Since increasing demand will be hard to meet then rationing will be necessary by price or otherwise. By very heavy overdemand blackouts and brownouts would fulfill the rationing. Otherwise it could be done electronically. Regulation of electronics, etc. would also fulfill a role. All electronics would have to fulfill efficiency standards befoe coming on the market.

Perhaps possession and availability could be controlled as well to limit the constant growth of the numbers of electronic gadgets in households. So each household/person would have an energy quota of electrical and electronic articles (so much theoretical maximal usage capacity if they were running X hours per month). On the other hand sales of electronics could be based on ration coupons for which one had to wait and demonstrate need.

One might argue that this would not work in a capitalistic system but if blackouts are the alternative it might be accepted. In the same way that the market system allows bubbles in demand to create a roller coaster ride of over and undercapacity in the economy as a whole, demand vs. supply for electricity will be similarly unplannable when no large scale flexibiity to the upside exists due to lack of capacity. So US is left with the choice of unplanned decline, like in Pakistan or some sort of planned decline as maybe Japan or Switzerland might attempt.

Demand-side management is one way to reduce demand and level out peaks and valleys.

Enter frequency responsive heat pumps, much cheaper & easier to store energy as useful heat (especially if it can offset peak demand)

http://www.nytimes.com/2007/11/07/business/businessspecial3/07cutoff.html

Alternatively, a utility can pay demand-response companies to reduce use. For instance, a utility might pay Comverge $70,000 to $80,000 a year for each megawatt that the company can commit to eliminating from the load. By contrast, if a utility builds a power plant, it could pay $400,000 to $2 million for each new megawatt of capacity, depending on the type of power plant.

There are such low return rates on cash savings, and huge potential for capital intensive demand management and renewable / nuclear energy schemes which fit the pension investment model, it could all work very well.