Abstract
This paper presents an original forecasting methodology for achieving the spatiotemporal future long-term expansion of small power photovoltaic (PV) systems in a region, taking into account the population density, ground usage and the type of small PV power application adopted. This methodology comprises three stages: a first stage based on a suitable PV technological forecasting method with a group of experts; a second stage consisting of an innovative and iterative process based on elimination of the possible numerical inconsistencies achieved in the first stage; a third stage with a new method for achieving PV power density maps, using a geographical information system (GIS), that provides significant quantitative GIS information and visual and geographically-disaggregated representation of future small power PV systems expansion. The proposed methodology is illustrated with a real example for the region of La Rioja, Spain. In this example, four different combinations of PV systems and geographical zones were considered, and they are referred to as four "PV technologies" in the paper. The forecasted period range was 20 years with steps of 5 years. The results offer very valuable information for electric utilities, PV systems sales and installation agents, investors and regional authorities responsible for energy plans.

Abstract
This paper discusses a novel load management strategy for the optimal use of renewable energy in systems with wind turbines, a battery bank, and a diesel generator. Using predictions concerning wind speed and power, controllable loads are used to minimize the energy supplied by the diesel generator and battery bank, subject to constraints imposed by the user's behavior and duty cycle of the appliances. We analyzed a small hybrid power system in Zaragoza, Spain, and the results showed load management strategy allowed improvement in the wind power use by shifting controllable loads to wind power peaks, increasing the state of the charge in the battery bank, and reducing the diesel generator operating time, when compared to a case without load management.

Abstract
This paper presents a comparison of two new advanced statistical short-term wind-power forecasting systems developed by two independent research teams. The input variables used in both systems were the same: forecasted meteorological variable values obtained from a numerical weather prediction model: and electric power-generation registers from the SCADA system of the wind farm. Both systems are described in detail and the forecasting results compared, revealing great similarities, although the proposed structures of the two systems are different. The forecast horizon for both systems is 72 h, allowing the use of the forecasted values in electric market operations, as diary and intra-diary power generation bid offers, and in wind-farm maintenance planning.

Abstract
The intermittence of the renewable sources due to its unpredictability increases the instability of the actual grid and energy supply. Besides, in a deregulated and competitive framework, producers and consumers require short-term forecasting tools to derive their bidding strategies to the electricity market. This paper proposes a novel hybrid computational tool, based on a combination of evolutionary particle swarm optimization with an adaptive-network-based fuzzy inference system, for wind power forecasting and electricity prices forecasting in the short-term. The results from two real-world case studies are presented, in order to illustrate the proficiency of the proposed computational tool.

Abstract
Wind energy experiences in Portugal an increasing interest. Slightly more than 1700 MW were operating by the end of 2006, in a system with a global capacity of about 12 GW (8,5 GW peak demand). Several new wind farms are under construction and a considerable amount of connection points are or will be granted in the coming years. More than 5000 MW are expected to be connected to the grid around 2012, the global generating capacity being then about 16 GW. Clearly, a wind power forecasting system must be implemented that will help to deal with the significant penetration of the technology in the electrical system. A group of wind farm promoters, owning the majority of the capacity installed so far, ordered to a consortium of universities and research institutes the development of a forecasting tool, giving rise of the EPREV project, wholly financed by them. The system will have the following main characteristics: Wind speed and active power forecasting up to 72 hours; Evaluation of the forecasting uncertainty; Possibility of using the predictions of physical models and the information from the wind farm Supervisory Control And Data Acquisition (SCADA); Capacity of predicting only with SCADA information for very short term. The main components of the system are: A human-machine-interface, allowing the control of the system, the selection and aggregation of forecasting models and the visualization of results; A power forecasting model for individual wind turbines and for wind farms. A cascade of models is used, starting in the mesoscale simulation, with up to 2 km resolution. The outputs of the mesoscale models are corrected and statistically adapted to the fine scale conditions. Two models and different boundary conditions are run, in three nested domains (54x54, 18x18 and 6x6 km). The advantage of using a 2x2 km resolution is also tested. The statistical models are fed with recent information from the wind farms, after a learning process that made use of the historical information of its operation. Three different types of statistical models are employed: Power Curve Model (PCM), Auto Regressive (AR) and Neural Network Assembling Model (NNAM). The wind simulation at the wind farm scale is done both by linearized physical models and Computational Fluid Dynamics (CFD) models, namely using VENTOS®, a code developed at the University of Porto. The duration of the project is planned to be 1 year, including off-line tests of the complete system for 3 wind farms, for performance evaluation purposes.

Abstract
The use of wind energy has developed significantly worldwide. Wind power is the strongest growing form of renewable energy, ideal for a future with pollution-free electric power. But the integration of wind farms in power networks has become an important problem for the unit commitment and control of power plants in electric power systems. The intermittent nature of wind makes it difficult to forecast wind-produced electric energy in a wind farm even in the next hours. This paper compares the results obtained with a set of selected models for hourly electric power production forecasting in a real-life wind farm. The results show a significant improvement if previous numerical weather forecasts are used as input in hourly power forecasting models.