Viscous and heavy oil reservoir is a challenge in oil production due to unfavorable mobility ratio. Thermal injection is a method to increase well productivity and oil recovery which is usually performs in shallow reservoir. One type of thermal injection is cyclic steam stimulation. Cyclic steam stimulation was conducted in two wells of Melibur field as a pilot project. Since it has 8–12 cP viscosity, 22–26° API oil gravity and relatively shallow reservoir depth, Melibur has an appropriate character to perform steam injection. The project started with well selection with the main consideration is amount of remaining oil in reservoir. It is also considering completion diagram and operational aspect for each well. The injection process was performed in 10–12 days with certain injection parameter to meet the heat requirement for reservoir and follow with 5 days soaking. This paper focuses on the result and effect of cyclic steam stimulation to well and offset wells production rate and fluid properties. Many experiences acquired from the project of cyclic steam stimulation perform in Sihapas formation, one of them is the effect to offset well that indicates there is a connection and high heat conductivity between wells. Incremental of initial production rate about 40% occurred in first well. In second well, this operation gives an effect to offset well with the incremental of production rate reach 100% in nearest well. Oil properties changes with different in viscosity, oil gravity and pour point value after cyclic steam stimulation.

Permeability estimation for un-cored wells is a classic issue. A simple model that widely used is using core porosity-core permeability cross plot to determine the linier regression. Then we estimate permeability in un-cored well after making adjustment for porosity log to porosity core. The difficulties using that method is most of cross plot did not show clear relationship (scatter data) due to effect of rock heterogeneity. Therefore another effort is needed by separate it based on rock type to get better relationship. Permeability itself is not depend only on porosity but also other properties like clay content, grain size, tortuosity and etc. Part of this phenomenon had been modeled by Carmen-Kozeny which illustrates strong dependency of permeability on average grain size, tortuosity and flow zone index. The conventional way to reduce data scatter is by using additional correlation parameters. Commonly shale content (Vsh) and reservoir facies are used to give reliable transform or regression analysis for estimating the permeability. Start with this concept; we try to simplify the correlation by modified the Carmen Kozeny eq. (used flow zone index term) using clay content as another parameter that influences permeability value. Because we assume that porosity and clay content are the most important properties that have significant effect on permeability. In this paper, we will describe permeability estimation for un-cored well as function of porosity and clay content using modified flow zone index-permeability cross plot. This cross plot has been test in three clastic reservoirs in Lower Sihapas formation either for consolidated or unconsolidated sandstone. The result shows this cross plot give better relationship compare to conventional cross plot and more simple transform to estimate permeability in un-cored well for input to geologic and reservoir simulation models. Introduction EMP Malacca Strait has develop reservoir integrated study which involves data review, G & G remodeling, reservoir characterization, reservoir modeling and prediction to identify the reliable hydrocarbon potential and developing a reliable reservoir model for choosing the optimum development plan for Lower Sihapas formation in three clastic reservoir1,2. One of the main subjects during the integrated study workflow is the reservoir rock characterization and core data analysis which includes the permeability transform to develop correlation between core data and log data. Permeability is one of the fundamental rock properties which represent the quality of a reservoir. The appropriate permeability value in each well is needed in order to represent the permeability distribution in a reservoir model. Usually the data that we have from cored well is very limited. Beside that, in order to get the data from each well we need to perform coring process that quite time consuming and require expensive laboratory measurements. For that reason, in term of practical use, we need a mathematical function that can represent the permeability value from each key well (cored well) by using the existing data (log data).