Formation Multi-Tester (FMT)Principles, Theory, and Interpretation

01987 Western Atlas International

CONTENTS

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1Operation of the FMT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2Calibrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Strain Gauge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7Quartz Gauge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Qualitative Indications from FMT Pretest. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8Reservoir Fluid Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

Density or Specific Gravity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14Resistivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16Viscosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Fluid SampleTest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20Variable Pressure Control (VPC). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Segregated Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Estimating Sampling Time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

Evaluating Recovered Fluid Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23When Sample Recovery is Primary Native Formation Fluids . . . . . . . . . . . . . . . . . . . . . . . . 23Prediction of Water Cut . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25Recovery of Small Volumes of Formation Fluid or No Formation Fluid. . . . . . . . . . . . . . . 26Technique for Various Size Recoveries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

Fluid Flow in Porous Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27Darcy’s Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

Pressure Drawdown - Permeability Estimate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30Pressure Buildup - Permeability Estimate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

Spherical Pressure Buildup - Permeability and Formation Pressure . . . . . . . . . . . . . . . . . . . 31Cylindrical Buildup - Permeability and Formation Pressure . . . . . . . . . . . . . . . . . . . . . . . . . 34

Comparison of Permeability Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35Bed Thickness Definition During Buildup. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37Subsurface Pressure Regimes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

Hydrostatic Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37Overburden Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38Formation Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

Applications of FMT Pressure Measurements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39Measured Depth vs. True Vertical Depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39Pressure Regimes in Water-Bearing Reservoirs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41Supercharging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41Selection of Test Intervals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42Pressure Gradients and Particular Pressure Regimes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42Determination of Movable Formation Fluid Density

in Zones with High Connate Water Resistivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42Defining Gas/Liquid and Oil/Water Contacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42Zone Isolation or Communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44Determination of Oil/Water Contact Below Total Depth of the Borehole. . . . . . . . . . . . . . 46Reservoir and Zonal Depletion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47Monitor Injection Program in In-Field Wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47Fracture Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47Extremely Tight Formations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49Grain Size Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49FMT Pulse Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

FMTREALITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50List of Symbols, Including Subscripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56Appendix A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

Permeability from Drawdown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58Permeability from Spherical Buildup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58Effective Bed Thickness Computation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58Permeability from Cylindrical Buildup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59Time Estimate for Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

Formation Multi-Tester (FMT)Principles, Theory, and Interpretation

INTRODUCTION

The Formation Multi-Tester (FMT) is a sophisticatedsystem of wireline testing equipment designed tomeasure formation pressures at any number of depthlocations per single trip into an uncased wellbore. Thesepretest pressure measurements also provide a means ofconfirming adequate packer seal and a quick qualitativeestimate of formation permeability prior to making adecision regarding the opening of one or both of the fluidsample chambers.

Run on a multi-conductor logging cable, the FMT canbe accurately positioned at selected depths by referenceto a sequentially recorded spontaneous potential (SP)or gamma ray (GR). Formation testing normally followsthe recording and evaluation of openhole logs fromwhich potential target intervals are identified. Wirelinetest information provides a fast and economical methodfor identifying theproduction potential of the targetedreservoirs.

The FMT provides more detailed and reliable verticalpressure profiles than can be expected from drillstemtests (DST). The ability to acquire multiple pressurereadings provides a much quicker and less expensivemethod for obtaining a reasonably accurate profile ofvertical pressure gradients. FMT pressure measurementsare acquired from specific depth intervals while DSTgauges are commonly located in the test tool string abovethe interval tested. The nature of the fluid between thezone being tested and the DST gauge depth provides forsome uncertainty.

A profile of several DST pressures through a commonpressure regime may not always define an accurate gra-dient. In a similar environment, FMT pressure data oftenprovides descriptive gradients for gas or fluid type. Thedata given in Fig. 1 exemplify such a case.

FMT pressure data are recorded on film and presentedin both analog and easy-to-read digital formats. The sta-tionary recordings at individual depths are a plot ofpressure vs. time. Direct digital readouts are observableon surface instrumentation during the test, allowing forquick decisions on whether or not to open the fluid sam-ple chambers. Two sample chambers can be filled at thesame depth and segregated, or the two chambers can befilled at two different depths per trip in the borehole.Several chamber sizes are available for sampling. A

Comparison of FMT pretest pressures and drillstem testpressures.

unique variablepressurecontrol (VPC) provides forim-proved sampling by avoiding excessive differentialpressure across the packer seal. An analysis of therecovered fluids is provided to aid in prediction of reser-voir production characteristics.

The capability for multiple pretest pressure recordingshas made the FMT the primary openhole system formeasuring vertical pressure distribution in a wellbore.The numerous pressure readings have been used toestablish the hydrostatic gradients of mud columns, fluidand gas gradients in reservoir rocks, and verticalpermeability barriers. Comparisons of FMT gradientson adjacent wells have been found useful in describingthe presence or absence of lateral communication. In-formation from the pressure measurements can besignificant in association with lost returns during cemen-ting operations. Selective perforating and selection ofmethods to best control acid or frac fluid placement canalso be improved through the use of FMT data. The in-formation derived from the FMT can therefore be usedto optimize completion methods and maximize ultimatereservoir recovery.

1

OPERATION OF THE FMT

The FMT makes pressure measurements while the toolis stationary at selected depths in an uncased borehole.If the pretest pressures indicate a good packer seal to theformation and a relatively permeable zone, fluid samplesmay be recovered by opening a sample valve and allow-ing one of the two sample chambers to fill. Photographsof the FMT packer section are shown in Fig. 2.

FIGURE 2Three views of the FMT packer section.

The downhole tool system includes the control elec-tronics, a hydraulic section, and the test samplechambers. A schematic diagram of the subsurfaceassembly is shown in Fig. 3. The control electronics andhydraulics are located in the upper part of the tool string,with the packer seal section and pistons directly below.Sample chambers are attached to the lower end of thetool assembly. A number of different chamber sizes arecurrently available (e. g., 1 gallon, 4 liter, 10 liter, and 20liter). Other unique sizes may be found, depending upongeographical location. Without sample chambers on thetool string, the packer section is located approximately5.5 ft off bottom, or 9 ft if the Hewlett-Packard quartzgauge is run. With two chambers of 10-liter capacityeach, the packer section is located approximately 31 ftoff bottom (34.5 ft with the H-P gauge).

