پیاده سازی انتشار تجارت مبتنی بر عامل برای کنترل انتشار نیروگاه مجازی

A methodology was developed and tested for controlling the emissions from a group of micro-generators aggregated in a Virtual Power Plant. The methodology is based on the EU Emissions Trading Scheme. A multi-agent system was designed and simulations were performed. The operation of the system was demonstrated experimentally using micro-generation sources installed in two laboratories. Two days of experiments were performed. Results show that system emissions have been controlled with a good accuracy, since only small deviations between desired and actual emissions output were observed. It was found that Virtual Power Plant controllability increases significantly by increasing the number of participating micro-generators.

Aggregation of Distributed Energy Resources (DER) has been considered as a promising solution for mitigating issues related with DER grid integration, such as network constraints, controllability, or resource intermittency [1], [2], [3] and [4]. Two aggregation concepts have been proposed [5]: (i) the micro-grids and (ii) the Virtual Power Plants (VPP). Emission Trading Schemes (ETS) have been adopted by several countries as a means to regulate carbon emissions, the most prominent of which is the European Union ETS [6]. This scheme allows the cost-efficient reduction of CO2 emissions among its participants. Allowances are issued by the regulator, in the form of transferrable Carbon Credits, representing one tonne of CO2 emissions [6]. The method of “cap and trade” is used, issuing fewer allowances than the participants actually need. Hence, the participants with the lowest cost of emissions reduction are given incentive to balance the emissions of the more costly participants. This results in a cost-efficient way of matching the total emissions to a desired value. The total emissions are controlled by the total amount of Carbon Credits supplied to the participants by the regulator (the European Commission). In this paper, a method for the control of carbon emissions induced by a group of aggregated micro-generators is presented. The purpose of this work was to test this methodology by performing simulations and an experimental demonstration of an agent-based implementation. This methodology is based on the EU ETS scheme, as presented in Section 2. A multi-agent system has been developed for implementing the methodology and is presented in Section 3. The methodology was demonstrated with a simulated and an experimental case study, presented in Sections 4 and 5, respectively. Results are presented in Section 6 and conclusions are given in Section 7.

A methodology for controlling the emissions of a Virtual Power Plant was presented. The methodology has been tested by means of simulations and experimental demonstration. The methodology is based on the EU Emissions Trading Scheme. Trading of Carbon Credits was used to balance the micro-generator emissions within the VPP. This method was termed Environmental Virtual Power Plant (EVPP). An agent-based control system was designed and developed. A hierarchical control structure was defined. Three types of intelligent agents have been used: (i) the EVPP aggregator agent, (ii) the micro-grid and (iii) the micro-generator. Simulations were performed, indicating high precision in the regulation of the micro-generator emissions. The EVPP methodology was also tested using equipment from two laboratories, installed in NTUA and CRES. Two experiments were performed. During both experiments, the EVPP was operated for approximately 8 h, using the mixed control policy (accounting for emissions and cost). In Experiment I, the EVPP included only the four sources installed in the two labs. In Experiment II, 55 additional sources were simulated. In the results from Experiment I, it was observed that the output of the EVPP was dominated by the Diesel engine, which was producing most (>80%) of the EVPP emissions. This disparity resulted in significant deviation of the EVPP output from the Carbon Credits that were supplied. In Experiment II, the EVPP included 59 sources. It was observed that the deviation from the Carbon Credits dropped significantly, compared to Experiment I. This was mostly due to two reasons: (a) The Diesel engine had access to much more micro-generation agents that could buy or sell Carbon Credits. Thus, incidents such as the deviation peaks in Experiment I were avoided. (b) Most of the sources (55 out of 59) were simulated, therefore lacking limitations of real systems, such as measurement errors, engine response delays, or generator start-up requirements. It was concluded that the controllability of the EVPP output primarily depends on the number of sources included in its portfolio. Individual micro-generation limitations are cancelled out as their number in the EVPP increases. Thus, the average EVPP output deviation can be expected to decrease, as the number of sources increases.