Methods of Using Logs to Quantify Drillability

Abstract

Abstract Correlations between sonic logs and the formation drillability for different lithology types have been developed from data taken from 10 wells in North America. The gamma ray log was used in conjunction with drilling data to calculate the drillability. The drillability from penetration rate models is back calculated from bit design and reported field wear in conjunction with meter by meter operating parameters, formation type and pore pressure. Then this drillability was correlated with sonic logs for different lithologies as defined by the gamma ray log. The different formation types clearly show different correlations for the normalized correlations between drillability and sonic logs. The non-homogeneous lithologies are also correlated and normalized to rock strength from the sonic logs where the percent formation type mixtures are determined from the gamma ray. Data from multiple wells is presented showing the accuracy of the presented approach where more then 100,000 data points were statistically analyzed and evaluated in the development of the equations presented herein. The drillability from inverted penetration rate models has been verified to give good representation of rock strength based on comparison with triaxial laboratory data and makes the use of this model more versatile. The correlations provide improved estimations of rock strength which can be used in drilling performance simulation and wellbore stability studies. Introduction The optimization of the drilling process by the use of a drilling simulator can significantly improve results with the input of proper rock properties. The Unconfined Compressive Strength (UCS) of the rock is one of the most important properties and is also the most difficult to obtain. The ultimate objective of well logging is to evaluate subsurface formations by providing an indirect measurement of fluid and rock characteristics. Well logging has been known to provide information on rock strength but the usage of electric logs for calculating drillability is less common. Correlations of rock strength with logs can be obtained from laboratory uniaxial tests where the stress at shear fracture in a core specimen is correated to the value of the log properties at the core sample depth. The disadvantage with the proposed procedure is that lab data is available only for the area where the well was cored and does not account for insitu strength. This gives less accurate representation of the entire well. The "Optimizer" is also capable of estimating unconfined rock strength as apparent rock strength (ARS). The current process of estimating ARS from the drilling simulator involves the use of a reference well that closely matches the characteristics of the planned well. The drilling inputs (WOB, RPM, Flow Rates, etc.) and outputs (Bit Wear, ROP, etc.) of the reference well are used to generate a drillability log or an Apparent Rock Strength Log (ARSL) for the planned well. The reliability of this procedure is dependent on the quality of the field drilling data recorded. Although this technique has proved to give accurate predictions with successful results, refinements can be made to provide greater confidence in the predictions by using log data as a second source for calculating rock strength. At times the drilling data are not recorded with great accuracy and thus adversely affects the prediction of the ARSL. Also problems encountered while drilling such as down-hole and stick-slip vibration will have an effect on the quality of the prediction. Background Apparent Rock Strength Log (ARSL)2–4 The ARSL is a strength log generated by the "Optimizer". The ROP and field drilling data are used to predict apparent rock strength under actual drilling conditions. The apparent rock strength found from the Optimizer is a function of operational parameters, bit properties, lithology, pore pressure, and rate of penetration.