Some datacenters price power differently if the customer is buying a redundant pair of circuits , versus if the customer is buying two circuits for capacity. For example,
http://www.baymountain.com/colocation offers:
"Redundant 120V/20A circuit. 20A max capacity, 16A max normal usage, per pair."
...and
"Additional 120V/20A circuit. 20A max capacity, 16A max normal usage, per circuit."
...as separate products. Is it common to price these differently? If so, how do you enforce that the customer won't grow their use until the service is no-longer redundant? Do you automatically "upgrade" the customer to non-redundant pricing if their utilization exceeds 80% of a single circuit?