Okay, take my local station Squires Gate:
Doing the sums on that station it has:
10 x (assumed) 75w lamps, burning for 16 hours a day regardless of time of year. That burns 12kWh per day @ 0.11p / kWh = £1.32 a day, or £481.80.
If we assume a photocell is fitted and use a rough rule of thumb guide of 16hrs burn for 183 days a year and 11hrs burn for 182 days, we get a cost of £406.725 a year.
Roll that across say 200 stations that may fit the same sort of design as Squires Gate, thats a saving of £15,015.
Add in the carbon saving of £4,095 thats a saving of £19,110.
Now lets assume a payback period of any investment of five years, that means the cost to fit a photocell to each station would have to be less than £477.75.
I will freely admit that it will be more complicated on some stations than others, but simple stations if it costs more than £477.75 to fit a single photocell to the incoming supply point, I'm in the wrong job.