Negotiation Aid System For Promotion of Distributed Generation and Renewables Vladimiro Miranda, Cláudio Monteiro and I.J.Ramirez-Rosado Abstract – This paper describes a new concept of a Negotiation Aid System, developed over a GIS (Geographic Information System) and designed to facilitate reaching compromises among agents such as investors, environmental groups and governmen- tal agencies, when deciding the location and sizing of new re- newable energy sources in a region. The core model of an Actor is similar to a Fuzzy Inference System of the Takagi-Sugeno type, built from a definition of preferences and levels of accept- ability. An outranking method is employed to define geographi- cal places of less conflict among the several Actors negotiating. An application to the region of La Rioja, in Spain, is described. Index Terms: New Renewable Energies, Decision Aid, Nego- tiation Aid, Fuzzy Sets, Geographical Information Systems I. INTRODUCTION he development of distributed generation has been condi- tioned in the European Union by constraints that limit the success of governamental incentives, aimed at increasing the penetration of renewables in the EU power generation portfo- lio, having in mind the Kyoto protocol. These incentives have several forms, from subsidies to the construction of new gen- eration facilities to guaranteed “green” prices and tariffs for energy from renewables including avoided costs for CO 2 em- missions. Therefore, investors put a heavy pressure in search- ing for places where to build new renewable plants and where to connect them to the grid with the purpose of selling energy. One source of constraints to the widespread use of wind generation derives from supra-national, national or local regu- lations, creating protected zones, natural or national parks, areas of protected bio-diversity or ecologically protected, besides zones close to buildings, airports, etc. or related with the military or possibility of radar confusion. Another source of constraints derives from the opposition of environmental organizations, who locally object to renew- ables while at national level continue to claim for their use. Therefore, Government Energy Directorates or Energy Agen- cies need to have at hand a comprehensive methodology to try to conciliate the interests of different agents (investors, envi- ronmentalists, state agents) in order to organize regional plans for the development of distributed renewable generation [1]. This paper presents one Negotiation Aid platform for such purpose [2][3][4]. It is built over a GIS – Geographical Informa- tion System[5], and it allows the identification, over a map of a region, of the zones or areas where the conflicting interests of several actors may best be conciliated. The model goes through two steps: a decision phase for each Actor in the negotiation process and a conciliation phase among all the Actors; the process may iterate, guided by a broker agent and, given certain circumstances, it may even use techniques of training and automatic learning to capture the essence of the preferences of each Actor. The core of the model is a decision rule for each Actor in the form of a Takagi-Sugeno Fuzzy Infer- ence System, which provides as output an aceptance or toler- ance index about the location of wind generation at each cell in a GIS map. The work described has been applied in a joint project de- veloped for the region of La Rioja, Spain, by INESC Porto and the University of La Rioja, and some results are displayed in the paper, under the form of maps and graphs. The model is a revised and corrected version of the one described in a previ- ous publication [6]. II. BUILDING THE CONCEPT A Negotiation Aid System (NAS) is different from a Deci- sion Aid System (DAS) in the following basic aspects: • a NAS targets a set of agents or Actors, instead of a sin- gle Decision Maker, as in the case of a DAS • in a NAS, each Actor defines his own interests and objec- tives and evaluates alternatives based on attributes and criteria, independently of the attitude of the other agents • the criteria adopted by an Actor are not necessarily the same nor in the same number as the criteria adopted by any other agent. This means that one cannot directly transport for a NAS environment the techniques usually employed in DAS where a single Decision Maker is present. For instance, if the actors do not share the same criteria, it is not possible to define a set of alternatives with the property of Pareto Optimality. To build a NAS, we have therefore proceeded to take two successive steps: 1) to develop a model for each Actor; and 2) to apply a model for the interaction among actors. T ______________________________ V. Miranda is with INESC Porto – Institute of Engineering in Sy s- tems and Computers at Porto, Portugal, and with FEUP – Faculty of Engineering of the University of Porto, Portugal (vmiranda@inescporto.pt) N. Fonseca is with the Power Systems Unit of INESC Porto, Portu- gal (email: nfonseca@power.inescn.pt) I.J.Ramirez-Rosado is with the University of La Rioja, Spain (email: ignacio.ramirez@die.unirioja.es)