For monitoring the photovoltaic performance of the system, a separate prototype with a hybrid absorber with polycrystalline silicon cells and a reflector was constructed.

The optical efficiency g(a) is defined as the ratio between the performance of the concentrating module and a vertical module of the same area as the concentrating aperture. It was determined through outdoor measurements. The short circuit current Isc of the concentrator module was monitored as a function of the angle of incidence p in the meridian plane. The optical efficiency (figure 6), was then derived according to

Isc -1000

I1000 • • G ■ cos(fl)

where 11000 is the short circuit current of the bare module at an irradiance of 1000 W/m2 at normal incidence, Cg is the geometrical concentration of the concentrator system, p is the angle of incidence of beam irradiance, and G is the global intensity perpendicular to the sun.

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

The measurements were performed during high irradiance and with a diffuse fraction of around 10%. The concentrator accepts all irradiance for solar altitudes exceeding 15° in the meridian plane, which means that the diffuse optical performance of the concentrator will be similar to that of a module tilted 20° with a correction for reflectance losses. This further means that the optical acceptance of diffuse irradiance will be around 70% of the beam efficiency. For this reason, the global intensity can be used in Eq (3) without significantly increasing the error of the model.

The optical efficiencies are functions of the projected angle of incidence in the transversal plane (i. e. the north south vertical plane) and the transmission of the glazing is given as a function of the conventional angle of incidence. Ray tracing represents the theoretical optical efficiency of the system at 85% reflectance. The graph labeled Optical efficiency — Isc in figure 6 contains contributions from measurements with corrections from ray tracing. The difference between measured values and ray tracing at 15°<0T<60° is due to resistive losses in the cells when the reflector is effective. The cells on the prototype absorber did not cover the whole width of the absorber, which meant that for angles above 40° the reflected beam partly missed the cell. The angulars above 40° are instead generated by ray tracing. The transmission of the glazing has also been included in the graph as it was used in the calculations of the annual output.

A simulation software, MINSUN (Chant and Hakansson 1985), estimated the annual output of electricity using the optical efficiencies at different angles of incidence. The model used to describe the incidence angle dependence of the system in MINSUN is defined by Eq. (4)

Пар, — RT (@T )fL (A )

RT describes the behaviour of the reflector as dependent of QT and fL the transmission of the window glass as dependent of 0, . QT is the projected angle of incidence in the transversal plane and в, is the conventional angle of incidence relative to the glass normal.

This model has previously been shown to describe the optical performance of an asymmetric compound parabolic reflector system such as this one well (Brogren et al, 2004).

The simulations show a 93% increase in electrical output for the concentrator module relative to the vertical reference module, which means that one square meter of this window annually would deliver 79 kWh of electric energy. The annual performance is 43% higher than that of an identical module tilted 20°.

The active area of the tested measured prototype covers only 87% of the total glazed area, which this has to be taken into consideration when an economical comparison is made with other systems. It is however possible to increase the active area of the window in a future full scale installation.