| dc.description.abstract | Regression analysis is used to develop models for the
total number of miles of EHV line needed in a power system.
These models are not meant to be used to design a transmission system but to examine existing EHV systems to see
if general patterns or relationships exist on a system-
wide basis.
The U.S. was divided into regions and data on regional
load and generation characteristics was obtained for the
years 1960, 1970, 1980, and 1990. Regressions were performed
on this combination cross-sectional and time-series data to
develop equations for circuit miles of EHV line and for
gigawatt-miles (miles of line X power-carrying capability)
of EHV line. t- and F-tests were used to determine the
statistical significance of the model parameters.
The independent variables (system characteristics)
found to be most significant in determining the miles of
EHV line include the generating capacity of the system,
the area (square miles) of the region, the percent of area
that is metropolitan, the number of generating plants in
the system, the percent of energy used for industrial
purposes, the percent of generating capacity which is hydro-
electric, and the average distance between plants and load
centers.
-The equations developed are multiplicative, of the
form miles of line = K7Xf i4' where the Xi's are system
characteristics. "Expansion" models attempting to determine
the additions to the existing grid in a 10 year period were
also postulated. However, they are not as statistically
significant as the "static" models.
Methods of using the models to investigate new plant
siting strategies, such as power parks or offshore nuclear
plants, are discussed. Analysis of the effect of trends
in plant siting and construction, such as the growing
scarcity of potential plant sites near load centers,
is also mentioned. The limitations, uses, and possible
extensions of this type of model are also described. | en |
