Here is a set of actual calculations showing the potential savings to an existing Opalco customer for using a TOU (Time Of Use) meter.
   
Assumptions:
The user consumes 1000 kwhrs/month
The user shifts peak period (6am to noon) power to the off peak period at the percentage shown in the column "peak kwhr percent", that is, the total kwhrs consumed during the peak are X% (see the column) of 1000 kwhrs. In the case of "september 2010" the total kwhrs consumed during the peak period for the month would be 150 kwhrs, as shown in the column "kwhr peak period".

Note that as the user consumes a higher percentage of power during the peak period, the savings decrease. At 40% peak period consumption, there is a negative savings by using a TOU meter.

if you discount the period midnight to 6am, then there are effectively 3 6 hour periods a day (6am-noon, noon-6pm, 6pm to midnight). If power were being consumed uniformly during these 3 6 hour blocks, then one could assume that the peak period block would only consume 33% of the power during the day. My experience for 2008 is that I use about 14-15% of my power during the peak period. (I can get you the exact number if you wish).

So, assuming the TOU meter user does use only 15% of their power during the peak, they save $19.14/month over what they would have paid, which would have been $101.30, so they are saving about 19% on their bill. For the year, that is $229 and change. (The actual total charges (TOU) numbers always round up to an even dollar because the program I wrote assumes the user contributes to project pal). 

If one assumes that a "normal" user, who did not make any effort to shift power off the peak, would consume 30% of their power during the 6am-noon peak period, then they would have consumed 300 kwhrs/month during this period, or approximately 10 kwhrs/day (assuming a 30 day month). The "peak-shaver" user, at 15% peak use, would use half this power, saving about 5 kwhrs/day during the peak. At even ten cents a kwhr (higher than the current rate in the NW), this is only 50 cents/day saved. Why bother?

For two reasons. First, as shown above, the user does save money, over $200 per year. Second, and this is the quiet but really big point, those 5 kwhrs/household/peakperiod/day add up if there are millions of players. If one million households in a metro area do this, that is 5 million kwhrs that do not have to be generated during the peak. Those reshuffled kwhrs means less need for power plant construction, particularly base load (think coal or nuclear) power plant production. Those reshuffled kwhrs buy time for the country to shift to renewables. 

One coal fired 500 mw power plant produces about 3.3 billion kwhrs/year. One million customers who shift power off peak for a year translates to about a half a power plant. There are well over 100 million households in America. If they all shifted power off peak, that translates to the equivalent of freeing up 50 existing power plants, thus buying time.