Reliability data for the initially installed sets for each collector is presented in Tables 1 through 3. The data includes failures resulting in loss of vacuum due to causes other than the glass cracking. The set used for the Novel ICPC is for the collectors in the first half of the production run. Several distributions commonly used in reliability modelling were fit to the data in the tables. The best fits were obtained with the lognormal distribution. Survival probabilities were then calculated from the fitted lognormal distributions. These probabilities are shown in Table 4 for four year periods extending to 28 years.

Collector Performance

Novel ICPC

Prior to the start of the 1998 Sacramento demonstration, individual fourteen tube modules were tested on Sandia National Laboratory’s two-axis tracking (AZTRAK) platform. The efficiency curve for insolation of 1000 W/m2 was given in [16] as

SHAPE * MERGEFORMAT

but with no control deficiencies are shown in figure 7. Here, daily efficiencies as high as

54.5 percent were achieved.

During 2002 the ICPC collectors were operated in the 50 to 70C temperature difference range [13, 14]. Performance closely matched the 2001 performance shown in figure 7. Allowing for some additional tubes that had lost vacuum, array performance during 2003 was essentially unchanged while operating in the lower temperature difference range [15].

Corning and Philips VTR261

During the 1982/83 period there ———- Regressed 1998 Banks Data

was only one Solarhaus tube with vacuum loss (one in the

Corning array) and no other Figure 6: 2001 Daily Collection Performance at a 90 to factors degrading performance. 110C Collector to Ambient Temperature Difference.

In 1982/83 the Corning and

Philips VTR261 collector arrays produced maximum daily solar collection efficiencies of about 56 percent and 53 percent respectively when operating in the 45C to 60C temperature difference range [17].

As of 1996/97 six tubes in the Corning array and one in the Philips array had vacuum loss and the Philips VTR261array reflectors and tubes backs had developed deposits due to growth of algae [18, 19]. The consequent reduction in performance is shown in table 5.

Daily performance in 1996/97 is shown in figures 8 and 9. The degree of degradation of performance during these years is nearly the same as for the novel ICPC operating in 2001. However, the degradation in the novel ICPC performance was characterized in a higher average daily temperature difference range (90 to 110C versus 15 to 60C) and for a higher average daily insolation level (7 versus 4 kWh/m2). Nevertheless, the two factors should offset somewhat so that the results for the novel ICPC in figure 7 may be roughly compared to those for the Corning and Philips VTR261 in figures 8 and 9.

Comparisons

Table 6 contrasts some of the important characteristics of the three different evacuated collectors that have been brought out in the paper. Some interesting observations are

• Higher survival rates are seen when the evacuated tubes are fabricated at some level of automation as opposed to being hand built.

High operating temperatures (130 to 160C) experienced by the novel ICPC was not a factor in tube failures from glass cracks

Table 5: Average Daily Performance of the Solarhaus Freiburg Collectors Corning Philips VTR261 1982/83 1996/97 1982/83 1996/97 Energy Delivered (kWh/m2) 2.18 1.754 2.01 1.76 Insolation (kWh/m2) 4.18 4.07 3.83 4.14 AT*Day Length (Kh) 342 275 320 237 Efficiency 0.521 0.431 0.525 0.426 Change in Efficiency -0.17 -0.19

SHAPE * MERGEFORMAT