Electricity is traded like a commodity, but it does not behave like a warehouse good. Demand and supply must remain balanced almost continuously. Transmission lines have limits. Generators cannot always start instantly. Some plants ramp slowly. Storage has state of charge. Renewable output changes with weather. A market therefore cannot simply choose the cheapest seller and ignore the electrical network.
Electricity markets and dispatch work by combining economic bids with physical constraints to decide which resources produce, consume, store or move electricity at each operating interval. Prices emerge from the marginal cost of serving the next unit of demand subject to network, reliability and operating limits.
Wait, what? The cheapest generator does not always run first
A generator may offer cheap energy but sit behind a congested transmission line. Another plant may be more expensive but located close to demand. A cheap unit may be unable to ramp fast enough. A thermal plant may require a minimum stable output once started. A battery may be cheap now but need to preserve charge for a later scarcity period.
Dispatch is therefore a constrained optimisation problem, not a simple price ranking.
The direct answer
Generators and other resources submit offers describing quantities, prices and operating capabilities. Demand is forecast. The system operator or market engine chooses a feasible combination that meets demand and reserves at minimum expected cost while respecting transmission and technical limits. The marginal accepted offer often influences the market-clearing price.
The full sequence is: forecast → bids → commitment → dispatch → balancing → settlement → verification.
Bids and offers
A generator can submit an offer specifying how much power it is willing to produce at different prices. Loads may also submit demand bids. Storage can participate as both a load while charging and a generator while discharging.
The market rules determine whether bids reflect energy only or also startup costs, minimum output, reserve capability and other services.
Merit order
In a simplified market without network constraints, offers can be sorted from lowest to highest marginal price. The system accepts resources until demand is met. This ranking is called the merit order.
Wind and solar often bid low because their short-run fuel cost is near zero. Gas or oil plants may bid higher because fuel and emissions costs are material. But the real dispatch adds technical and network constraints to this simple stack.
Marginal pricing
In many wholesale electricity markets, the market-clearing price is linked to the marginal resource needed to serve one additional unit of demand. Generators accepted below that price can earn a margin above their short-run offer cost.
This design rewards low-cost generation while allowing higher-cost resources to set price when scarcity increases. Exact settlement rules differ by market.
Why marginal price can be much higher than average cost
Suppose most demand is served by low-cost solar, nuclear or efficient generators, but the final 50 MW requires an expensive peaking plant. If that plant sets the marginal price, all cleared energy in the relevant market interval may be settled at the higher price.
The price reflects the cost of the next unit, not the average production cost of every unit already running.
Unit commitment
Large thermal generators cannot always start or stop instantly. They may have startup costs, minimum run times, minimum output levels and ramp limits. The operator must therefore decide hours ahead which units should be online.
This is the unit-commitment problem. A plant may be committed even if it is not the cheapest for every interval because starting it later would be impossible or more expensive.
Economic dispatch
Once units are committed, economic dispatch determines their output levels over shorter intervals. The optimisation minimises operating cost while satisfying demand, reserves, ramp limits and network constraints.
Dispatch can run every five minutes or on another interval depending on market design. Faster dispatch helps the system respond to changing wind, solar and demand.
Transmission congestion
Cheap electricity cannot always reach every customer. Transmission lines have thermal, voltage and stability limits. When a corridor is constrained, the market may dispatch more expensive local generation while reducing output elsewhere.
Congestion therefore creates location-dependent value. A megawatt near a constrained demand centre can be worth more than a megawatt trapped behind a full line.
Locational marginal pricing
Some markets calculate prices at different network nodes. The locational marginal price reflects the marginal cost of serving additional load at that location, including energy, congestion and sometimes marginal losses.
Two neighbouring regions can therefore have different prices during congestion even though they are part of the same larger grid.
Balancing markets
Day-ahead forecasts are never perfect. Real demand and generation deviate from schedules. Balancing markets procure or dispatch resources that increase or decrease output in real time.
Batteries, flexible generators, imports and demand response can all participate if market rules allow. The balancing price can become high when the system has little remaining flexibility.
Ancillary services
Energy is not the only product a power system needs. It also needs services that maintain frequency, voltage, reserves and restart capability. These are commonly grouped as ancillary services.
- frequency response,
- operating reserve,
- reactive power and voltage support,
- black-start capability,
- inertia or equivalent fast frequency support.
A resource can therefore earn value even when it is not producing much net energy.
Scarcity pricing
When reserves become scarce, market prices can rise sharply. Scarcity pricing signals that another unit of demand is expensive to serve because the system is close to its physical limits.
High prices can encourage flexible demand to reduce consumption and storage to discharge. Over the long term, scarcity revenue can support investment in capacity that runs only rarely.
Price caps
Markets often impose price caps to limit extreme outcomes. Caps protect consumers from unlimited short-term prices but can weaken the revenue signal for rarely used peaking capacity.
