Abstract:
In a deregulated competitive electricity market, Load Serving Entities (LSEs) and generation companies (GenCos) have to conduct their own portfolio optimization for different viable trading options before reaching any bilateral agreements in order to optimize their price risks and returns which affect their profitability. Portfolio optimization procedure assume that bilateral transactions are agreed by match making for power quantities at fixed energy prices. In an electricity market, market participants’ decision making includes match makings and bilateral negotiations which are two distinct phases of decision making process for bilateral transactions. Agent-based models are particularly suitable for bilateral transactions. A novel matchmaking algorithm has been developed and implemented in an open-source software named ‘agent-based model for electricity markets simulations (AMES)’. The matchmaking algorithm enables each LSE agent to undertake its own matchmakings separately, considering different time spans of a day, to find optimal trading allocations over a range of prices, before engaging in bilateral negotiations. In this algorithm, matchmaking is achieved by direct-search without any organized bulletin-board, broker or match-maker. Portfolio optimization based matchmaking, instead of random matchmaking, systematically explores available electricity trading options throughout the market. It also explores local and non-local bilateral trades as well as day-ahead auction. Proposed matchmaking algorithm is unique as it scans all trading options over the entire range of negotiable prices. Each LSE agent individually finds its matchmaking results depending on market price history, risk-aversion preferences and private profit-seeking goals. The matchmaking algorithm finds utility of each bilateral trading option and also explores utility variation over negotiable price set which helps a lot in offering prices during subsequent bilateral negotiations.
