Abstract: Based on the data of 244 wells, the sandstone reservoirs of the Neogene in Huatugou oilfield, Qaidam Basin is tentatively evaluated with Q-model Principal Factor Analysis(PFA) and Fast Clustering Method(FCM). In order to quantitatively analyze the characteristics of the reservoirs and the relation between the characteristics and the sedimentary facies of the reservoirs with the evaluation results, the method is put forward as follows: firstly, the scores of the principal factors obtained by PFA are normalized; then, the normalized scores are multiplied with the weights of the corresponding principal factors and added to get the scores of the samples. The reservoirs are evaluated by applied FCM with the scores of the samples. After analyzing the characteristic of the sandstone reservoir in the area with the statistic of the evaluation results, the authors conclude that, for the characteristic of the reservoir, the lower Youshashan Formation of the Neogene in Huatugou oilfield is better than the upper Ganchaigou Formation, the braided channel deltaic front is the best sub-type facies, the river-mouth bar sand body is the best microfacies sand body, and the submarine distributary sand body and turbidite sand body follow. Authors' method is quantitative and contains the explicit geological significance, so it is worthy of being generalized.

Key words: reservoir evaluation, Q-model principal factor analysis(PFA), fast clustering method(FCM), Neogene, Qaidam Basin

摘要： 利用244口井的小层数据，尝试用Q型主因子分析与聚类分析相结合的方法对柴达木 盆地花土沟油田新近系砂岩储层进行评价。为使评价结果定量化，以便利用评价结果深入分 析储层储集性能及与沉积相的关系，提出如下处理方法：首先对由Q型主因子分析得到的各 主因子得分进行归一化处理；然后用归一化得分与对应主因子的方差贡献率相乘后相加，得 到样本总得分。根据样本总得分，运用聚类分析，进行储层分级。对储层评价结果进行统计 分析后认为，就储层储集性能而言，花土沟油田新近系下油砂山组明显好于上干柴沟组；辫 状河三角洲前缘是最好的沉积亚相；河口坝是最好的沉积微相，其次为浊积砂与水下河道。 研究表明，本文提出的储层评价方法具有定量化、地质意义明确等优点，有一定应用价值。

关键词: 储层评价, Q型主因子分析, 聚类分析, 新近系, 柴达木盆地