If you work in power project development, asset management, or energy trading, you will encounter the term Locational Marginal Pricing (almost always abbreviated to LMP) within days of stepping into US wholesale electricity markets. It is the single most important price signal in those markets, and yet it remains poorly understood by many professionals who rely on it every day.
This guide explains LMP from first principles: what it is, how it is calculated, what its three components mean, and why it matters directly for merchant revenue.
The Short Definition
Locational Marginal Pricing is the cost of serving the next incremental megawatt-hour (MWh) of electricity demand at a specific location on the grid, at a specific point in time.
Three words in that sentence carry most of the weight: incremental, specific location, and specific time. LMP is not a single average market price. It is a granular, real-time price that varies across hundreds or thousands of nodes on the transmission network, and it changes every five minutes in real-time markets or every hour in day-ahead markets.
Why Location Matters: The Node
Electricity cannot be stored at scale and travels through physical wires that have finite capacity. When power flows from a generator in the west to a load center in the east, it must traverse transmission lines. If those lines are already carrying as much electricity as they safely can, the operator cannot simply dispatch the cheapest generator on the other side of the constraint. It must find a generator closer to the load, even if that generator is more expensive.
This is why LMP varies by location. Each node on the grid (a physical point where generators, loads, or transmission lines connect) can have a different LMP. In PJM, the largest US power market, there are over 10,000 nodes. In CAISO (California), several thousand. In ERCOT (Texas), several hundred settlement points plus load zones.
A generator’s revenue depends on the LMP at its node, not on a market average. A wind farm in a congested corner of West Texas can see dramatically lower prices than a gas peaker in Houston, even at the same moment.
The Three Components of LMP
Every LMP value is the sum of three distinct components. Understanding each component is essential for diagnosing price differences and forecasting merchant revenue.
1. Energy Component (System Marginal Cost)
This is the baseline cost: the marginal cost of the next MWh if the grid had no transmission constraints and no electrical losses, a theoretical “copper plate” grid. It reflects the bid of the marginal generating unit dispatched to serve total system demand at that moment.
The energy component is the same at every node within a market at a given interval. If the system marginal cost is $35/MWh, every node shares that $35 as a starting point.
2. Congestion Component
When transmission lines are operating at their limits, the system operator cannot freely move power from cheap generators to load. It must dispatch more expensive generators near the constrained area and back down cheap generators behind the constraint. The price difference this creates is the congestion component.
- Nodes downstream of a congested line (toward the load) typically see positive congestion values: prices are pushed higher because only expensive local generation can serve demand.
- Nodes upstream (behind the constraint) see negative congestion values: cheap generation there is backed down, and the price drops below the system marginal cost.
Congestion is often the most volatile component of LMP and the one most relevant for financial transmission rights (FTRs) and congestion revenue rights (CRRs). Learn more about how congestion shapes locational capacity and deliverable capacity values.
3. Loss Component (Marginal Loss Component, or MLC)
Electricity is lost as heat when it travels through transmission lines, a physical reality called resistive losses. The loss component of LMP accounts for the extra cost of generating power to compensate for what is lost in transit.
Generating units electrically close to the load they serve face lower marginal losses; units far away face higher ones. In practice, the loss component is usually the smallest of the three components but can become meaningful over long distances.
The formula is simple:
LMP = Energy Component + Congestion Component + Loss Component
Real-Time vs. Day-Ahead LMP
US wholesale markets operated by ISOs (Independent System Operators) and RTOs (Regional Transmission Organizations) run two parallel markets:
- Day-Ahead Market (DAM): A financial forward market where generators, loads, and traders submit bids and offers for each hour of the following day. LMPs are settled hourly and known the evening before.
- Real-Time Market (RTM): An operating market that dispatches generation every five minutes to balance actual supply and demand as conditions evolve. Real-time LMPs are 5-minute prices and can be significantly higher or lower than day-ahead prices.
Most physical generators settle their energy revenue against the day-ahead LMP at their node. Any deviation between their scheduled output and their actual output settles against real-time LMP. This means real-time price volatility (spikes during heat waves, negative prices during wind oversupply) directly impacts generators that are out of balance.
