Grigoris A. Dourbois

European market coupling algorithm incorporating clearing conditions of block and complex orders

In this paper an algorithm for the solution of the European electricity market coupling is presented, considering all block and complex orders available in the European Power Exchanges. The model takes into account the clearing conditions of profile and regular block orders, linked block orders, exclusive group of block orders and flexible hourly orders, as well as the clearing conditions of Minimum Income Condition and Load Gradient orders, possibly under a scheduled stop condition. The model considers also hourly flow ramping constraints on single interconnections or group of interconnections, net position ramping constraints, interconnection losses and tariffs. The flow-based approach is implemented, using the zonal PTDF matrix. The algorithm eliminates possible paradoxically accepted block and MIC orders within an iterative process. The proposed algorithm is evaluated in a pan-European day-ahead electricity marketplace.

European-Type Electricity Market Clearing Model Incorporating PUN Orders

A novel model for the clearing of European day-ahead electricity markets, with all associated types of orders, namely simple hourly orders, block orders, complex orders and PUN (“Prezzo Unico Nazionale”) orders is presented in this paper. An iterative algorithm is employed for the market clearing, bearing an inner iteration process for the handling of Paradoxically Accepted block and MIC orders, and an outer iteration process for the handling of PUN orders. During the iterative process a Master Problem and a PUN subproblem are sequentially solved and coordinated, for the efficient handling of all order clearing conditions and market and/or system constraints. The clearing conditions of each order type are explicitly incorporated in the PUN subproblem, formulating a Mixed Complementarity Problem bearing both primal and dual decision variables. The PUN subproblem concerns the sequential clearing of 24 hourly subproblems, employing a special technique for handling the possible inefficiencies due to the myopic nature of the hourly subproblem solutions, leading to enhanced behavior in terms of social surplus with respect to the state-of-the-art day-ahead market clearing solver. The model and the algorithmic process are evaluated in a west-European test case, demonstrating minimal computational requirements and proving its excellent escalation capabilities.

European Market Integration With Mixed Network Representation Schemes

A centralized market-splitting algorithm is implemented in this paper in a Europe-wide level, comprising both power pools and Power Exchanges, with each local/national market respecting the standard constraints imposed by its own regulatory framework, including the full set of unit technical/commitment constraints and system operating constraints in power pools. In view of the forthcoming large-scale RES penetration, physical markets with unit-based offers (either pools or PXs) check the feasibility of the electricity market solution against their internal (intra-zonal) transmission network constraints, considering full network topology. This is accomplished through an iterative process, iterating between the overall optimization algorithm, and intra-zonal power flows of the countries/regions, which identify possible congestions and incorporate additional constraints in the central pan-European market-splitting problem. The iterative process terminates when all internal transmission constraints are satisfied. Locational marginal prices (LMPs) can be computed in the physical markets, whereas zonal/system marginal prices are computed in markets with portfolio-bidding schemes. The proposed algorithm is tested in terms of computational efficiency using the full UCTE network.

European market integration with both physical and non-physical markets

In view of the forthcoming large-scale RES penetration, physical markets should check the feasibility of the electricity market clearing against their internal (intra-zonal) transmission network constraints, considering full network topology. Three methods for the solution of a market-splitting problem in a Europe-wide level are implemented in this paper. Two iterative processes are employed, iterating between the overall optimization algorithm and intra-zonal power flows of the countries/regions that identify possible internal congestions; the iterative processes terminate when all internal transmission constraints are satisfied. The results of the iterative processes are compared with the results of a one-stage solution of the overall problem, which incorporates the full set of transmission constraints of all European electricity markets. The proposed algorithms are also compared in terms of computational efficiency using the full UCTE network.

European Power Exchange day-ahead market clearing with Benders Decomposition

A European Power Exchange day-ahead market with both simple and combinatorial products is modeled in this paper and cleared using Benders Decomposition. Except from simple hourly supply offers and demand bids, the products comprise supply and demand profile block offers/bids, and supply and demand linked profile block offers/bids. The problem constitutes a Mixed Integer Linear Programming model having a decomposable structure, which is decomposed in terms of binary and continuous variables, and solved using the Benders Decomposition approach. The master problem includes the binary variables, the decisions of which are transferred to the sub-problem, which incorporates the problem continuous variables. The iterations of Benders method constitute inner iterations; outer iterations are also performed in order to handle the paradoxically accepted and rejected blocks. The overall algorithm converges to the solution of the original MILP problem. The proposed approach is evaluated using the IEEE RTS-96 three-area test system.

A novel method for the clearing of a day-ahead electricity market with mixed pricing rules

A novel approach for the clearing of a European-type day-ahead electricity market is presented in this paper. All market products that are currently tradable in European day-ahead electricity markets, namely simple hourly orders, block orders, complex orders and PUN (“Prezzo Unico Nazionale”) orders are incorporated in the model. A one-stage approach is formulated for the efficient handling of the market orders clearing conditions, the non-intuitive bilateral exchanges and the transmission constraints. The clearing conditions of block orders, complex orders and PUN orders are explicitly incorporated in the problem formulating a Mixed Complementarity Problem. The algorithm eliminates possible paradoxically accepted block and non-intuitive bilateral exchanges within an iterative process. The proposed algorithm is evaluated in a simple 14-zone system in terms of solution efficiency and computational requirements.

A new concept for the clearing of European Power Exchange day-ahead markets with complex orders

A new concept for the clearing of the day-ahead market of a European Power Exchange is proposed in this paper, bypassing the problems caused by the indivisibility of complex orders present in the European markets. The presence of complex orders may lead to paradoxically accepted and/or paradoxically rejected orders, necessitating the use of iterative heuristic procedures and empirical simplifying criteria for their handling. In this paper, the concept of “adjustable orders” is introduced, leading to a significant simplification in the handling of complex orders. The adjustable orders can be partially cleared, leading to orders that can be marginally accepted in the market clearing, setting the market prices accordingly for the hours involved in the orders. The complex orders that are used in this paper comprise profile block orders (offers/bids), flexible hourly orders and joint demand profile orders. The modeling of adjustable blocks is not possible as a MILP problem due to the inherent indivisibility denoted by the use of binary variables; instead, the day-ahead market clearing is formulated as a Mixed Complementarity Problem. The solution of the proposed MCP model is compared with the solution of the classical MILP model solution within an iterative procedure (employed in CWE and Nordpool) for the handling of paradoxically accepted/rejected orders.