Spreadsheets have been used to compare some 90 possible small PV concentrat
or designs that might be suitable for use at remote sites. They have apertu
res of about 2 m(2), use BP Solar LEG cells, and employ small aperture modu
les to reduce heat sinking and construction costs. Designs include fixed V-
troughs and CPCs, single axis tracked cylindrical lens and mirror systems,
and two-axis tracked spherical-symmetry systems. Performance and volume pro
duction costs were estimated. Four promising systems were constructed as pr
ototypes:
(A) Point-focus Fresnel lenses, two-axis tracking; Cg = 32 X; and 69 X with
secondaries.
(B) Line-focus mirror parabolic troughs, one-axis tracking, Cg = 20 X.
(C) SMTS ('single-mirror two-stage'), one-axis tracking, Cg = 30 X.
(F) Multiple line-focus mirror parabolic troughs, E-W 1/day manual tracking
, Cg = 6 X.
The prototypes were tested at Reading, and three for up to a year's field t
rial at ZSW's test site, Widderstall, in Germany. The best system efficienc
ies, normalised to 25 degrees C and excluding the end losses of linear syst
ems, were 12.5%, 13.2%, 13.6% and 14.3% for collectors A, B, C, and F, resp
ectively. The collectors were practical and robust, and the performances of
collectors B, C and F are only 10% below the estimates in the spreadsheet
calculations. The best collectors have estimated production costs between 1
.5 and 1.8 US $/Wp, yielding energy costs at a good site (excluding BOS and
overheads) of between 5 and 7 cents/kWh (18 and 25 cents/MJ). On the same
cost basis a conventional PV array costs 4.3 $/Wp, and 18 cents/kWh (65 cen
ts/MJ). (C) 1999 ISES. Published by Elsevier Science Ltd. All rights reserv
ed.