Early kick detection has been identified as being of primary importance in slimhole wellbores. Small annular volumes mean that, to maintain the integrity of the well, allowable kick volumes must be small. Gas influxes must therefore be detected and shut in rapidly. This paper describes an early kick-detection system developed for slimholes to detect and confirm the presence of an influx rapidly. This system has been run successfully on a number of slimhole operations.The early kick-detection (EKD) system is based on real-time analysis of drilling data obtained directly from a comprehensive mud-logging system on the rig. The analysis technique compares predictions of mud flow out and standpipe pressure from a dynamic wellbore model with corresponding values from the rig. The predicted values are derived from a model driven in real time by rig data such as pump rate and pipe rotation rate. Kick detection is based on deviations between measured data and idealized model predictions.The EKD system has been incorporated into an operational engineer-oriented graphical interface, which has provided easy access to the model for both input and output of data, and for the interpretation of results. This paper describes the design considerations and technology behind the EKD system and the engineering interface. The paper also presents examples of the system running in real time at a slimhole rig site.

Ttis PaW was selec!ea f., PrEse"la!ion by a" SPE P[ogmnl CVmmiUe@ following review 01 tn{onmtio" con!.ained 1" m abstract s"bminmi by Ihe aulhw(s). COn!e"!S of the m,%,, as presented, have not been r@wed @ w Society of Petroleum E.g(neers and are subJect.to .wmectlon by me wthw(,]. The rna!er,d, as prwe.ted, does noi pcesswily ref)acl my Fowon of the Socie$y of Petroleum En@ers, {% olflcars, m members, PWW8 pmemed at SPE meetings am s.biec! to wblicallo. wiEw by Ed(kxbal Committee% 01 me .Sadety of Petroleum Engineers. Pmliss!mto copy is restricted 10 an ab,tracl of ,0! mme than 3C+I.+, Illustrations may "01 be CCQled, The ,bsl'ad should ca"l.." CO"@CUO"S .Wk"owl.dgmen! of where and by whom the paper is presented. AbstractOptimum drilling conditions have to date been determined using either trial and error at the rigsife or by repetitive and time consuming calculations, generally involving a number of different PC based programs. Calculating optimum drilling fluid flowrate conditions to ensiure the highest possible penetration rate for example, involves manipulating variables such as drillstring configuration, nozzle diameters and 'rnud propeflies to satisfy constraints such as the ability to clean the hole, bit specific en"ergy, system pressure distribution and rig capability. This optimization" process is ideally suited to automation, however, the number of variables and constraints on the problem requires careful design of the user interface.