![]()
Summary. More than 1,000 directional surveys from wells in the Prudhoe Bayfield are used to examine differences in well location at 8,000 ft [2438 m]subsea, as measured by gyroscopic and magnetic surveying instruments. Thelateral, radial, and vertical errors are plotted as a function ofhorizontal offset. Mean lateral differences appear to be about 100 ft[30m]. Drillstring magnetization causes magnetic surveys to read north ofgyros. Vertical and radial differences are negatively correlated, are chiefly a result of inclinometer errors, and are on the order of tens of feet. The Wolff-de Wardt (WdW) error model, with small modifications, does good job of quantifying the error magnitudes, of describing thedependence of error upon horizontal offset, and of producing the ellipsesof uncertainty.
Introduction Walstrom et al, proposed a model to predict theuncertainty of the bottomhole position obtained from agiven survey. Their main assumption was that the errorsmade on the measurement of a given parameter (measured depth, azimuth, or inclination) were random from onestation to another. Walstrom et al. introduced theellipse-of-uncertainty concept to describe the boreholeposition. Truex described some discrepancies observed on surveys calculated on the East Wilmington field. Truex foundthat surveys in high-inclination wells (above 45 degrees[0.79 rad]) had a possible error of 20 ft [6 m] verticallyand 100 ft [30 m] horizontally for a well of 6,000-ft [1830-m] measured depth. These errors were higher thanthose expected from the Walstrom model. Wolff and de Wardt showed that the errors were notrandom from station to station in a given survey, but ratherwere systematic from one station to the next and randomfrom one survey to another. The consequence of thisassumption is that the errors do not compensate each otherfrom one station to the other, but tend to stack in the same direction, producing a larger ellipse of uncertainty thanthat calculated with the Walstrom model. Most wells in the Prudhoe Bay field have been directionally surveyed several times. The surveys are goingto be compared on a well-by-well basis, which is anapproach similar to Truex's work. Truex, however, emphasizedthe extreme discrepancies between surveys. Here both largeand small discrepancies are treated, facilitating thecomparison between real data and the WdW model.
Directional Surveys at Prudhoe Bay Drilling and surveying of production wells have beenunder way at Prudhoe Bay since 1969. Wells are drilled2,000 ft [610 m] through permafrost(Fig. 1), deviatedfrom vertical at some distance below the permafrost, anddrilled to a total subsea depth of around 9,100 ft [2775m]. Fig. 2 shows the histogram of the depths of the kickoffpoints for 675 Prudhoe Bay wells. The kickoff point isdefined as the point at which the horizontal offset of thewells becomes greater than 100 ft [30 m]. To guide the drilling, magnetic single-shot surveys, S, are run. Since 1982, measurement-while-drilling(MWD)instruments that incorporate fluxgate magnetometers havesupplanted the single shots. Single shots are used to checkthe MWD results, however, and the resultant surveys, X, are almost always mixes of MWD and single-shot data. MostX surveys are predominantly MWD with only a few single-shot data included, although a few X surveys contain as much as40% single-shot data. Early wells were also surveyed with magneticmultishots, M, but most of the postdrilling surveys were runwith gyroscopic instruments. Gyro instrumentation variesamong contractors; both subsurface recording and surfacereadout instruments also have been used. However, wedo not distinguish here among the conventional gyros, G. The data used for this study are taken from 496 wellsin which more than one of the following survey types isavailable: gyroscopic, single-shot, multishot, or mixes ofMWD and single-shot data. Of these wells, 414 have twosurveys, 73 have three, 8 have four, and 1 has five. Thisrepresents a total of 1,084 surveys. Of these surveys, 500are gyroscopic, 408 are magnetic single shots, 119 are magnetic multishots, and 57 are mixed. We recomputedall surveys with the minimum curvature method. Surveysare compared at a reference depth of 8,000 ft [2438m] subsea, which is about 300 to 700 ft [90 to 210 m] above the top of the productive formation. The projected plan map of Fig. 3 shows the x, y locationat 8,000 ft [2438 m] subsea of the 1,084 available surveys. In a given well, the coordinates x, y, z of the surveys arecalculated as follows: pick one survey of the well to useas a base(when available, the gyro was used, but thechoice does not affect the results). Find the measureddepth, which corresponds to 8,000 ft [2438 m] subsea on this base survey.
SPEDE
P. 257^
Site News