How ISOs Calculate LMP
ISOs use a mathematical optimization called Security-Constrained Economic Dispatch (SCED). SCED solves for the least-cost combination of generators that meets total system demand while respecting every transmission security constraint on the grid simultaneously. The shadow prices (the marginal value of relaxing each constraint by one unit) become the congestion components at each node.
This calculation runs continuously. In ERCOT and PJM’s real-time markets, it runs every five minutes. The result is a set of nodal prices that reflect the true economic cost of serving load at every point on the system. See how this connects to broader concepts of grid reliability and supply and demand dynamics.
Why LMP Drives Merchant Revenue
A merchant generator (one without a long-term power purchase agreement) sells its output at the LMP at its node. Its revenue per MWh is literally equal to that LMP. This creates several critical implications:
- Location risk: Two identical gas plants in the same state can earn materially different revenues if they are on opposite sides of a chronic transmission constraint.
- Time risk: LMP varies by hour, season, and year. Peak summer afternoon prices in Texas can be 10x or 100x midnight prices. Annual merchant revenue depends heavily on when a plant can generate relative to the price stack.
- Market design risk: Rule changes (new transmission builds that relieve congestion, storage mandates that shift the dispatch stack, renewable integration policies) can fundamentally reshape LMP patterns at a given node.
For project finance, merchant revenue is projected using LMP merchant curves: forward-looking price forecasts at specific nodes, built from fundamental dispatch modeling of the ISO. The accuracy of those curves (and understanding their assumptions) is central to generation asset valuation and any renewable asset valuation. Understanding merchant vs. PPA offtake trade-offs is equally important at the development stage.
LMP Across the Major US ISOs
| ISO / RTO | Region | No. of Pricing Nodes (approx.) | Key Characteristics |
|---|---|---|---|
| PJM | Mid-Atlantic, Midwest | ~10,000+ | Largest US market; deep FTR market; capacity market (RPM) |
| CAISO | California | ~4,000+ | High renewable penetration; duck curve dynamics; EIM integration |
| ERCOT | Texas | ~4,000 (settlement points + zones) | Energy-only market; scarcity pricing via ORDC; weather-driven spikes |
| MISO | Midwest, South | ~35,000+ | Large footprint; complex transmission topology; capacity accreditation reforms |
| SPP | Great Plains | ~8,000+ | High wind penetration; growing renewables; integrated marketplace since 2014 |
| NYISO | New York | ~3,500+ | 11 load zones; NYC congestion premium; capacity zone structure |
| ISO-NE | New England | ~8,000+ | Natural gas import constraints; winter fuel security; Forward Capacity Market |
Negative LMPs: When the Price Drops Below Zero
One of the most counterintuitive features of LMP markets is that prices can go negative. This happens when there is too much generation on the grid and not enough demand, a situation increasingly common during periods of high wind or solar output with low overnight demand.
In a negative price environment, generators with must-run characteristics (wind and solar with production tax credits, nuclear plants that cannot easily cycle) are willing to pay to keep generating. The LMP reflects this: the clearing price becomes negative, and load (buyers of electricity) are effectively paid to consume.
For renewable developers, negative price hours are a key risk factor in revenue modeling. A solar project in CAISO that generates primarily during midday hours is directly exposed to negative or near-zero prices during high-solar periods, a phenomenon explored in depth in our guide on the CAISO duck curve.
Key Takeaways
- LMP is the marginal cost of serving the next MWh at a specific node, at a specific time.
- It has three components: energy (system-wide), congestion (transmission constraints), and losses (electrical distance from load).
- LMP varies by node and changes every five minutes in real-time markets.
- Merchant generator revenue = output x LMP at the generator’s node.
- Day-ahead and real-time LMPs can differ significantly; generators settle against both.
- Negative LMPs occur during oversupply and are a key risk for renewables.
- Understanding LMP is foundational for generation asset valuation, capacity merchant curve analysis, and energy trading.
