Not very likely, IMHO, but potentialy very serious if it does happen.
In a 2002 assessment of the electric power transmission system of the mainland United Kingdom, the consequences for multiple scenarios that varied by storm severity and location of the electrojet (flows of electric current in the Earth‘s ionosphere) were expressed as estimates for reactive power losses (Erinmez et al., 2002). Such an analysis provides information on the reactive reserve power requirements for a power system, but does not forecast whether the system can or cannot sustain such an event and maintain service. The third approach is to make that determination of continuity of service. Kappenman (2007) illustrated the potential effects of a geomagnetic storm of the estimated severity of a May 1921 storm. In that simulation, a geomagnetic storm was estimated to have introduced 4,800 nT/min at 50 degrees geomagnetic latitude, creating exceedingly large demand for reactive power to maintain voltage in the transmission system (over 100,000 megavolt-ampere reactive [MVAR]). This, in turn, could lead to ―probable large-scale voltage collapse‖ and ―major power grid blackout.‖ (Kappenman, 2007). North American Electricity Reliability Corporation (NERC) (2010) drew from the 2007 Kappenman analysis and stated that a severe storm ―could entail the potential for widespread damage to EHV transformers,‖ which could lead to ―prolonged restoration and long-term chronic shortages of electricity supply capability.‖
http://www.oecd.org/sti/futures/globalprospects/46891645.pdfGeomagnetic storms have it would seem a much higher probability than most would suspect, no doubt in part to the fact that the human experience of electricity grid supply within a large urban/city environment is no more than 100 years. i.e. in the age before the sky scraper and the electric powered elevator.
Measurement of the cyclic nature of large geomagnetic storms with 100-200 year periods are actually comparable to say a New York category 3 or 4 hurricane. Consequences could be much more severe that the recent storm called Sandy, which wasn't even technically a Hurricane. (although it did have a storm surge comparable to a category 1 hurricane direct hit across a wide area of the coast.)
Of course the New York and New Jersey area will probably fully recover by Christmas of this year because other areas of the country are able to function and respond to the crisis.
A wide scale Geomagnetic storm could potentially destroy many of the grid transmission transformers. These along with the HV transmission lines from the Power stations including Nuclear form the backbone of the Electrical Grid and not just the smaller transformers situated up a top of a localised wooden pole knocked over by the trees and the gale forces winds.
Rather than weeks or a month or two the ability to repair and reactivate the electrical grid could potentially take up to a year or even longer not just across one nation but many nations.
It actually has been a surprise to myself just how chaotic a modern city can become within 72hrs with even just a partial loss of the electrical grid. i.e. such as disruption to the just in time food, water and liquid fuel distribution and even communications and the lack of any fall back strategies being highly dependent of a computer packet network which is effectively only a decade or two old.
But then again I grew up in an age (1970/80s) where my Gran didn't have a telephone, had 3 TV channels and heated her home with a open coal fire with a ton of the highest quality anthracite at the side of her house stored in a large metal bunker.