun): where the sun is not particularly close to the runway aiming point, a Level 4 (negligible) impact is attributed. Other factors should then be considered to modify this (upwards or downwards on the scale). Where such reflections are a significant angular distance away from directly ahead of an observer, or the sun is either very close to the reflecting object, or from more directly in front of the observer, any impact is be expected to be less than this. 7.7.4.2 Impact on Rwy 24 Approach At more than approximately 4km from touchdown, the sun will be close to any reflecting panels, or to the runway ahead (and invariably closer to the runway ahead than any significant solar reflections) so impact is assessed as Level 3 (negligible).

Poulshot Lodge PV Array 52 RAF Keevil Glint & Glare Assessment

Within 4km from touchdown, the solar farm (and any reflections) will be from below or behind the aircraft and impact is assessed as Level 1 (nil) applies for the remainder of the approach. There can be no question of ‘cumulative effects’ from these various stages of approach as solar reflections could only ever affect a small portion of any one approach (since the geometry of reflections doesn’t change appreciably during an approach to land), so the overall worst case impact of Level 3 (negligible) applies to approaches to Rwy 24. Due to the negligible impact and the proximity of the sun to the approach vector, no increment to the level of impact is considered necessary to accommodate any change to the environment as reflections may be encountered. 7.7.4.3 Impact on Rwy 06 Approach Solar reflections could only affect aircraft approaching this runway if they are excessively displaced to the north of the runway’s extended centreline. Such reflections (if they ever occur) would be in the far distance, hence well attenuated, so impact is assessed as Level 2 (nil). Due to the negligible impact and high ambient brightness when reflections may occur, no increment to the level of impact is considered necessary to accommodate any change to the environment as reflections may be encountered. 7.7.4.4 Impact on Phases of Flight Other Than Final Approach and Landing Impacts on other phases of flight will vary, but solar reflections will either be:  viewed from a long distance (hence attenuated);  the sun will be shining sufficiently close to reflecting panels or the main direction of a pilot’s view to minimise the significance of effects;  transient in nature; or,  from a direction of little interest to a pilot. In no instance will they be greater than Level 4 (negligible). The worst case impacts will be for transiting aircraft (i.e., Level 4, negligible) and/ or during the downwind leg of wide and high circuits to Rwy 24 on the north side of the airfield, and will be similar to what may occur to any transiting aircraft. Other impacts will be less than this. No increment is necessary to accommodate any change to the environment as reflections may be encountered in other phases of flight as impacts are only negligible. 7.7.4.5 Overall Impact on RAF Keevil Given the impact levels of the most significant impacts on aircraft (most notably on those approaching to land on some runways), the overall impact of solar reflections from the proposed solar farm on RAF Keevil is assessed to be no greater than Level 4 (negligible). This impact assessment ignores the limited time in which the solar reflections can occur, only considering the impact level when solar reflections occur. Extending this to consider the overall impact of solar reflections across airfield operations through an entire day (and year), the overall impact must be even less than this due to the limited periods of impact (and the requirement for the sun to be shining).

