An Analysis System Of Exploration While Drilling Based On Date Mining

Oil and gas exploration is a technology-intensive industry of high risk and high investment. Its major technologies include seismic survey, well logging, drilling, coring, well test, chemical assay and analysis etc. In each procedure of the exploration a huge quantity of precious data will be acquired. These exploration data will be classified and analyzed and combined with traditional experience and knowledge, and then identified by drilling to search for oil in place. This is the major exploration routine. As the exploration process develops continuously exploration researchers will accumulate more and more experience and knowledge for different exploration targets. Some of the experience and knowledge will upgrade into theories while most of them will be only stored in the brains of the exploration researchers or recorded in archives. So there is a saying in the oil exploration industry that oil is in the brain of exploration researchers. With the technological advance, the exploration data gathered tends to increase in huge amount. If they are combined closely with traditional exploration experience the exploration will be more effective. The old way of exploration which is heavily relying on the experience in the brain of the exploration researchers or on checking data from a pile of archive paper files can not meet the requirement of exploration of fast pace, high precision and high efficiency. Therefore it is necessary to study the exploration data mining technology, the management and application technology of exploration experience and knowledge, so to set up a specialist decision support system which can help the coordination of multi-subject and multi-specialty, provide the richest information for decision making and improve the effectiveness of production measures, lower the risk of exploration decision, evade risk of investment, and finally upgrade the exploration work. This paper will focus on the data mining in the drilling process and try to set up a drilling data analysis system.For the purpose of making effective application of information resources to push forward the exploration technologies, many international oil companies are active in using information technology to build exploration and exploitation knowledge database, develop specialists decision making support system so to improve the scientific management and decision making. Their methods mainly include construction of model database, method database, high level usage of present data and knowledge etc. Many inland large oil companies has also begun research on data mining such as the Daqing oil field who asked Landmark company to design its data management and application system, Shengli oilfield has set up its exploration data management and query system, exploration well location decision making support system, some companies has created initial knowledge management system etc. However the author has not witnessed any report on a system to make data mining and experience and knowledge management while drilling.In the paper, the artificial technology has been used to study key technologies supporting the oil and gas exploration decision making. Artificial intelligence is a branch of computer science which studies how to make machines intelligent. That is to use computer to simulate or realize the human intelligence which include technologies like data mining, neural network etc. specialist system is the knowledge processing system dealing with three basic questions: knowledge expression, knowledge utilization and knowledge acquirement which is the most mature area of artificial intelligence. As its nature is knowledge the specialist system is normally called knowledge based system.The target of the paper is to make analysis of the drilling decision process and use data mining technology to set up a support system for the decision work of specialists. First of all, the operation of deepening of exploration well and sidewall drilling decision of the drilling work will be studied. All the factors impacting the decision will be defined and their relations be made clear. Then the clustering algorithm will be studied to analyze and choose typical examples for reference. The storage and utilization method of theories, experiences, and knowledge will be studied for the following two aims. One is to inherit the successful traditional and right experience and lessons gained in the long history of operation of decision making which will be stored and managed in the knowledge database. The other is to make high level usage of these data and knowledge through setting up model database and method database. Based on the above, data mining technology will be used to anatomize existing examples to find out the impacting knowledge and logics on decision making. Finally a knowledge database for automatic analysis will to created and enriched for making new decision. The knowledge database and Inference engine will be used to reason and predict possible decisions so to prove or suggest on the decisions made by specialists.The paper has made on new approach of the mode of supporting decision while drilling. Its initial application shows some feasibility of the system which has laid good foundation for making data mining study and experience and knowledge management in the whole exploration process. It can also be referenced for future study of making automatic decision by computers to simulate specialists.