Effect of Relative Permeability on History Matching a Permian Basin Oil Well
Master thesis
Permanent lenke
http://hdl.handle.net/11250/2351064Utgivelsesdato
2015Metadata
Vis full innførselSamlinger
Sammendrag
Production from shale oil and gas has increased significantly the last couple of decades. The industry is constantly developing horizontal well and stimulation technology to increase production from these wells. Data acquisition is fairly well established and reliable in conventional reservoirs. This, however, is not the case in unconventional shale reservoirs. In shales the data is usually limited to drilling logs, temperature and pressure measurements and fluid samples. Even the fluid samples can be non-representative. In fact, what you produce at the surface might not the same as what is present in the reservoir [Whitson and Sunjerga, 2012]. Reservoir parameters such as: Matrix permeability, matrix porosity, initial GOR and initial saturations is history matched to best fit observed production rates and pressures. It is important to input the known, tangible data as detailed as possible to reduce error in the well model. Operators of wells in the United States are obliged to report drilling reports and monthly production rate data to state regulatories. This thesis has compared data acquistion from a subscription based online database, IHS Enerdeq Browser, with data acquired directly from the operator of the well for a well in the Bone Spring Formation, in the Permian Basin. The public available data for this well has proven to be sufficient in order to build a well model. Some reservoir characteristics must be estimated based on neighboring wells, such as: reservoir height, reservoir temperature and reservoir pressure gradient. Simulating a well controlled on monthly average rates instead of daily rates has proven to be sufficient after a couple ofmonths of production. Relative permeability models are based on simple, analytical relationships in unconventionalshale well modeling. The only input, saturation endpoints and curvature, are estimates based on conventional rocks. The curves are then calculated based on Corey Power Law in SENSOR. This thesis investigated effects of relative permeability on the behavior of the gas-oil ratio, history matching and production forecasting on an oil well located in the Permian Basin, refered to as RHW8. One additional well in the same area has been history matched and production forecasted, refered to as RHW22. The analysis show that GOR is clearly affected by relative permeability of oil and gas. As the oil relative permeability decreases, the producing GOR is increases. When the flowing bottomhole pressure is lowered, a big increase followed by a long transient period (more than a year) is seen on the producing GOR. This proves the importance of numerical modeling of these type of wells, because conventional decline curve analysis will not capture these effects. By history matching the well on relative permeability of oil to gas and matrix permeability it has been shown that three different relative permeability models, nog Æ 3.5, 4.5 and 5.5, yield equally good history matches for both wells. The production forecast is affected by the relative permeability. Lower relative permeability of oil leads to a higher producing GOR, i.e. lower cumulative oil recovery. However, the effect of this is not significant until after ten years of production. For the RHW8 well the optimal constant flowing bottomhole pressure, in terms of oil production, from time = 0 is 2000 psia for the first 10 years of production. The highest cumulative oil production after 3.5 years is obtained by gradually lowering the FBHP.