Poulshot Lodge PV Array 53 RAF Keevil Glint & Glare Assessment

7.7.5 Discussion of Solar Reflections with respect to the Air Navigation Order 7.7.5.1 Background In Section 4.2.3 above, the CAA’s concerns that solar farm development should be in accordance with the requirements of the ANO were specified, specifically relating to 3 articles within the ANO (Articles 137, 221 and 222). These are discussed more fully here. The CAA guidance also refers to that from the US Federal Aviation Administration (FAA), however the ‘current’ FAA guidance is under review and caveated such that the FAA has distanced itself from being held accountable for its use; it is discussed briefly in Section 4.2.2 above. It appears that the solar PV industry and operational experience of the benign nature of glint and glare from solar farms near airports has taken the lead in its assessment in the USA. 7.7.5.2 Article 137 – Endangering safety of an aircraft The commissioning of this assessment shows that the developer is taking reasonable measures to ensure that it is not recklessly or negligently acting in a manner likely to endanger an aircraft, or any person in an aircraft. Flight safety considerations are discussed next in relation to Articles 221 and 222. 7.7.5.3 Article 221 – Lights liable to endanger This article prohibits lights that are liable to endanger aircraft taking off from or landing at an aerodrome due to their glare, or that may endanger aircraft if liable to be mistaken for aeronautical ground lights. Glare Effects. The difference from the definitions of terms used in this report is noted: glare (as defined in Section 3 above) from the solar farm is unlikely to dazzle pilots; glint (as defined in Section 3 above) has been discussed in depth. As almost every pilot will have had to land or take off into the sun on occasion, the much dimmer solar reflections (whether in the form of brief flashes or more prolonged exposure) from a solar farm is unlikely to significantly affect a pilot other than perhaps as a nuisance. The specular reflections of sunlight from the proposed solar farm are not considered to be a significant nuisance for the reasons given in this assessment and there will be no impact on flight safety. Misidentification as Aeronautical Ground Lights. The solar panels do not emit light, and solar reflections are only possible during the day with the sun above the horizon. Therefore, solar reflections will not be a cause of misidentification of aeronautical ground lights. Cultural and other lighting may be very dimly reflected by solar panels at night. Such reflected lights are unlikely to be mistaken as the reflections will be so dim and if misidentification was likely, the source lights would not be permitted under this article. Furthermore, surface lights from near horizontal angles would generally be reflected by the inclined panels at high angles unless (almost close to due east or west of the panels) and would therefore only be noticeable to pilots for short periods as they fly through the reflection. Screening from hedges is likely to block either the source of such lights or near-horizontal reflections from the panels.

Poulshot Lodge PV Array 54 RAF Keevil Glint & Glare Assessment

If reflections at low angles were observed (to east or west), the light sources would also be visible to the west or east very close to the dim reflection and would be much more prominent. Therefore it is virtually inconceivable that reflections from the solar panels could be misidentified as aeronautical ground lights. 7.7.5.4 Article 222 – Lights which dazzle or distract This report has considered the effects of specular reflections of sunlight, which will be the most intense form of reflected light and found effects to be negligible. There is no question that effects of solar reflections during bright daylight with the sun near to the reflecting panels could approach the dazzling effect of, say, looking towards the sun, into a laser beam, or even of suddenly looking into the bright headlights of another car at night. Therefore the proposed solar farm cannot be construed as ‘lights which dazzle or distract’. 7.7.5.5 Summary The articles of the Air Navigation Order that CAA guidance directs solar farm developers to are all satisfied by the proposed solar farm. 7.7.6 Other Considerations 7.7.6.1 Proximity of the Sun For most aerial and surface receptors of solar reflections, when they occur, the sun will be shining relatively close to the reflecting object (as it appears to the observer). This reduces the impact of solar reflections as their relative intensity will be reduced by the ambient light level, an observer’s eyes will be attuned to brightness further reducing any nuisance from reflections, and the sun is likely to be the predominant nuisance to an observer. 7.7.6.2 Effects of Trees, Hedges and Fencing around the Solar Farm Existing and future treelines, hedgerows and fencing around the solar farm (particularly to the east and west) will reduce solar reflections in 2 ways as follows.  When between the solar farm and a receptor, they will generally block ‘near- horizontal’ reflections, especially from panels nearest to the hedge. As the height of the hedge increases, reflections from further inside the solar farm are progressively blocked.  Hedges on the far side of the solar farm will cast long shadows over the panels nearest to them when the sun is low in the sky (when ‘near-horizontal’ reflections are most likely to occur), so preventing them from reflecting and hence reducing overall reflection effects on a receptor. These long shadows will not significantly reduce the solar farm’s output as the sun will be low in the sky (hence there would be little electrical generation potential) and long shadows are only cast for short periods after sunrise or before sunset. This reasoning also applies to many aerial receptors, as most of these may be considered close to the horizontal plane through the solar farm. However, it is accepted that aircraft approaching to land at RAF Keevil are likely to have an unrestricted view of the airfield and large areas of the solar farm.

