GRIDRA

Lesson 2 of 7

The Day-Ahead and Real-Time Markets

7 min read

Electricity isn't bought and sold once — the same megawatt-hour typically passes through several rounds of trading, each closer to the moment it's actually delivered, each one correcting the forecast errors of the round before it. This lesson walks through that sequence.

Why not just trade once?

Demand and renewable output can't be forecast perfectly, and forecasts get more accurate the closer you get to real time. Trading in stages — a rough plan made a day ahead, refined with better information as delivery approaches — lets the system commit expensive-to-start generation (like large thermal plants) early enough to be ready, while still leaving room to correct the plan cheaply as forecasts improve.

The day-ahead market

Generators submit bids (the price they're willing to accept to produce a given quantity) and, in many markets, large consumers submit demand bids, all for delivery the following day, broken into hourly blocks. The system operator solves an optimization — essentially, merit order dispatch across the whole network subject to transmission limits — and produces a schedule: which generators run, how much, and at what price. This schedule is financially binding, but it's still a forecast of tomorrow, so it's rarely exactly right.

The real-time (balancing) market

As delivery hour approaches and finally arrives, actual conditions inevitably differ from the day-ahead forecast — a cloud bank drops solar output, a plant trips unexpectedly, demand runs hotter than predicted. The real-time market (sometimes running every five minutes) continuously re-solves the same kind of optimization using live conditions, dispatching additional generation up or down to keep the system balanced. Any deviation from what a generator or buyer scheduled in the day-ahead market is settled at the real-time price.

Two prices, one physical event

A single hour of electricity typically has both a day-ahead price and a real-time price, and they can differ meaningfully — real-time price spikes happen precisely when the system is under the most stress and needs the last, most expensive generator willing to help close the gap. This is also exactly the moment when demand response and fast-acting storage (from the Renewable Energy track) become most valuable.

How the price itself is set: marginal pricing

In most of these markets, every generator that's dispatched gets paid the same price — the bid of the most expensive generator needed to meet demand at that moment (the "marginal" unit), not each generator's own bid. This is called uniform (or marginal) pricing. It sounds generous to cheaper generators, but the underlying logic is that it gives every generator, including cheap ones, an incentive to bid their true cost rather than try to guess and game a pay-as-bid price.

Key takeaways

  • Electricity trades in rounds — day-ahead first, then real-time — refining the plan as forecasts improve.
  • The day-ahead market produces a financially binding schedule based on forecast conditions.
  • The real-time market continuously rebalances against actual conditions, often every few minutes.
  • Most markets use marginal (uniform) pricing: all dispatched generators are paid the highest accepted bid, not their own.

Further reading

  • D. Kirschen & G. Strbac, Fundamentals of Power System Economics, Wiley — detailed treatment of day-ahead and real-time market clearing.
  • S. Stoft, Power System Economics: Designing Markets for Electricity, Wiley-IEEE Press — the classic derivation of marginal pricing logic.