FIGURE 3Schematic diagram of the subsurface assembly of FMT withVariable Pressure Control.

2

Operation of the downhole equipment is controlled from permeability indication. Tracks II and III are dividedthesurfacelogging unit. A film recording is made of each into half-track digital scales of pressure (1000’s, 100’s,complete cycle of operation. Recorded measurements are 10’s, units), providing for a more accurate reading of thein time (seconds) rather than depth as the tool is sta- recorded pressures. Hydraulic pressure may also betionary (see Fig. 4). The time grids or time lines are com- presented in Track I (dashed trace) to help identify theparable to the depth grid on traditional log recordings various stages in the tool’s set and retract cycles. Pressurewith each line representing two seconds (rather than two listings are also available. Temperature may also befeet) when English depth measurements are used. presented in Track I when the series 1966 electronics isPressures are recorded as an analog trace in Track I, pro-viding a quick-look profile of packoff effectiveness and

run.

PUMP PRESSURE

.---------------I 3001

ANALOG

(psi)

SET,PACKER

I---.\\

1

HYDR(PRES

IO set

su‘ATIC‘RE I

t0

t = 8 set

HYDROtTATlCPRES+lRE

STATION DEPTH - XX840

HYDROSTATIC PRESSURE - 4546 psi

FORMATION PRESSURE - 3978 Dsi

RECORDED DIGITAL SAMPLING PRESSURE

(psi)

.....................

.....................

.............................................

.........................................................................................................................................................................................................................................................................................................................

...............

FIGURE 4Example of pretest pressure recording.

At each designated test depth, operational practice in-cludes both a before and after recording of hydrostaticmud pressure, i. e., before actuating and after retractingthe hydraulic packoff section (see Fig. 4). The FMT hasan internal motor, pump, and hydraulic system which areused in actuating and retracting of the packoff section.Hydraulically activated setting pistons cause a rubberdonut-shaped packer element to press tightly against theborehole wall. A special nitrile rubber, which is sulphurcured, with peroxide-cured o-rings, is available for usein H2S environments. Hydraulic pressure is recorded atthe surface, indicating proper (or improper) setting ofthe tool. The tool mandrel is held away from the boreholewall to reduce the possibility of differential sticking.

A snorkel tube is then forced into, or pressed tightlyagainst, the formation. This is followed by the movementof a small piston, called the pretest piston, allowing 10cm3 of formation fluid to enter the pretest chamber ata more or less constant flow rate. The effect of thisvolume extraction on the formation pressure is observedand recorded at the surface (see Fig. 4). A 5-cm3 plug isavailable for use in wells where extremely tight forma-tions are expected.

At this point the logging engineer must be patient andallow the pressure to build up adequately to formationshut-in pressure (see Fig. 4). It is essential that shut-inpressure be recorded as long after the flowing portionof the pretest as possible in order to allow the buildupto approach actual formation pressure. If the pressuretest is terminated too early, the shut-in pressure readingwill be too low since sufficient buildup did not occur.Early termination of the pretest will also prevent lateruse of spherical and/or cylindrical model analysis.Pretest shut-in pressure is often referred to as initial shut-in pressure.The tool is then (1) retracted prior to moving the tool,or (2) a sample valve is opened to allow formation fluidto flow into a sample chamber. In either of these casesthe tool is ultimately retracted, at which time an equaliz-ing valve is opened and the pretest fluid is expelled fromthe tool. A system schematic is shown in Fig. 5.

Operation of the pretest piston and the concurrentmeasurement of pressure are the keystones to determina-tion of formation pressure and inferences of permeabili-ty. Formation fluid enters the tool through the probe dur-ing the pretest. As the probe is pressed against the faceof the formation (it may even penetrate softer, uncon-solidated rock) it also defines the area of flow from theformation as well as eliminating lateral mud cake entryinto the tool. A schematic of the probe in both the retractand set position is given in Fig. 6. A sleeve, inserted inthe probe, is slotted to function as a filter. The slits are0.015 in. wide to prevent any debris larger than the slits

from entering the tool. The sleeve is easily removed atthe surface for cleaning between runs.

PRESSURETRANSDUCER

PRETESTPISTON

u

I JEQUALIZINGVALVE

--L WELLBORE

SAMPLE VALVES

SAMPLE TANK

p SAMPLE TANK

#l

#2

FIGURE 5Schematic of FMT system.

FIGURE 6Schematic of the FMT probe in both the retract (upper) andset (lower) positions.

4

If the pretest data indicates adequate conditions for fluidsampling of the formation, a sample valve may beopened. As many as two fluid samples may be recoveredper descent into the wellbore. These samples may betaken at the same depth or at different depths. If bothsample chambers are filled at the same set depth, theywill be segregated from one another.

Atlas Wireline Services has developed a unique VariablePressure Control (VPC) system for use with the FMTThe VPC is a variable orifice valve located upstreamfrom the sample tank valve. Both valves are closed whenno sample is being taken. To obtain a sample, the tankvalve is opened first, followed by the variable orificevalve. The variable orifice valve is controlled from thesurface and is opened only until a suitable flowingpressure is attained. The VPC avoids excessive pressuredifferentials and samples are obtained successfullywithout guesswork. This feature also prevents formationplugging in unconsolidated sands. The logging engineercan adjust the VPC to pressure increments as small as5 to 10 psi. Samples can be obtained without damage orplugging from formation collapse. In addition, samplescan also be taken above bubble point pressure, therebyeliminating non-representative gas/oil ratios caused bythe effects of relative permeability.

Testing at multiple flow rates and multiple drawdownpressures may be useful in evaluating formationmechanical properties relating to sand control and con-solidation. An analog pressure record showing VPCsampling at three different flow rates is shown in Fig. 7.

Segregated sampling at the same set depth usually in-volves filling the larger sample chamber first, the ideabeing to drain the flushed zone (mud filtrate) as muchas possible. The second sample chamber is then ideallyfilled with representative reservoir fluids.

The tank valves in the VPC-FMT system can be openedand closed as often as required. This feature allows thelogging engineer to check for plugging and enables himto reuse the first sample tank in the event of an earlypackoff failure during segregated fluid sampling.

After a sample chamber is filled, the sample valve isclosed by a spring and kept closed because of the bal-anced seal design, thereby sealing the fluid in the sam-ple chamber at formation pressure. The pressuretransducer transmits the final shut-in pressure to the sur-face where it is recorded. A film record of the pretest andsampling steps is illustrated in Fig. 8.