Poulshot Lodge PV Array 55 RAF Keevil Glint & Glare Assessment

7.7.6.3 Non-Reflective Panels Solar panels are very dark in colour – much darker than normal glass – as they are designed to absorb light to convert it to useful energy rather than reflect it (reflected light is wasted), and their surfaces may be further treated to scatter any reflected light rather than cause specular reflections. There are various sources of public domain information to support this. It is therefore considered that any solar reflections from the solar panels will be significantly dimmer than other common sources of such reflections and insignificant compared to the brightness of the sun (which will always be observed close to any solar reflections from the panels at Poulshot Lodge).

Poulshot Lodge PV Array 56 RAF Keevil Glint & Glare Assessment

8 Conclusions 8.1 General Effects of Solar Reflections 8.1.1 General Characteristics Solar reflections are commonplace occurrences for most people either from wet roads, expanses of water, or windows and mirrors of cars and buildings. Solar cells are designed to absorb light to generate electricity, not reflect it, and so are much less reflective than other sources of solar reflection. Solar reflections can only occur when the sun is shining. It has no significance when the sun appears very close to – that is, in almost the same direction as – the reflecting object as seen by an observer (i.e., the observed angle between the sun and its reflection is close to 0°) as the much brighter sun will completely mask any reflections and the observer’s eyes will be attuned to brightness when looking in that direction, thus reducing the observed intensity of any reflections. Conversely, solar reflections are at their worst when an observer is facing the reflecting object, is in shade from the bright sun so that his/ her eyes aren’t attuned to brightness, and the sun is behind the observer (i.e., the angle between observed reflections and the sun is close to 180°). Solar reflections from PV panels may typically have intensities of 20 W/m² (about 2,000 lux). Ambient light levels on a sunny day in shade, but illuminated by the entire blue sky, are typically 20,000 lux so the worst problem that may be caused by reflections from solar panels is a nuisance from looking at or near them. 8.1.2 Characteristics of Solar Reflections from Poulshot Lodge 8.1.2.1 Significant Aerial Reflections Solar reflections from the horizontal plane up to 45° above it may be significant for aircraft operating at RAF Keevil near the proposed solar farm. Reflections within this vertical band are described as ‘significant aerial reflections’. Such reflections can occur all year round:  in the mornings at receptors to the west of the solar farm from 05:59 to 09:38 Greenwich Mean Time (GMT), or 06:59 to 10:38 BST (after the last Sunday in March until the last Sunday in October);  in the evenings at receptors to the east of the solar farm from 14:36 to 18:17 GMT, or 15:36 to 19:17 BST (after the last Sunday in March until the last Sunday in October). Aircraft would need to be in extremely close proximity to the solar farm to receive reflections at steeper angles than 45°, although in some cases such reflections may occur in any azimuth at certain times. Aircraft 1000ft above the solar farm could only observe such reflections within 305m of its boundary; this increases to within 915m for aircraft 3,000ft above it. The first and last times of reflections on any day are later and earlier, respectively, than for ‘near-horizontal reflections because in this case, reflections below the horizontal plane are not considered. 8.2 Effects of Solar Reflections on RAF Keevil Solar reflections will never affect aircraft as they land or take-off at RAF Keevil. Aircraft operating in the wider vicinity, e.g., in a wide and high circuit to the north of the airfield

Poulshot Lodge PV Array 57 RAF Keevil Glint & Glare Assessment

or on approach at some distance (more than 4km) from touchdown at RAF Keevil. Such aircraft may receive solar reflections at various times. Overall solar reflection impacts for aircraft operating at RAF Keevil are assessed as negligible; no specific impact on a specific operation was assessed to be greater than negligible. This does not consider the very limited times when impacts are possible, and the requirement for the sun to be shining: true impacts on the operations at RAF Keevil will therefore be even less than those assessed here. All Civil Aviation Authority guidance on solar farms near aviation sites is met by the Poulshot Lodge proposal.

Poulshot Lodge PV Array 58 RAF Keevil Glint & Glare Assessment

Appendix 1 – Solar Farm, Reflection Zones and RAF Keevil Receptor Points Chart N

Limits of significant aerial reflections zone

Limits of near-horizontal reflections zone

Poulshot Lodge PV Array site boundary