The representation made of the network in electricity markets must allow the consideration of main constraints related to the existence of the network that are shaping system operation. These can be of two main types: network congestion and losses. Implementing a network model within markets that is reflecting, accurately enough, network constraints should result in market prices corresponding in each point of the network to the local marginal supply cost of electricity. At the same time, simplifying to the extent possible the representation made of the network should provide some benefits, like the reduction of the complexity, an increase in transparency, of the energy dispatch; or the increase in the liquidity of markets.

In order to achieve this objective, a set of options for the representation of the network have been defined. A list of the follows:

* Nodal Pricing:
it involves the application of a separate short term marginal cost of supply in each node. All network constraints are considered.

* Zonal Pricing:
defining a set of zones that reflect congestion in the grid so that a uniform price of energy is applied in each zone. Currently, zonal pricing is applied in Europe, with most zones being coincident with countries. However, we are considering here a redefinition of zones so that these, as just mentioned, manage to appropriately reflect congestion.

* Hybrid Zonal Pricing:
this is a variant of the zonal pricing scheme whereby one or more zones are defined within each control area so that zones defines are compatible with political and administrative borders (also those borders defined traditionally in electrical systems). A example of this is the scheme applied in the Nordic countries.

* Single Node Dispatch:
this is the traditional networkless dispatch. Implementing it at European level would imply having a single price applied all over Europe.

*Average Zonal Pricing:
under this scheme, first, a fully-detailed representation of the network is considered when computing the energy dispatch. Then, the price applied to non-flexible load within each of a set of zones defined (the same as for a zonal pricing scheme) is computed as the average of the nodal prices in this zone. Flexible Load and Generation is earning a price equal to their bid in the market (pay as bid pricing).

As already mentioned, defining price zones that reflect active constraints effectively conditioning the dispatch is necessary to produce an efficient and feasible dispatch.

The pattern of congestion and losses in the system, together with the topology of the grid, may affect the viability of the definition of price zones, i.e. whether it is possible to define a reduced number of areas so that congestion in the system, and losses, limit transfers of power taking place among these zones, but not transfers of power occurring within each zone.

As mentioned above, the computation of efficient prices out of the application of a network model that reflects system congestion must be made compatible with achieving a sufficient level of liquidity in markets that limits to the extent possible, the exercise of market power by local generators.

Bidding zones may need to be updated with a change in the topology of the network (construction of new lines) or changes in the distribution and operation profile of demand and generation in the system. This may be necessary in order to have, at all times, zones that are reflective of local marginal supply costs.

Discrimination is another aspect to take into account when assessing options for network models. Some systems have claimed that applying different prices to different consumers within each country, or local system, especially if these are households, is unfair discrimination that should be avoided. The same might also be applied to generation (different generators within a country earning different prices based on their location is unfair).