Quantifying the severity of the intermittencies in solar irradiation is important for (i) understanding the potential impacts of high solar power penetration levels on the electric grid and (ii) evaluating the need for technologies that may be necessary to complement solar power in order to balance the electric grid. This study uses a spectral method to distinguish between cloud-induced and diurnal cycle-induced transients and to quantify the severity of intermittencies occurring over a range of timescales. The method is used to quantify variability between specific sites as well as to evaluate the sensitivity of solar power variability to spatial diversification of the solar farm portfolio. Results indicate that increasing the spatial diversity of the solar farm portfolio reduces the magnitude of the fluctuations in power output as a fraction of the total system capacity. This behavior is associated with two forces: (i) a reduction in the influence of fluctuations occurring at an individual site on the total profile and (ii) should the sites in question be sufficiently spaced apart, the fluctuations in solar irradiation that each site exhibits are uncorrelated and do not generally add up in tandem at short timescales. These effects reduce the degree of uncertainty and variability associated with solar farm output and demonstrate a reduction in the maximum magnitude of solar power fluctuations for a given solar penetration level. The rate of increase of the maximum solar power deviation from the 1-h average associated with increases in desired solar penetration level decreases in an inverse exponential manner with the number of sufficiently spaced sites composing the solar farm portfolio. These results imply that a lower amount of regulation or energy storage capacity is needed to regulate solar intermittency if solar installations and the accommodating transmission infrastructure are designed and operated appropriately. Copyright (c) 2012 John Wiley & Sons, Ltd.