Permeability Determination With a Multiprobe Formation Tester

Abstract

Summary This paper presents analytic models to interpret the pressure transientsmeasured by a multiprobe formation tester. The pressure transients measured bya multiprobe formation tester. The multiprobe tester discussed consists ofthree probes. The sink probe generates a pressure pulse by withdrawing fluidfrom the probe generates a pressure pulse by withdrawing fluid from theformation while the resulting pressure response is measured at the sink probeand at each of two observation probes. One observation probe is positioned onthe opposite side of the borehole on the probe is positioned on the oppositeside of the borehole on the same vertical plane as the sink, and the other isdisplaced vertically on the same azimuthal plane as the sink. The effect of thevarious reservoir parameters on the pressure response is discussed. Applicationof the models is demonstrated through an actual field example. Introduction The introduction of the Repeat Formation Tester (RFT SM) tool in themid-1970's coincided with the exploitation of the North Sea. The RFT tool, which is primarily a device for measuring vertical pressure distribution inopen hole, soon gained industry pressure distribution in open hole, soon gainedindustry recognition. While the RFT technique has been successful, it has notreached its full potential because of its inability to give reliablepermeability estimates. The estimated formation parameters and the collectedfluid samples generally are from the parameters and the collected fluid samplesgenerally are from the contaminated zone rather than the native reservoir;consequently, the associated degree of uncertainty is high. With productiondecline of many of the world's large reservoirs, reservoir management isreceiving increasing attention as emphasis shifts from primary to secondaryoperations. As secondary and tertiary schemes become the dominant means ofrecovery, the measurement of permeability in heterogeneous and anisotropicformations will become increasingly important. Because of its effect on manyreservoir displacement processes, vertical permeability often is the singlemost important reservoir parameter. permeability often is the single mostimportant reservoir parameter. Although vertical permeability measurementsroutinely are made from cores, reliable in-situ measurements of verticalpermeability over a large rock volume are much more desirable. The transientwell tests proposed for estimating vertical permeability may be classified asvertical interference testing. permeability may be classified as verticalinterference testing. These techniques have not been widely accepted becausethey usually require cased holes, and the measurements are masked by wellborestorage effects. To eliminate these effects, longer (hence costly) testingtimes are required. For these reasons, industry needs an alternative means ofestimating large-scale, in-situ vertical permeability. permeability. Moran and Finklea first proposed quantitative methods for estimating permeability usingpressure-transient data obtained from the wireline formation tester, thepredecessor to the RFT tool. Their method corresponds to a finite, sphericallyshaped sink of radius r in an infinite, isotropic medium. During drawdown, thespace around the perforation can be divided roughly into three sphericalregions: a steady-state zone near the sink where the total flow rate isindependent of distance r from the sink, the undisturbed formation far from thesink where the flow rate vanishes, and the transition zone between the twowhere the flow rate decreases continuously from the constant steady-state valueto zero at some distance far from the sink. During drawdown, as time increases, the boundaries between successive zones move farther from the sink. Forsufficiently large times, the Moran and Finklea spherical solution yields asteady-state pressure difference, (1) The actual geometry at the RFT probe is not of a spherical source but of adisk source. Therefore, an effective probe radius, r, must be defined. For anisotropic medium, r, was defined as one-half the probe radius. Sharma and Dussan later showed that this definition was not valid and that the correctresult is r =2rp/. The RFT probe enters the formation from the borehole. Because the boreholewall is considered impermeable, the flow pattern cannot be spherical. At thesink probe, in the limit of steady-state flow, Stewart and Wittman accountedfor this wellbore effect in terms of a shape factor by writing Eq. 1 in theform (2) The shape factor, C, is essentially one-half the ratio between the pressuredrop at a specified point with a borehole to the pressure pressure drop at aspecified point with a borehole to the pressure drop without a borehole. Depending on the borehole diameter, C varies between 0.5 for spherical flow and1.0 for hemispherical flow. For an 8-in. borehole in an isotropic formation, Stewart and Wittman defined C as 0.645 when r was taken as one-half the probeaperture radius. Wilkinson and Hammond developed analytical derivations forcalculation of steady-state shape factors as a function of rp /rw andanisotropy. They found that, for isotropic formations with a smallprobe-to-wellbore radius, the effect of the wellbore was small. When rp/rw=0.05, the effective shape factor was found to be 0.96 (it would be unity foran infinite-radius well) when the correct definition of the effective proberadius r =2r / was used. For anisotropic formations with kz less than less thankr, however, the wellbore correction was appreciable. No matter howsophisticated the interpretation, the measurements obtained from the RFT tool(sink probe for the multiprobe configuration) are uncertain for the followingreasons: mudcake blocking the probe, damage to the formation resulting from themechanical setting of the probe, non-Darcy flow near the probe, mud particleshaving migrated into the formation, gas evolution in the near-probe region, andinternal fines migration. The non-Darcy flow results from the high fluidvelocities expected in the near-probe region. According to Muskat, deviationsfrom Darcy's law occur at Reynolds numbers exceeding unity. At high Reynoldsnumbers, the effects of inertial forces and turbulence begin to contributesignificantly to the pressure gradient. It would not be unusual to observe flowin the near-probe region with a Reynolds number greater than 4. Because the Reynolds number is proportional to the fluid velocity and, in spherical flow, the velocity decreases as 1/r, the non-Darcy flow phenomenon is localized tothe sink probe only and has negligible phenomenon is localized to the sinkprobe only and has negligible effect on the observation probes. This paperemphasizes permeability derived from the pressure data recorded away from thepermeability derived from the pressure data recorded away from the sink probe. The multiprobe concept originally was proposed in a U.S. patents in 1956. Fig.1 shows the probe geometry of the multiprobe patents in 1956. Fig. 1 shows theprobe geometry of the multiprobe tool discussed here. In operation, one probe(the sink) withdraws fluid from the formation. The resulting pressure transientis measured simultaneously at the sink and the observation probes. From thesepressure data, both vertical and horizontal permeabilities can be determined. Also, the observation probes are not affected (at least, not to leading order)by the adverse flow effects at the sink probe. Because of the distance betweenthe observation probes and the sink probe, the permeabilities arerepresentative of probes and the sink probe, the permeabilities arerepresentative of a larger length scale than those determined by a single-probetool. SPEFE P. 297

A New Technique to Determine Horizontal and Vertical Permeabilities from the Time-Delayed Response of a Vertical Interference Test

“…However, based on the discussion in the previous section, the pressure drop is inversely proportional to horizontal permeability. Goode and Thambynayagam (1992) had the same observation in multiprobe testing. Hirasaki (1974) proposed a vertical permeability pulse test.…”

“…In a multiprobe testing tool in petroleum industry, the flow interval isolated by straddle packers is replaced by a single probe, as presented by Goode and Thambynayagam (1992). Then the straddle packers are not needed.…”

Lessons learned from Wireline Formation Testing with Single and Multiple Probes

“…This work along with other studies 15,17,18 shows that the formation testing with single probe or multiprobe is very valuable in intergreting reservoir data. …”

Test Design for Vertical Permeability Determination From a Conventional Pressure-Buildup Test

“…Single and multiprobe formation tests may provide horizontal and vertical permeability at the time a well is logged in open hole. 1,2,3 Single well vertical interference tests 4 have been proposed for cased hole completions with more than one set of perforations in the same production interval, but these tests are expensive to conduct, and the results can be subject to error in the case of a cement leak. A hybrid approach by Pop, et al, 5 performs the vertical interference test with a wireline-conveyed straddlepacker tool.…”