After completion of the sampling, the hydraulic systempressure is released and hydraulic pressure reversed inorder to retract the packer and backup shoes from the

I-----me -i

SECTION OFLOG REMOVED

Analog pressure record showing VPC sampling at three dif-ferent flow rates.

face of the formation. As with a standard pretest,hydrostatic pressure is again recorded to provide averification of transducer stability, repeatability, andreliability. Hydrostatic pressures recorded before andafter tool setting should read within +l psi of oneanother (assuming no change in the mud column) (seeFig. 8).

After the FMT tool and sealed samples are returned tothe surface, apressureregulator, separator, andgasmeterare used to extract the samples individually. Recoveredgas is bled from the sample chamber through theseparator and measured by the gas meter (in ft3 at sur-face conditions). Water and oil are drawn off in theseparator and then poured into a calibrated vessel wheretheir volumes are measured in cm3. When H,S issuspected in the sample, the gas is bubbled through an

5

ANALOG

(PW

RECORDED DIGITAL SAMPLING PRESSURE(PSI)

~ I~i’-i~~~:i_ :::::::::::FIGURE 8.Film record of pretest and sampling steps.

6

H,S scavenging bottle to remove the H,S. The remaininggas is vented into the atmosphere. Recovered water is

logging tools in that they provide a basis for correcting

routinely tested for chloride concentration and anystrain gauge measurements for temperature effects en-

recovered oil is measured in terms of ‘API gravity. Ancountered in the borehole environment. A high-qualitydeadweight tester is used. A calibration test strip is shown

example of a sample evaluation is shown in Fig. 9. Fur-ther analysis, if needed, can be made in the laboratory.

in Fig. 10.

If an undisturbed sample is required for laboratoryanalysis, a breakoff tank can be utilized. The breakofftank is a 4-liter capacity high-pressure tank which canbe detached from the FMT at the surface withoutbleeding off any pressure or fluids. The breakoff tankmeets U. S. Department of Transportation safety stan-dards for transport by any common carrier.

CALIBRATIONS

Strain Gauge

The routine shop calibration of FMT strain gauges isessential to obtaining accurate pressure measurementsin the borehole environment. These routine checksshould be performed within a 30-day period and areanalogous to shop calibrations performed on other

FIGURE 10Calibration test strip for the FMT strain gauge.

FORMATION TESTER RECOVERY & INTERPRETATION DATA

T E S T N U M B E R 3 Depth 2588.0 -0RMATION and MUD DATA

Test Type: 0 Open Hole 0 Cased Hole Formation SWAN HILLS

Porosity % SourceTOOL DATA Rt

R W zi

“C“C

Tool Type 1965 Source of RwProbe Type DUAL PASSAGE NaCl (Chart) ppm; Cl (Titrated) PPnSample Unit Size 3550 cm3 Water SaturationFlow Control OPEN Rmf 1.55 ii? 20 “CTool Number 71577 Source of Rmf

NaCl (Chart) ppm; Cl (Titrated) wn

PRESSURE DATAInitial Shut inBuild Up TimeSampling Range

Sampling TimeFinal Shut inFinal Shut in TimeHydrostaticSurface Chamber

241700.1

400 - 100003.75

241106

295107067

RECOVERY INFORMATIONkPa Gas 0.0127 msmin Distillate cm3 APl/15.5”(

kPa Oil 200 cm3~APl/15.5”(min GOR

kPa Water 2950 Cm3; Res 0.18 18 0~

min NaCl (Chart) ppm; Cl (Titrated) wmkPa Mud cm3; Sand cckPa Formation Water %

INTERPRETATION

Remarks 2nd TANK OF SEGREGATED SAMPLE

May be Expected At This Depth.

FIGURE 9Typical sample evaluation.

7

The Atlas FMT is electrically calibrated prior to beinglowered into the well. The wellsite calibration methoduses deadweight tester data and temperature correctionsdetermined in the laboratory on the strain gauge pressurerecordings. The calibration process is repeated on eachsurvey to ensure that proper response is maintained. Thewellsite calibrations also verify the reestablishment of theshop calibration response.

Quartz Gauge

The Hewlett-Packard quartz gauges are accuratelycalibrated by the manufacturer. These gauges are moreaccurate than the deadweight testers used to fieldcalibrate strain gauges, therefore a field calibration is notrequired. However, it is important that periodic shopcomparisons with a deadweight checker be made to en-sure/verify stable quartz response with time. Quartzgauges do require significant temperature correction andAtlas’ H-P probes are modified to measure thetemperature of the most temperature-sensitive compo-nent in this gauge.

High-precision quartz gauges are typically used whenstudies of formation pressures require the utmost ac-curacy These gauges typically have an accuracy asfollows.

l If temperature is known to l°C accuracy, f 0.025%of pressure reading.

l If temperature is known to 10°C accuracy, + 0.1% ofpressure reading.

l If temperature is known to 20°C accuracy, + 0.25%of pressure reading.

The temperature is accurate within + 5’F (+ 3OC).

Quartz gauges also provide good repeatability (+ 0.5psia) and a large amount of pressure data. Their singlelimitation is the time factor necessary for stabilizationbefore measuring true pressure, i. e., several minutes maybe required before reaching stabilization. It is alsonecessary to depth-correct pressures read from thequartz gauge to a pressure reference level due to the factthat the quartz gauge is physically located lower than thestrain gauge on the tool string.

The quartz gauge consists of two crystal oscillators, bothbeing sensitive to temperature and pressure. However,one crystal performs as a sensor of fluid pressure andtemperature while the other crystal is used as a referenceto provide frequencies suitable for transmitting onwirelines. They are calibrated as a pair and both musthave the same temperature for equilibrium pressure.Temperature changes of a few OF per minute can cause

errors of 100 psia for standard H-P gauges. SelectedAtlas modified H-P gauges will have a maximum errorof 20 psia under the same conditions and will read within?5 psia within 2 minutes of the temperature change. Theaccuracy will be within +2 psia when the rate oftemperature change is less than 0.5 OF per minute.

QUALITATIVE INDICATIONSFROM FMT PRETEST

The curve character of the pretest analog recording ofpressures is affected by the pretest and sampling se-quences. A schematic illustrating flow during pretest isshown in Fig. 11A. The analog pressure recording for atypical test is shown in Fig. 11B. The setting of the toolbegins on the left of both figures with time increasingto the right. At the left of Fig. llB, hydrostatic pressureis recorded but when the equalizing valve is closed, therubber packer engages and is pressed against the mudcake. A hydraulic seal is likely to occur before the packerand mud cake are fully compressed, therefore thepressure in the tool flow line is often observed to momen-tarily build up slightly above hydrostatic pressure.