This tension helps explain why some systems use separate capacity mechanisms or reliability payments in addition to energy markets.
Negative prices
Prices can become negative when supply is abundant and some generators prefer paying to remain online rather than shutting down. Startup costs, subsidies, minimum-run constraints and excess renewable generation can contribute.
A negative price is not negative physical energy. It is an economic signal that the system currently values additional generation below zero.
Curtailment and markets
When supply exceeds what demand, storage and networks can accept, renewable output may be curtailed. In market systems, low or negative prices can provide an economic reason to reduce production before direct operator curtailment is required.
Persistent curtailment signals that storage, transmission or flexible demand may have increasing value.
Storage arbitrage
A battery can charge when prices are low and discharge when prices are high. This is energy arbitrage. The battery also affects prices by increasing demand during surplus periods and increasing supply during scarce periods.
Arbitrage profits must cover round-trip losses, degradation and capital cost. Perfectly flattening every price difference would eliminate the very spread that pays for the storage.
Demand response in markets
Consumers can become active market participants by reducing or shifting demand when prices are high. A factory may pause a flexible process. A building may adjust cooling. A fleet may delay charging.
This makes demand part of the dispatch stack rather than a fixed quantity that supply must always follow.
Renewable forecasting and bidding
Wind and solar plants forecast their expected output before market intervals. If actual output differs, imbalance charges or balancing settlements may apply depending on market rules.
Better forecasting therefore has economic value as well as operational value.
Fuel prices affect electricity prices
When gas-fired plants are marginal, changes in gas price can move wholesale electricity prices strongly. Carbon prices, emissions rules and plant efficiency also affect offers.
This is one way global fuel markets propagate into local electricity bills even when the physical grid itself has not changed.
Market power
If a generator controls capacity that is essential during constrained periods, it may have the ability to raise offers above competitive levels. Market monitoring, bid rules and competition policy are therefore important parts of electricity-market design.
A technically feasible market can still perform poorly if participants can manipulate scarcity or if entry barriers prevent competition.
Capacity markets
Some systems pay resources for committing dependable capacity years ahead. The objective is to ensure enough capacity exists even if energy-market revenue alone is too uncertain.
Capacity markets must decide how to value storage, demand response, imports and variable renewables fairly. Nameplate megawatts are not all equally dependable.
Contracts and hedging
Retailers and large consumers often use forward contracts, futures or bilateral agreements to reduce exposure to volatile spot prices. Generators can lock in revenue while buyers lock in cost.
Hedging changes financial risk but does not eliminate physical scarcity. If generation is unavailable, someone must still balance the system.
Markets follow physics
Electricity markets are economic layers built on a physical machine. Prices cannot make a conductor carry unlimited current. A contract cannot make a battery discharge after it is empty. A bid cannot start a cold generator instantaneously.
Good market design makes economic incentives align with physical needs rather than pretending the two can be separated.
Singapore as a market case
Singapore has a dedicated wholesale electricity-market architecture in which generators, retailers and consumers interact under local market rules and system constraints. That detailed mechanism belongs to the existing Singapore-specific market article.
The general lesson here is broader: every electricity market must translate bids into a dispatch that the physical grid can actually carry.
Three worked examples
1. Cheap generator behind congestion
A wind farm bids nearly zero. Its transmission line is full. The market reduces wind output and dispatches a more expensive local gas plant near demand. The higher price reflects the network constraint, not a failure to notice the cheap generator.
2. Battery arbitrage
Prices are low at noon because solar output is abundant. The battery charges. Prices rise in the evening as solar falls. The battery discharges. It earns the spread while smoothing the system’s net load.
3. Scarcity interval
Demand is high and several plants are unavailable. The final available resource is expensive. Prices rise sharply. Demand response activates and storage discharges. The high price is a signal of scarce marginal capability.
Common misconceptions
- Dispatch is not always simple cheapest-first merit order.
- The market price often reflects the marginal unit, not the average cost of all generation.
- Negative prices do not mean electricity has negative energy.
- Storage participates on both the demand and supply sides.
- Network congestion can make prices differ by location.
- Ancillary services can be valuable even when little net energy is produced.
- Markets cannot override physical constraints.
A universal market-dispatch audit
- Forecast demand and renewable output.
- Collect bids and technical capabilities.
- Commit slow-starting resources.
- Dispatch feasible resources to meet energy and reserves.
- Respect transmission, ramping and minimum-output constraints.
- Calculate market or nodal prices according to the rules.
- Balance forecast errors in real time.
- Settle energy and ancillary services.
- Monitor market power and performance.
- Compare incentives with long-term adequacy and transition needs.
How markets fit the wider Energy series
Markets connect forecasting, flexibility, curtailment, resource adequacy and electricity-grid physics.
The deeper lesson is that electricity prices are not floating abstractions. They are economic signals generated inside a physical machine whose wires, generators, storage and demand must still obey physics every second.