Transmission & congestion

The short version

Electricity wholesale prices are not the same across the grid. At any given moment, the locational marginal price (LMP) at one substation can differ by hundreds of dollars per megawatt-hour from the LMP at another substation a hundred miles away. The difference is called basis, and it exists because transmission lines have physical capacity limits. When those limits bind, the lowest-cost generation can't reach the load that needs it — and the system has to dispatch more expensive local generation instead. The cost difference shows up as congestion, paid by load in the constrained area and (in a properly designed market) flowing back to holders of Financial Transmission Rights.

How LMPs work

Every ISO that uses nodal pricing — PJM, MISO, ERCOT, SPP, CAISO, ISO-NE, NYISO — calculates an LMP at every node on the transmission network, typically thousands of nodes per ISO. The LMP at any node equals the marginal cost of supplying one additional megawatt-hour at that specific location, given all current physical and operational constraints. Mathematically, LMP decomposes into three components: marginal energy (the system-wide cost of the next MWh), marginal congestion (the additional cost imposed by transmission constraints), and marginal losses (the cost of resistive losses delivering power to that location).

In an uncongested grid, all nodes would have similar LMPs differing only by the small losses component. In a heavily congested grid, LMPs can differ dramatically. ERCOT's West Hub vs Houston Hub basis has at times exceeded $300/MWh during specific intervals when wind generation is high in West Texas but transmission to Houston is constrained. CAISO's South-of-Path-26 vs North-of-Path-26 basis reflects congestion across the constraint between Northern and Southern California. PJM's Western Hub vs DOM Zone basis reflects the congestion driving billions in 2024 transmission upgrade costs.

The data center congestion story

The fastest-growing source of US transmission congestion is the geographic mismatch between new electricity demand (concentrated in Northern Virginia, Phoenix, Dallas-Fort Worth, Columbus, and a handful of other data center hotspots) and the transmission infrastructure to serve it. The Union of Concerned Scientists found that local transmission upgrades made specifically to connect data centers added more than $4.3 billion to customer bills in PJM's seven core states (Illinois, Maryland, New Jersey, Ohio, Pennsylvania, Virginia, West Virginia) in 2024 alone — with billions more coming as additional projects move through construction.

Crucially, under existing PJM rules, these connection costs are allocated to all customers of the same utility, not just to the data center loads themselves. This has prompted regulatory and legislative pushback in multiple states, with proposed tariff changes to assign more of the cost to the load drivers. Similar dynamics are playing out in ERCOT (where the 2025 Texas legislature passed multiple bills addressing large load tariffs), MISO, and other regions. The cost allocation question is unresolved — and the answer will shape billions of dollars of cost flow over the next decade.

Hedging congestion with FTRs

For commercial buyers with persistent exposure to congestion (a load served at one location, paying a basis premium over the system hub price), Financial Transmission Rights provide a hedge. An FTR is a financial contract — typically purchased through monthly, seasonal, and annual auctions held by each ISO — that pays the holder the difference between LMPs at two specified points. If you hold an FTR from a generation hub to your load zone and basis between the two widens, your FTR settles for cash payments offsetting the higher LMP you're paying on the physical side.

FTRs are funded by congestion charges collected from market settlements — when LMPs differ across the grid, that "congestion rent" flows to FTR holders proportional to their positions. This makes the FTR system self-financing in expectation, though in practice FTR auction revenues sometimes exceed and sometimes fall short of congestion collections, with the difference allocated to load. FTR strategy has become a sophisticated procurement discipline at large industrial loads, and an entire trading industry exists around taking speculative FTR positions.

What grid-enhancing technologies can do

Building new transmission takes 10-15 years (see NEPA & transmission permitting). Grid-enhancing technologies (GETs) can increase the capacity of existing lines in months, not years, at a fraction of the cost. Four major categories: Dynamic line ratings — sensors that measure actual conductor temperature and ambient conditions, allowing line capacity to be set based on real-time conditions rather than conservative static ratings (typical static ratings underestimate capacity 20-40% of the time). Advanced power flow control — devices that actively route current away from congested lines onto parallel paths with capacity available. Advanced conductors — carbon-composite-core conductors that carry more current at higher temperatures without excessive sag, often allowing line ratings to be doubled by reconductoring without rebuilding towers. Topology optimization — software that reconfigures network switching to find optimal flow patterns.

FERC Order 1920, issued May 2024, requires transmission planners to formally consider GETs as alternatives to new lines. Adoption has been slower than advocates would like, but pilot deployments are accumulating evidence that GETs can deliver 20-40% capacity gains at 5-15% of new-line construction cost. For commercial buyers, GETs deployment in your transmission region matters because it directly affects how quickly congestion in your area gets relieved — and therefore the trajectory of your basis charges over the next several years.