The pretest piston is then drawn back, allowing 10 cm3of formation fluid to enter the pretest chamber at a con-stant rate as shown in Fig. 11A. The end of the flowingor drawdown phase is indicated at time tl in Figs. 11Aand 11B. As the piston motion ceases, flow stops and thepressure builds up to formation pressure as shown on theright side of Fig. 11B. When the tool is retracted, thedrawdown piston is reset thereby purging the pretest fluidinto the wellbore, and the equalizer valve is opened allow-ing the pressure to return to hydrostatic. The differencebetween flowing (drawdown) pressure and shut-inpressure is AP and the time necessary for flow to stopfrom the beginning of the drawdown is referred to as At.Both AP and At are illustrated in Fig. 11B. These valuesof AP and At are used to determine permeability fromthe drawdown.

A pressure record of the FMT internal hydraulic systemduring the pretest sequence is illustrated in Fig. 11C. Thesteps indicated are (1) the closing of the equalizing valve,(2) the packer engagement, (3) the pretest piston move-ment, and (4) the completion of the pretest piston move-ment. This internal pressure record is important formonitoring tool performance and is usually recorded asa dashed trace in Track I (see Fig. 4). Furthermore, thismeasurement is not used directly in the evaluation of for-mation pressure data.

During pretest, flow, shut in, and the stopping of thepretest piston will coincide only if formation perme-ability is adequate to allow the formation fluids to flowfast enough to fill the volume created during the move-

Aq PRETEST FLOW RATE’ (HIGH PERMEABILITY)

A

P PRETEST PRESSUREHYDROSTATlC

PRESSURE-

LW- SHUT INFORMATION PRESSURE

B

- At

‘1 f

INTERNAL HYDFjAULlC PRESSURE

C

EQUALIZING VALVE CLOSURE

FIGURE 11Schematic illustration of flow during pretest.

ment of the pretest piston. If formation permeability istoo low, the pretest piston will cause the pressure to dropbelow the bubble point and multiphase flow may occurat the tool/formation interface. Although the piston hascompleted its stroke, the formation will continue totrickle fluid into the tool until 10 cm3 has flowed. Atypical analog pressure record under these conditions isillustrated in Fig. 12A, B, and C. Observed pressures willeventually build up to formation pressure if sufficienttime is allowed. An example of a long duration pretestis shown in Fig. 13.

The effects of formation permeability in the vicinity ofthe FMT probe on the pretest pressure record are il-lustrated in Figs. 14A, B, C, D and E. These comprisea family of typical FMT pretest pressure analog recordsfor permeabilities varying in order of magnitude in-crements from 100 md to 0.1 md to tight. Note that theflowing time increases between 10 md and 1 md, in-dicating that the formation permeability is sufficientlylow so that it cannot flow fast enough to fill the volumedisplaced by the pretest piston. The illustrations in Fig.14 are intended only as guidelines since the actualpermeability depends on the drawdown from formationpressure, flow rate, and nature of fluid.

Other factors often affect the character and quality ofthe pretest pressure record. Debris drawn into thedrawdown line during pretest may cause plugging and

PRETEST DISPLACEMENT

A

(PHYOROSTATlC

PRETEST PRESSUREPRESSURE

SHUT INFORMATlON PRESSURE

INTERNALHYDRAULICPRESSURE

FIGURE 12Typical analog pressure record in a low permeability formation.

in extreme cases prevent 10 cm3 being drawn into thetool. In Fig. 15A, light plugging is indicated by a rough,irregular response during the drawdown or flowingperiod. Severe plugging, if it occurs immediately upondrawdown, is virtually indistinguishable from a tight test(compare Figs. 15B and 14E). The presence of gas in theflowline causes the abrupt changes in pressure to occurmore gradually due to gas compression and expansionas shown in Fig. 16A. If the packer is set on a tight for-mation, the effect of pretest is to expand the gas in theflowline by 10 cm3 and reduce its pressure to some non-zero constant value as shown in Fig. 16B. In either case,gas may be eliminated from the flowline by opening thesample chamber when set against a tight formation. Thisprocedure in effect allows the gas to distribute itself overthe small flowline and much larger sample volume,thereby allowing it to be captured in the sample jug (thisprocedure cannot be done with all tools without wastingthe sample test).

Seal failure is caused by the inability of the rubber packerto isolate the probe flow channel from the mud columnand may occur at any time during the pretest sequence.Figure 17A illustrates a catastrophic seal failure such asmight occur in a washout or highly rugose hole. Thepretest pressure record remains at hydrostatic eventhough the pretest piston goes through its cycle. In Fig.17B, a relatively low permeability is indicated; however,upon closer inspection the final formation pressure and

9

RECORDED DIGITAL SAMPLING PRESSURE

FIGURE 13A long duration pretest.

10

APPROXIMATE PERMEABILITY

I 1 RECORD

C

FIGURE 14Family of typical analog pretest pressure records for different permeability ranges.

FIGURE 15Analog pretest pressure recordings for (A) irregular light plugging and (B) severe plugging.

11

FIGURE 16Examples of (A) gas compression/expansion in the flowline and (B) tight gas zone, where gas expands to 10 cm3 in the flowline.

II1 I

B

FIGURE 17Example of (A) catastrophic seal failure and (B) case where apparent formation shut in pressure is suspiciously similar to hydrostaticpressure.

12

initial hydrostatic pressure are nearly identical. Thesesituations should be viewed very carefully since eithera seal failure has occurred or virtually no overbalanceexists with respect to the formation. The latter situationmay indicate the need to weight up the mud as the wellmay be near blowout conditions. Measurements of thesetypes should be repeated to determine which situationexists so that appropriate action can be taken.

The FMT presentation includes both analog recordingsof the pretest pressure record and the internal FMThydraulic pressure. Both are recorded in the left-handtrack (see Fig. 18). Where greater accuracy and resolu-

right of the depth track as shown in Fig. 18. The recordersteps the pressure data in digital increments of 1000, 100,10 and 1 psi from the track nearest the depth track andthen to the right. For example, at time 80 sec. in Fig. 18,the digital record indicates a formation pressure of 3927psi and still building up slightly. It is apparent thatresolution in this case is 1 psi. This example is an FMTmeasurement with the strain gauge. The Hewlett-Packard gauge is designed to improve the resolution to0.1 psi and the 1000 must be read from the analog (TrackI) data with 100, 10, 1, and 0.1 values read from the digitaltrack. All pressure data (temperature corrected and un-corrected) is recorded on magnetic tape and may be re-

tion are required, four digital tracks are placed to the tained for later processing.

