Most spot electricity markets establish locational marginal prices (LMPs) that reflect the unique value of energy at each location or node on the network. Components of the LMP include the system marginal price, losses, and the cost of transmission congestion.
When a transmission facility reaches its operating limit, higher-cost units must be dispatched instead of lower-cost ones so that the facility does not over load. This is called a binding transmission constraint. If meeting an increment of demand at a location increases flow on a binding transmission constraint, the congestion component will be positive, reflecting the redispatch cost of relieving congestion to serve the increment of load. Likewise, if meeting an increment of demand at a location decreases flow on a binding transmission constraint, the congestion component will be negative, reflecting the easing of congestion and the associated redispatch cost savings.
The congestion component of the LMP measures the change in redispatch cost on a binding constraint from supply one MW of demand at a location. More technically, the congestion component of the LMP is based on (1) the marginal cost savings from relieving the constraint (the shadow price) and (2) the flow on the constraint from injecting at the specific location. The shadow price of a transmission constraint represents the marginal change in total production cost from relieving the constraint by one MW. The flow on the constraint at a specific location is measured by “shift factors”. Because electricity will flow over different facilities based on physical properties of the network, a generation shift factor (GSF) is used to measure what portion of marginal production at a location will contribute to flows over an individual constraint. For example, if the GSF at a location is 5 percent and the shadow cost is $60 per MW of flow, then the congestion component of the LMP is $3 per MWh.
Typically, a system will have multiple binding transmission constraints at any given instant in time so the congestion component at a location will reflect the congestion cost associated with any number of constraints by adding the product of the shadow cost and GSF for each constraint. Some of these individual congestion cost values will be positive and others negative, so that the LMP will reveal the net system-wide impact of supply load at the location.