Os métodos visuais para a detecção de defeitos na madeira, ainda que de grande utilidade, podem ser falhos e exigem a utilização de mão-de-obra relativamente bem treinada. Muitas vezes, a presença de nós, rachaduras ou, ainda, regiões com medula no interior de uma peça serrada, podem ser imperceptíveis na avaliação visual. Um dos grandes avanços obtidos nos últimos anos na caracterização de materiais, tanto do ponto de vista mecânico quanto da qualidade, é a aplicação de técnicas que utilizam a propagação de ondas, destacando-se, dentre elas, o ultra-som. O objetivo deste trabalho é avaliar, estatisticamente, a possibilidade de se utilizar o método do ultra-som na detecção de defeitos em peças de madeira serrada. Para isto, foram utilizadas 180 peças de dimensões nominais: 0,027 m de espessura; 0,10 m de largura e 0,25 m de comprimento, retiradas de vigas de Pinus sp obtidas em serraria da cidade de Campinas. As peças foram ensaiadas na umidade de equilíbrio ao ar. Para a realização dos ensaios utilizou-se o equipamento de emissão de ondas de ultra-som marca Steinkamp BP-7 com transdutores de 45 kHz. Inicialmente foi realizada uma análise visual das peças e, posteriormente, a determinação da velocidade de propagação das ondas nas mesmas peças. Com os resultados, realizou-se uma análise exploratória das variáveis e obteve-se um modelo de regressão logística visando verificar a relação entre a presença ou não de defeitos e a velocidade de propagação da onda do ultra-som na madeira. Os resultados demonstram uma forte relação entre a velocidade de propagação da onda de ultra-som e os defeitos detectados pela análise visual

The Congro field (Campos Basin, Brazil) contains considerable reserves located in a extremely heterogeneous, very low-permeability carbonate reservoir. For many years after its discovery, this reservoir was considered non-economic. However, a newly drilled horizontal well showed encouraging results. An advanced, integrated reservoir study showed that the exploitation of the reservoir can be economically attractive if non-conventional wells are used. In this work, we demonstrate how cutting-edge technology plus interdisciplinary efforts involving geophysics, geology, and reservoir engineering were the key to make a noneconomic asset into an economic one. Two major problems were faced: the short available time (only two months for the whole study) and the lack of data. The lack of data was overcome by using analogy with a similar, well-sampled reservoir. The time constraint was handled by taking an integrated approach where the geophysical, geological, and numerical simulation models were built almost simultaneously. Stochastic reservoir models were generated using a geostatistical method called truncated Gaussian technique to assess uncertainty on facies distribution. Three different types of wells were considered: vertical fractured, horizontal, and multilateral. As multilateral wells showed better economics, we optimized the number of legs and their length based on numerical simulation. Introduction Technology has more and more made the difference between success and failure of hydrocarbon exploitation projects, especially when the asset is on the economic borderline. New drilling techniques, as well as reservoir description techniques, have allowed the development of fields that had been considered economically unattractive in the past. Throughout the world, many projects have been reviewed and had their economic indicators (net present value, payout, etc.) significantly improved when new technologies were applied to them. Multilateral drilling and seismic interpretation are among the techniques that suffered major breakthroughs in the last few years. The first allows one to greatly improve well productivity index as well as enhance sweep efficiency; furthermore, the use of multilateral wells reduces drilling and equipment costs such as flowlines and christmas trees. The second can give a pretty good idea of where the best-quality rock facies are located and help identify reserves that may not be produced optimally. A third technique that can add value to a project is geostatistics. Multiple equiprobable reservoir models can be generated to assess uncertainty on reservoir description and on reservoir production forecast. Geostatistical models are generally more detailed than conventional mapping techniques, allowing a better description of the reservoir heterogeneity. Another advantage of geostatistics is that it can integrate different types of data independently measured at different scales. In this work, we show how the techniques mentioned above were applied in an integrated way to make the carbonate reservoir of Congro field economically viable. Team integration (geophysicist, geologist, geostatistician, reservoir engineer, drilling engineer, economist, etc.) was another key element to the success of this project. Pressured by time, the team did not take the conventional approach of taking one step at a time to build the reservoir model. That is, to build the geophysical model first, then the geological model, and, finally, build the numerical simulation model. Instead, the geophysicist, the geologist, and the reservoir engineer worked together so that their models were built almost simultaneously. This approach saved time and ensured that the models were coherent with all available data.