PUMP PRESSURE- - - - - - - - - - - - - - - - - - - - - - - RECORDED DIGITAL SAMPLING PRESSURE(Psi)

ANALOG

................ . . . . . . . . . . . . . .

........................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................ ..................... ................................... ................................... ........................................................................................................................ ....... ........

FIGURE 18Analog pressures allow a quick qualitative reference. The digital pressure record provides greater accuracy and resolution for quan-tification of pressure data.

13

RESERVOIR FLUID PROPERTIES

In conjunction with quantitative well log evaluations, thefluid samples and pressure data obtained from the FMTcan be used to estimate formation pressures, perme-abilities, hydrocarbon production rates, and depths ofoil/water, gas/oil, and gas/water contacts. The samplesrecovered also supply information on the type of forma-tion fluids, gravity of oil, water cut, and gas/oil ratio.FMT data interpretation involves considerations of fluidpressure behavior and other physical properties of for-mation fluids such as density, resistivity, and viscosity.The pore space of reservoir rocks may contain water, oil,and gas as a single phase or in any combination of thesefluids. It is imperative that these properties of reservoirfluids be known or reasonably approximated in order toreliably evaluate the production characteristics of thereservoir rocks.

Density or Specific Gravity

Density of water depends upon its salt content,temperature, and pressure. The specific gravity of asubstance is the ratio of its density to that of water atspecified temperatures. Density of water decreases with

increasing temperature but increases with higher totalsolid concentration and pressure. The effect of pressureon the density of water is comparatively little, as can beseen on the chart in Fig. 19 which can be used to deter-mine the density of water. Alternatively, if density at acertain temperature and pressure is known, total dis-solved solids or chlorinity (in ppm) can be read from thechart.

Specific gravity of oil is related to its API gravity by therelation

141.5Yo = (1)

OAP1 + 131.5

where y0 is the specific gravity of oil at 60°F referred tothat of water at 60’F. When dissolved gas is present inoil, the specific gravity of the latter depends upon thegas/oil ratio, decreasing as the gas/oil ratio increases.Figure 20 can be used for determining reservoir densityof oil in g/cm3 for a known value of GOR. Figure 21shows the variation of specific gravity of oils withtemperature while dry gas density, as a function of reser-voir pressure and temperature, is illustrated in Fig. 22.

CDNSTRVCTED FROM DATA INTABLES OF INTERNATIONALCRlTlCAL VALUES AND LABORA-TORY DENSITY MEASUREMENTS

t i i i i i I

TdTAL DI’SSOLVED SOLIDS ;20 2 8 0, 1 , ppm I IO’ 1 , ,, , ,

240 I ,“A \,-t”.\rb”\l\,I

11 6 0

tH E X A M P L E RESER~~IRT~MPERATUR~ = 175 OF ’

FIGURE 19Chart for determination of water density.

14

I 1600

I 1800

I 2000

30 40

OIL GRAVITY (OAPI)

50 60

FIGURE 20Chart for determination of reservoir density of oil.

C,H, = EthaneC,H. = PropaneCal,,, = ButaneIC.H.., = lsobutana

FIGURE 21Gravity-temperature relationship.

15

FIGURE 22Density of dry gas vs. temperature and pressure.

Resistivity

All porous rocks contain some water. By virtue of ion-ized salts contained in solution, these formation watersare electrically conductive but may exhibit resistivitiesranging from 0.01 ohm-meter to several ohm-meters. Thepredominant salt in these solutions is sodium chloride.Resistivity of such an electrolyte decreases with increas-ing salt concentration (due to the higher concentrationof ions, which carry electric charges) and highertemperature (which increases the mobility of those ions).

Resistivity of formation water may be determined bydirect resistivity measurement on a sample, chemicalanalysis, or an estimation of equivalent NaCl (in ppm)from well logs. The nomograph in Fig. 23 shows theresistivity of a brine solution as a function oftemperature and equivalent NaCl concentration.

Viscosity

Viscosity of a fluid is a measure of its resistance to flow.The lower the viscosity of a fluid, the more readily itflows.

Viscosity of water decreases with increasing temperaturejust like honey thins on warming. Water viscosity alsodepends upon salinity. These variations due totemperature and salinity are shown graphically in Fig.24. Changes in water viscosity are significant when deter-mining permeability from drawdown. The effect ofpressure on the viscosity of water is negligible.

Viscosity of gas-free crude oil also decreases withtemperature (Fig. 25). From a knowledge of crude oil‘API gravity and formation temperature, the viscosityof gas-free crude oil can be determined (Beal, 1946). Theamount of gas dissolved in oil has an important bearingon viscosity at reservoir conditions. Figure 26 is used tocorrect the viscosity of dead oil (at reservoir temperature

and atmospheric pressure) to the viscosity of gas-saturated oil for the known GOR at reservoir conditions.

Viscosity of natural gas depends upon its gravity withrespect to air at standard conditions. Effects oftemperature and pressure on the viscosity for naturalgases, ranging in gravity from 0.6 to 1 .O, may be approx-imated by use of Fig. 27.

Nomograph for determination of resistivity or salinity of brinesolutions.

16

011 I I I I I I I100 150 200 250 300 350 400 d

RESERVOIR TEMPERATURE (“F)

FIGURE 24Water viscosity vs. temperature and salinity (in ppm NaCl equivalent).

4000 - \

2000 - \

600 j \400 -\ \

I\\2oo \\ \

100 -&

60: \'40- \

20 -

10 7

6:4-

2 -

1.0 7

0.6 ::

0.4 -

0.2 -

0.1 ..ILL10

\

t\\1

-

Ju

\-5

\

t\

-

CRUDE OIL GRAVITY OAP1 AT 60°F AND ATMOSPHERIC PRESSURE

FIGURE 25Viscosity of gas-free crude oils.

17

M SOLUTION GAS/OILRATIO (ftalbbl)

VISCOSITY OF DEAD OIL (cP)(AT RESERVOIR TEMPERATURE AND ATMOSPHERIC PRESSURE)

FIGURE 26Viscosity of gas-saturated crude oils.

