B. J. Huang Professor. S. W. Hsieh Research Assistant. Department of Mechanical Engineering, National Taiwan University, Taipei, Taiwan 10764 An Automation of Collector Testing and Modification of ANSI/ASHRAE 93-1986 Standard The steady-state performance test of solar collectors using ANSI/ASHRAE 93-1986 Standard was revised and an automation for the testing was carried out in the present study in order that the test can be easily performed outdoors in areas with variable weather conditions. It was shown that the 95 percent settling time of the collector T 95 can be adopted as the time basis in the selection of steady-state period for the test. To make the best use of the time available for the testing, the steady-state period defined by ANSI/ASHRAE 93-1986 Standard was changed to the r 9S plus five minutes, or ten minutes, whichever is larger. To reduce scatter uncertainty in the test results, the test periodfor the efficiency calculation was chosen as the segment of the last five minutes in the steady-state period and a steadiness condition defined statistically was adopted. To shorten the time for each test run a PC-based expert testing system, which is completely automatic and requires no operator, was de- veloped in the present study. Using this expert system associated with the modified ANSI/ASHRAE 93-1986 Standard, we can effectively carry out the collector test at variable weather conditions with small scatter uncertainty and can substantially shorten the duration of a test. I Introduction ASHRAE 93-77 Standard (1977) and its revised version ANSI/ASHRAE 93-1986 Standard (1986) for the performance test of solar collectors is basically a steady or quasi-steady testing method and has been adopted worldwide as a reliable method for the rating of collectors. However, it has been noted that several problems may arise for testing outdoors in use with this standard and by manual operation. In many parts of the world, e.g., southeastern Asia and northern Europe, where the weather is usually not at clear sky conditions for most of the time in a year, the maximum time suitable for the steady-state testing is thus quite limited. If manual operation for the testing is employed in these areas, it might take a very long time (sometimes it takes a month, for example, in Taiwan) to complete a test. To cope with this steady-state testing problem, different transient testing methods have been proposed recently (Gillett et al. 1983; Emery and Rogers, 1984; Oreszczyn and Jones, 1987; Wang et al., 1987). Two different testing standards have even been established by U.K. (1987) and China (1984). How- ever, implementation of the Chinese transient testing method (Wang et al. 1987) requires a prior test of the collector to determine the impulse response or weighting function of the collector and needs a presumed second-order quasi-dynamic model of the collector for the filtering manipulation of the Contributed by the Solar Energy Division of THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS for publication in the JOURNAL OF SOLAR ENERGY EN- GINEERING. Manuscript received by the ASME Solar Energy Division, Nov. 1989; final revision, June 1990. measured dynamic data. Thus, it will be subject to errors in the final results due to the uncertainty in the determination of the impulse response function and in the use of the quasi- dynamic model. The British transient testing method basically employs the deconvolution approach (an off-line nonparametric system identification technique) to identify the collector impulse re- sponse function first and then proceed to evaluate the collector parameters through some mathematical manipulations. It has been known that system identification using the off-line de- convolution method involves tedious computation and hence probably needs a mainframe computer with large storage ca- pacity. Besides, the identified results will be biased if the system input (e.g., solar irradiation) is not persistently exciting or the measured output signals are corrupted by non-Gaussian ran- dom disturbances or noises (Hsia, 1977). It is conceivable that non-Gaussian external disturbances such as variations of am- bient temperature, wind speed/direction, etc., could be intro- duced during the collector testing period (one hour for each testing point required by the British standard). Thus, the im- pulse response function identified may be seriously biased. The steady or quasi-steady testing dictated by ASHRAE 93- 77 or ANSI/ASHRAE 93-1986 Standards is thus superior to the transient testing from the points of view described above. However, several problems may arise if the steady-state testing was performed not under the conditions of "clear sky" (i.e., substantially free of clouds). ANSI/ASHRAE 93-1986 Stand- ard is a revised version of ASHRAE 93-77 Standard by adding the "steadiness" requirement for solar irradiance, collector Journal of Solar Energy Engineering NOVEMBER 1990, Vol. 112/257 Copyright © 1990 by ASME Downloaded 05 Nov 2008 to 140.112.113.225. Redistribution subject to ASCE license or copyright; see http://www.asme.org/terms/Terms_Use.cfm