|
eAuction
eAuction is an environment for simulating auction designs (e.g. descending price clock auction, pay as bid) for auctions of electricity supply contracts and considering different bidder risk profiles and coalitions.
Modeling aspects
The core of the environment is an agent-based simulation model that contains software modules implementing a wide range of bidding strategies, from simple heuristic rules to stochastic optimization models. The auction simulation can also be run with “human” bidders or any combination of these two options. The system is implemented in a network (LAN or Internet) composed of bidding agents (representing each of the auction sellers) and an “auctioneer”, which implemented the auction rules. The auction simulation may take place in real time or in “fully automated” mode (Repeated auction simulations where all bidders are software modules, with random sampling of bidding strategies or strategy parameters in each simulation). eAuction is has a Windows GUI interface and outputs are easily managed by Excel and other spreadsheet editors.
Some recent applications
The agent-based market simulator was developed by PSR by request of the Ministry of Energy and by the Regulator, with the purpose of simulating bidding strategies and market dynamics of the first contract auction under the power sector framework. This first auction was carried out in December 2004, for a required capacity of 17 thousand MW (average energy, not peak). Given that this was by far the largest electricity contract auction carried out anywhere, involving tens of billions of US$ in contracts, there was a special concern about auction rules and strategic behavior by some of the largest bidders (about twenty candidates were expected, but two of them controlled more than 20% of the country’s generation capacity). The eAuction environment was extensively used to detect undesirable behaviors (market power) and to simulate the effect of different auction rules on final prices.
After the auction rules were defined and published, the simulator was commercialized to the majority of bidders in the auction (80% of total supply). Since then, the model has been actively used for the simulation of other energy supply contract auctions (forwards and call options) in Brazil.
Tela 1.5.2 |
|
|