18

0.040

0.030

cv

ci7

8 0.020

5

0.015

0.0100 100 200 300 400 500

TEMPERATURE (“F)la\1-1

0.050

0.040

iE‘0

cz I -8 0.020”>

0.0100 100 200 300 400 500

TEMPERATURE (“F)

Cc)

0.050 -

0.040 -

0.030 -

a^0

E

22 0.020 \ I \ I

TEMPERATURE (“F)fb)

IGRAVITY = 10

RICH GAS

TEMPERATURE (“F)

Cd)

FIGURE 27Charts for determining viscosity of natural gas.

19

FLUID SAMPLE TEST samples per descent into the wellbore. Several differentsizes (capacity) and combinations of sample chambers

The original purpose of wireline formation tests was to are available (see Table 1).provide a means to obtain a sample of formation fluidand bring it to the surface. The multi-set pretest capa- With the FMT, the percentage of successful fluidbility of the FMT has tremendously improved the abili- recoveries has shown tremendous improvement and rigty to determineif an adequatepacker seal and sufficient- time has been reduced. The ability to acquire segregatedlypermeable zone are present prior to opening a sam- samples from the same zone improves the chance of ac-pie chamber. The FMT is also capable of gathering two quiring representative reservoir fluids.

TABLE I

VARIABLE PRESSURE CONTROL FORMATION MULTI-TESTER (VPC-FMT) SPECIFICATIONS

Length Without 18 ft 4 in. (w/o temp & H-P gauges) 5.59 m (w/o temp & H-P gauges)Sample Chambers 19 ft 4 in. (wltemp & H-P gauges) 5.89 m (wltemp & H-P gauges)

(sample chambers lengthen Packer is set 5 ft 5 in. above Packer is set 1.65 m above thethe distance below the the bottom of the tool w/o sample bottom of the tool w/o sample tanks;packer, e. g., two tanks; 9 ft w/H-P gauge 2.74 m w/H-P gaugeten-liter tanks and H-Pgauge would provide adistance of 34.5 ft)

Weight, Overall Without 982 lb 445 kgWater Cushion

Maximum Tool Diameter, 5.125 in., w/slim hole pad 13 cm, w/slim hole padRetracted 6.125 in., w/standard pad 15.56 cm, w/standard pad

7.875 in., w/16 in. extension kit 20 cm, w/16 in. extension kit9.188 in., w/20 in. extension kit 23.34 cm, w/20 in. extension kit

Maximum Hole Diameter 12.0 in., w/standard pad 30.48 cm16.0 in. or 20 in. available with 40.64 cm or 50.8 cm available withextension kits extension kits

Maximum Pressure Rating 15,000 psi 103,350 kPa(20,000 w/specially equipped tool) (137,800 kPa w/specially equipped tool

Maximum Temperature 350° F 177ocRating (425°F w/specially equipped tool) (218’C w/specially equipped tool)

Pretest Chamber 10 cm3 (5 cm3 plug is available) 10 cm3 (5 cm3 plug is available)Fluid Capacity

Sample Chambers, 1.06, 2.64, and 5.28 gal tanks are 4, 10, and 20 liter tanks are standardFluid Capacities standard (other tank sizes are (other tank sizes are available in(water cushions are available in some specific some specific geographical areas)available) geographical areas)

Strain Gauge AccuracyWithout templpressure + 0.8% * 0.8%

correctionWith templpresssure +0.13% +0.13%

correctionResolution -+l.O psi k6.89 kPa

Hewlett-Packard Quartz GaugeAccuracy with temperaturecorrection

Resolution kO.1 psi f 0.6894 kPaRepeatability r0.4 psi k2.76 kPaAccuracy at thermal (+ 1 .O psi + 0.1% of pressure (k6.89 kPa + 0.1% of pressure

equilibrium reading) reading)

In addition to identifying the production potential oftargeted reservoirs, sampling is also an effective meansof identifying fresh water aquifers. Drilling fluids aretypically more saline than potable waters. The resistivi-ty of the recovered fluid (Rrr), if some percentage of for-mation fluid is obtained, can provide the fingerprintdefining such potential water supplies. Regulatory agen-cies are interested in identifying such aquifers so as tomake assurances that the water-bearing zones are ade-quately isolated from any potential contamination.Segregated samples, taken at the same packer setting,often provide some uncontaminated formation water.

Sampling has also been used effectively to pinpointgas/oil, gas/water, and oil/water contacts in high-permeability, high-porosity reservoirs.

Variable Pressure Control (VPC)

Atlas Wireline Services’ unique Variable Pressure Con-trol (VPC) allows for better sampling of unconsolidatedformations where excessive drawdown or excessive flowrate might cause formation collapse, resulting in sealfailure, toolplugging, or formation plugging. Earlier ver-sions of FMT tools used a flow line restrictor or watercushion to combat the problem of excessive drawdownsand excessive flow rates. The flow line restrictor wasplaced in the flow line upstream from the samplechamber to limit excessive flow. Water cushions wereused to accomplish the same effect by causing the fluidfilling the sample tank to displace a piston which pushedwater through an orifice into an air-filled chamber. Theflow rate was controlled by installing an appropriatesized orifice prior to the job.

The VPC is located upstream from the sample tank valveand has a variable orifice. Both VPC and sample tankvalves are closed when no samples are being taken. Whena sample is desired, the tank valve is opened first followedby the opening of the variable orifice valve, which is con-trolled from the surface. Once opened, the variableorifice responds to pressure in the sample line by slight-ly opening or closing to maintain a constant pressure.Excessive packer differentials are avoided and samplescan be successfully retrieved without guesswork.

The VPC also permits sample retrieval from zones whichare above bubble point pressure, eliminating npn-representative gas/oil ratios caused by the effects ofrelative permeability.

Segregated Samples

In tight, invaded formations it is often difficult to ob- The time estimate equation only approximates sampl-tain a sample which is representative of formation fluids. ing time because other factors (e. g., relative perme-The two-chamber capability of the FMT improves the ability, flowing pressure, turbulence, debris, plugging,chance of obtaining a representative sample since both etc.) will influence the flow rate into the FMT. If samples

tanks can be filled at one set depth and segregated fromone another. The premise is that the first tank will drainoff most of the invaded filtrate from the flushed zonesurrounding the packoff and the subsequent sample ob-tained in the second chamber will be more representativeof native formation fluids.

