The reference case was chosen early in this study only as a starting point from which to conduct sensitivity analyses and as a base for comparison of system configuration. It displays an adequate solar fraction and annual cost but, as was determined, is not an "optimum" system, though. it is representative of the various configurations studied.

Figure 2 shows a heat flow diagram for the reference case.

Figure 3 is a storage temperature profile for the year. The temperature variation is 13.4°C.

Figure 4 presents the results of varying different parameters using duct storage.

• solar fraction was highest in Boston and lowest in Madison,

but the greater solar fraction did not improve the cost of energy as can be seen from the following specific costs:

Madison 7.2 <(:/kWh

Copenhagen 7.8 t[:/kWh

Boston 8.0 <j:/kWh

• Increasing bore hole separation reduces the cost of storage per unit volume but also reduces the solar energy available to load. The optimum point appears to be a spacing of about 3m.

• Solar fraction increases with collector area but so does cost. The economic optimum occurs at a collector area of about 100,000m2 and little increase in solar fraction

is achieved above 150,000m2,

• Duct storage cost is relatively insensitive to volume.

Solar fraction can be increased cheaply by increasing storage volume and bore hole density.

Figures 5 to 9 present the results of varying different parameters using the tank model to simulate duct storage so that the three dimensional graphing function could be used. The shape of the cost surface and differences between curves are more meaningful than actual numeric values because of the inaccuracies in the tank/duct simulation. Tank volumes in the 3D graphs are one quarter of the corresponding duct storage volumes.

• As the cost of system components decreases (or the cost of fuel increases) a local optimum forms and develops into a global optimum.

• The position of an optimum changes little with component cost, but its intensity varies directly with cost.

• In cases where there is no optimum the shape of the system annual cost curve changes abruptly along a line of nearly constant collector area. At collector areas greater than this line, system cost increases sharply with collector area and is insensitive to storage volume. At collector areas less than this line, system cost increases with storage volume and is insensitive to collector area.

Updated: August 22, 2015 — 2:21 am