In brief
A module throws sunlight back like a mirror casting a patch of light onto a wall. This chapter asks whether that patch of light falls on the window, i.e. on the observation point. Because modules are not perfectly smooth, the patch is not a sharp dot but a somewhat larger, soft-edged area.
A reflection reaches the observation point if the direction to a point on the PV area lies within the reflection lobe. The lobe surrounds the reflection direction, which follows from the sun position and the module orientation; its size depends on the module type and the diameter of the solar disc (0.53°).
For every time step, the solar position (Ch. V.1) and the surface normal (Ch. V.2) give the reflection direction, which is the centre of the reflection lobe. The lobe spans its angular range around this direction. A cell of the PV area counts as hit if the direction from the observation point to it lies within the lobe; in this direction the observer sees a reflection.
V.3.1 Reflection direction and lobe boundary
The sun must illuminate the front of the modules. The ideal specular direction follows from the law of reflection; the reflection offset δ rotates it within the plane of incidence towards flatter angles.
The deviation between the reflection direction and the viewing direction to the cell is split into two components: one in the plane of incidence (major axis) and one perpendicular to it (minor axis). A hit occurs if the point lies within the ellipse whose semi-axes are half the lobe width plus half the solar diameter. The solar diameter is taken as 0.53° because the sun is not a point source: every point on the solar disc produces its own lobe.
V.3.2 Multiple zones
For module types with zones, the ellipse is checked per zone, from the inside out. The innermost zone that still contains the point is decisive; its reflection profile determines the luminance. If the point lies outside all zones, there is no hit. All zones share the same centre and the same offset δ.
V.3.3 Perfect mirror
For the perfect mirror, the hit region is the solar disc (radius 0.265°) without offset; the check is carried out point by point (Ch. V.2, section V.2.3, deep dive "Accuracy of the hit check").
Deep dive: formulas
Notation: d direction of propagation of sunlight (from the sun to the area), n surface normal (unit vectors in the east, north, up system).
- Front side. A time step qualifies if d·n < 0. Angle of incidence θ = arccos(−d·n).
- Ideal specular direction. r = d − 2(d·n)n. The direction in which the observer must look to see the reflection is −r.
- Offset δ. Rotation of −r by δ(θ) about the axis n × d, i.e. within the plane of incidence. For δ = 0 the rotation is skipped; at normal incidence the plane of incidence is undefined and the rotation has no effect.
- Lobe axes. Minor axis eᵩ = normal of the plane of incidence (n × d, normalised); major axis eₗ perpendicular to it and to the reflection direction.
- Hit. With v = unit vector from the observation point to the cell centre (or calculation point), u = v·eₗ, w = v·eᵩ:
(u / a)² + (w / b)² ≤ 1
a = (major axis(θ) + 0.53°) / 2 + capture margin
b = (minor axis(θ) + 0.53°) / 2 + capture margin
Capture margin: half the cell diagonal (0.354°) on the cell grid, half the point-grid diagonal for the point-by-point check. For a circular lobe, a = b = (spread(θ) + 0.53°) / 2 + capture margin. a and b are used in radians; for small angles u and w are the angular deviations in radians. - Zones. The test is run per zone with that zone's a, b. - Hit: whether a cell is hit is decided by the cell centre with the capture margin; any zone suffices. This determines the glare minutes. - Zone of the cell: at zone boundaries, the zone that supplies the luminance is decided by the calculation points of the cell: the innermost zone containing one of them (with half the point-grid diagonal as margin), i.e. the part of the cell that lies on the area; never a zone further in than that of the cell centre with capture margin. - Only the capture margin hits: the outermost zone containing the cell with margin applies. - Solid angle: the zone of the cell gives its luminance. At zone boundaries, each calculation point contributes to the solid angle with the ratio of the luminance of its own zone to that value (innermost zone containing it with half the point-grid diagonal, never one further in than the cell's). The illuminance is thus the sum over the zones, and a point in a ring does not count with the luminance of the core. - Point-by-point check: the cell receives the luminance of the innermost zone containing one of its hit points; each point contributes to the solid angle with the ratio of its zone luminance to that value. The illuminance thus remains the sum over the zones. - Curves. Values between the 10° support points are interpolated linearly; this applies to the reflection profile, lobe axes, zones and offset.