Estimating Sampling Time

Another advantage of the multi-set capability is that thepretest pressure data allows for a quick approximationof the time required to fill a sample chamber. The longerthe tool sets stationary and packed off to the formation,the greater becomes the risk of sticking the tool. For thisreason it is significant to know how much time will beinvolved in filling a sample tank.

An estimate of the time period (in minutes) required tofill a one-gallon sample chamber can be made from thefollowing:

t = 63.1 (AP,,~)

qpt (AP,)(2)

where:

t = time required to flow one gallon, inminutes

Appt = drawdown during pretest(Pformation - Pflowing), psi

APS = drawdown during sampling(Pformation - Pflowing), psi

qpt = flow rate during pretest (chambersize/time to fill), cm3/sec

63.1 = conversion factor =3785 cm3/gal

60 sec/min

Equation 2 is a simple extrapolation of flow duringpretest vs. flow during sampling. When sampling is per-formed without a flow line restrictor or water cushion,the sampling flowing pressure is typically very low and

APS = Pf (3)

where pf = formation pressure.

21

larger than one gallon are to be retrieved, the timeestimate derived from Eq. 2 is simply multiplied by thedifference in chamber size (in gallons). For example, a2.75gallon tank would take 2.75 times the valuecalculated in the equation.

Example

A log of a pretest followed by a sample test is given inFig. 28. The flow rate (q) is determined to be 10 cm3/4set, or 2.5 cm3/sec. The pressure drawdown duringpretest is the difference between the shut-in and flowingpressures, which is indicated to be

ap*t = 2263 psi - 2215 psi = 48 psi

If the sample was recovered without a flow line restric-

minigal. Following the pretest, a 2.56-gal (9700-cm3)sample was retrieved in 1.47 min for an actual flow rateof 0.57 min/gal.

If a VPC tool had been used and flowing pressure wasadjusted to 2000 psi, the expected rate would be 4.61min/gal and the retrieval of a 2.56gal sample would havetaken 11.8 minutes. By using the VPC, sampling wouldhave taken a few minutes longer but the danger of for-mation collapse, erroneous gas/oil ratio, and/or debrisplugging of the flow lines would be lessened.

The ability to predetermine a sampling time provides theresponsibleperson at the surface with information whichhelps him to decide whether to chance sampling that par-ticular zone or to move the tool and find a morepermeable depth to sample. The time sampling estimatealso helps thelogging engineer make a judgment on the

tor or water cushion, the flow rate is estimated to be 0.54 proper VPC pressure setting to utilize.

.._--..-- 3 DIGITAL SAMPLING PRESSURE

FIGURE 28Recording of pretest pressures followed by sample pressures.

EVALUATING RECOVERED FLUID SAMPLES

Fluids recovered in the sample tank are mud filtrate,native formation fluids, or a mixture of the two.Recovered mud filtrate is not representative of forma-tion fluid. Recovered formation fluids are presumed toflow into the sample chamber in the same proportionsof gas, oil, and water as they would in production of thezone.

The quantity of recovered fluid is a function of time,fluid viscosity, and pressure in addition to permeability.Quick chamber fillups occur in high permeability for-mations; however, sample chamber fillup can occur intight formations if sufficient time is allowed. Therefore,the amount of fluid recovery is not diagnostic ofpermeability. Fluid recovery in excess of 1000 cm3 is suf-ficient to allow realistic estimates of:

l Gas/fluid ratio, i. e., gas/oil ratio (GOR) and gas/waterratio (GWR)

l Production prediction, i. e., hydrocarbon vs. water,

mud filtrate vs. formation water, per cent water cut,and gas solubility in water and/or oil

l Viscosity of recovered fluids

Samples recovered may be substantially formation fluid,substantially mud filtrate, or any mixture in between.Several methods have been developed to evaluate thesediffering conditions.

When Sample Recovery is Primary NativeFormation Fluids

When a relatively large fraction of the sample volumeis native formation fluid, the empirical chart of Fig. 29may be used to predict the production from the forma-tion. This chart was developed for porosities greater thanabout 25 5’0 and shallow filtrate invasion. The volume ofrecovered gas at surface conditions (in ft3) and recoveredoil (in cm3) is all that is required to utilize the chart. Thischart was prepared for a one-gallon chamber, thereforeall values of recovered volumes must be divided by thesample chamber size used (in gallons) to normalize the

I GAS-OIL RATIO/(ft3/bbl) A / / /

WATER ZONE

OIL RECOVERY - cm’ PER GALLON SAMPLE CHAMBER SIZE

FIGURE 29Empirical interpretation chart for l-gallon sample tank size and high-permeability formations.

23

recovery to a volume-per-gallon basis. The chart givenas Fig. 30 was prepared for a 2.75gallon chamber. Thesetwo charts were empirically derived from a large numberof tests carried out by Shell Oil Company. The chartshave been found to yield realistic estimates when the sumof recovered volumes (converted to subsurfacetemperature and pressure) is not appreciably less thanthe volume of the sample chamber.

Example

The recovered fluids in a l0-liter (2.64-gallon) sample are4.0 ft3 of gas at surface conditions and 1550 cm3 of oiland 8000 cm3 of water (both filtrate and formationwater). The recovery data must first be normalized to avolume-per-gallon basis, so the recovery becomes

4.0Gas Recovery = - = 1.52 ft3 per gallon

2.64

1550Oil Recovery = - = 587 cm3 per gallon

2.64

8000Water Recovery = - = 3030 cm3 per gallon

Entering the above oil and gas recovery data on Fig. 29,the data point falls clearly in the oil zone. Hence, oil pro-duction is predicted with a gas/oil ratio of 410 ft3 perbarrel of oil. In this case the formation shut-in pressure(SIP) is 2800 psi. Since the data point falls well above andto the right of the 2800 psi shut-in pressure (SIP) line,no water production is predicted with the oil. If the datapoint had fallen below the SIP curve, water productionwould have been predicted. An indication of water pro-duction should not necessarily condemn a zone sincethese empirical charts (Figs. 29 & 30) have a tendency tobe pessimistic. Any use of these charts should always besupplemented with other information on the tested zone.

