The elegant math that tells you electricity's true cost.
Here's a puzzle: two power plants are running simultaneously, both burning natural gas, both with marginal costs of $40/MWh. One is in Pittsburgh. One is in Philadelphia. Should both receive the same price?
In a world without transmission constraints, yes. Power can flow freely from Pittsburgh to Philadelphia, so the price difference between the two cities is effectively zero. But what if the transmission line connecting them is fully loaded? Then the Pittsburgh plant can't deliver more power to Philadelphia — even if Philadelphia desperately needs it. The market needs to signal this scarcity. Philadelphia's price should be higher than Pittsburgh's.
This is the insight behind Locational Marginal Pricing (LMP), the pricing methodology used by PJM, MISO, CAISO, ISO-NE, NYISO, and SPP.
Every LMP is composed of three parts:
Energy component: The cost of an additional megawatt-hour of energy, ignoring transmission. This is the system-wide shadow price of the supply-demand balance constraint — effectively the marginal cost of the last generator dispatched.
Congestion component: The cost of transmission congestion. When a line is constrained, delivering power across that line is more expensive than delivering it uncongested. The congestion component reflects this additional cost — and can be positive or negative depending on which direction the constraint binds.
Loss component: The cost of electrical losses in the transmission network. Moving power over long distances loses some fraction to heat. LMP accounts for these losses by making electricity at distant locations slightly more expensive.
An LMP can be positive, negative, or zero. During periods of high renewable generation and low demand — common in the Pacific Northwest during spring floods — LMPs can go sharply negative, because generators must pay to inject power that the grid cannot absorb.
The congestion component at a given bus is the difference between that bus's LMP and the system energy price. Its sign depends on whether the bus is on the "receiving" or "sending" side of a constrained line.
Positive congestion component — the bus is on the import-constrained (load) side of a bottleneck. Power is trying to flow in but the transmission line is full, so local supply is scarce and the price is bid up above the system energy price.
Example: A major gas plant sits west of a congested mountain crossing. East-side cities want that cheap power, but the 500 kV line is at its thermal limit. The eastern buses see LMPs of $75/MWh while the system energy price is $40/MWh — a +$35/MWh congestion component reflecting the scarcity of import capacity.
Negative congestion component — the bus is on the export-constrained (generation) side of a bottleneck. There is surplus cheap generation locally, but it cannot be exported because the line is full. The local price is depressed below the system energy price.
Example: Overnight wind output in West Texas is high, but the HVDC ties to the eastern grid are saturated. Wind generators in that zone bid negative or zero, driving the local LMP to $5/MWh while the system price is $30/MWh — a −$25/MWh congestion component reflecting stranded surplus generation.
One of LMP's most elegant features is that it enables Financial Transmission Rights (FTRs): financial instruments that hedge the congestion cost between two locations. An FTR holder is entitled to collect the congestion revenue that flows between two points on the grid. If you own an FTR from Pittsburgh to Philadelphia and the congestion component between those points is $10/MWh, you collect $10/MWh on your FTR — offsetting the higher price you pay to buy power in Philadelphia.
FTRs allow market participants to lock in the delivered cost of power across congested paths, reducing price risk. They're also traded speculatively by financial firms that understand the grid's congestion patterns better than anyone.
PJM Interconnection, April 1, 1998. When PJM launched nodal pricing across its transmission network, it did something no U.S. grid operator had done before: it assigned every generating bus its own real-time electricity price, calculated from the cost of serving one additional megawatt-hour at that specific location. The theoretical foundation came from Harvard economist William Hogan's 1992 paper, which proved that nodal prices simultaneously clear energy markets and send honest signals about transmission congestion. On the first day of operation, price differences between nodes were immediate and dramatic — generators near load centers commanded premiums over those facing congested transmission paths. Transmission owners, generators, and industrial customers all had to rebuild their understanding of electricity economics around a new question: not just what electricity cost, but where it was delivered. PJM's implementation became the international template for nodal pricing.
PJM — LMP Overview and History Harvard Kennedy School — William Hogan, "Contract Networks for Electric Power Transmission" (1992)