Gas/oil ratio anticipated in production may be estimatedwithout the use of Fig. 29 by the following equation:

GOR =Gas Recovery (ft3)

Oil Recovery (cm3)x 159,000 (4)

The recovery used in Eq. 4 does not have to be nor-malized to a per-gallon basis. Measured values can beused directly regardless of sample tank size. This equa-tion plots as the straight lines of gas/oil ratio (GOR) in

2.64 Figs. 29 and 30.

OIL RECOVERY (cm3) h

FIGURE 30Empirical interpretation chart for 2%-gallon sample tank size and high-permeability formations.

24

Prediction of Water Cut(6)

A prediction of the potential water cut may also be madefrom the recovered fluids. A nomogram given as Fig. 31can be used to predict water cut. Water cut prediction canalso be determined from the following equation:

Formation Water Recovery (cm3)Water Cut (070) =

Formation Water Recovery (cn?) + Oil Recovery (cm3)(5)

Recovered water is typically comprised of both filtrateand formation water. The filtrate must be deducted fromthe total water recovery prior to using Eq. 5. The frac-tion of formation water in the recovered sample can befound from the following equation:

ff,(~O> =%.v %nf - Rrf)

x 100Rrf %lf - &v)

where

ffw = fraction of formation water in theFMT sample, (070)

Rrf = resistivity of recovered fluid

R, = resistivity of formation water

Rm f = resistivity of mud filtrate

The necessary resistivity information is obtained asfollows. First, measurements of the recovered water

100

8 0

6 0

10

8

6

1 2 3 4 5 6 8 10 2 0 30 40 50 60 80 100

- 2 0 0 - 1 8 0 - 1 6 0 - 1 4 0 - 1 2 0 - 1 0 0 - 8 0 - 6 0 - 4 0 - 2 0 0SP c : : : : : ! ! : : : : : ! : : : : : : 1

FIGURE 31Nomogram for estimation of percent water recovered.

25

resistivity are made by either resistivity cell or titrationmethods, with the latter being more accurate. Second,formation water resistivity, R,, is determined from welllogs W, R,,, etc.), produced water samples from off-set wells, water catalogs, etc. Third, the mud filtrateresistivity, R,,, must be determined, again by resistivitycell or titration. A word of caution when determiningR,, is in order, however. R,, values are often observedto be too low, sometimes less than the resistivity of therecovered fluid, R,,. This may result from conditioningthe mud prior to logging or by an ion exchangemechanism. In any case, the values used for Rmfand R,should be checked with logs and mud company reportswhen their reliability is questioned. Fourth, allresistivities must be adjusted to the same temperature fordetermination of the fraction of formation waterrecovered. This is accomplished by using the chart shownin Fig. 23 or the following equation (ARPS) for NaClsolutions:

for OF

for OC

(7)

where

Rl = resistivity at temperature, T,

R2 = resistivity at temperature, T,

T,,T, = temperature, OF or OC as indicatedabove

The fraction of formation water is calculated using thesetemperature-corrected values of resistivity. Formationwater recovery is then determined by multiplying thiscalculated fraction, f, (To), times the total volume ofwater recovery and predicted water cut can be estimatedusing Eq. 5.

Recovery of Small Volumes of Formation FluidOr No Formation Fluid

Recovery of small volumes of formation fluid or no for-mation fluid results when zones are deeply invaded andrecovery is primarily filtrate. This usually occurs in lowerporosity formations. Interpretation under these condi-tions is inconclusive at best.

When deep invasion occurs, flushing of the formationis usually not complete. The recovered sample, eventhough apparently all water, must therefore be examinedvery carefully for traces of hydrocarbons. Any hydro-carbon present, even if only detectable through

fluorescence tests, may be significant. Detection of gasmay also be important provided the gas is free gas andnot solution gas associated with formation water (see thefollowing section). As a rule of thumb, if less thanlo-15% of the water recovered is formation water andonly a small volume or trace of oil is present, the forma-tion is predicted to produce water-free. A high water cutwould be predicted if larger amounts (>15%) of forma-tion water are recovered.

Easy recovery of filtrate is indicative of a permeable for-mation. This factor, coupled with indications ofhydrocarbons from openhole logs (even though onlyfiltrate is recovered in the sample tank) may indicate thezone to be a candidate for well completion. The FMTmeasurements verify a permeable zone which can be pro-ductive if hydrocarbons (indicated from other informa-tion) are present.

Technique For Various Size Recoveries

The following method analyzes the recovered fluids byconverting the surface-measured volumes to downholeconditions. It is presumed that the sample which enteredthe tool at downhole conditions is representative of an-ticipated production. This interpretive approach at-tempts to break out the formation water, oil, and free gas,if any, at formation pressure and temperature. Note,however, that small amounts of formation waterrecovery (<IO-15%) may be caused by mixing of con-nate and flushing waters and such small formation waterrecovery may not indicate water cut.

A schematic showing the various fluids entering the sam-ple chamber at downhole conditions is given as Fig. 32.When the sample is analyzed at the surface, the characterof the sample is altered due to gases coming out of solu-tion with the formation water and oil. Consequently, thegas volume recorded is greater than any free gas whichmay be present. Furthermore, the loss of gas from solu-tion may cause some change in the volume of oil andwater from downhole to surface conditions.

The analysis is based on the following steps:

Step 1: Computation of Filtrate Volume

To compute the filtrate volume, first determine the frac-tion of formation water, f, (vo), in the total waterrecovery, a technique discussed earlier. The fraction,I-ffw (olo), equals the fraction of filtrate which, whenmultiplied by the recovered water volume, equals thevolume of filtrate.

V, (cm3) = (l-fr,) (Yo) x W, (cm3)

26

FM7 FMTSAMPLE SAMPLE

DOWNHOLE Al’ SURFACE

FIGURE 32Schematic illustrating the differences between downhole andsurface conditions.

The formation water recovered is then

Vfw (cm3) = W, (cm3) - V, (cm3)

Step 2: Computation of Gas Soluble in Water

Figure 33 is a chart which indicates the solubility ofnatural gas in water, measured in ft3 of gas at the sur-face per 1000 cm3 of water. To use this chart, enterdownhole formation pressure and temperature, and readsolubility (S,) as indicated. This solubility value mustbe corrected for water salinity, which is accomplished bydetermining the solubility ratio (SR) from the formationwater resistivity and downhole formation temperatureusing Fig. 34. The volume of solution gas associated withthe formation water recovered is then given by

Vwsg (ft3) = S, x SR x vf;;;;3) (8)

If only water and gas were recovered and the volumecalculated is greater than or equal to the gas recovered,no free gas is present and the zone is 100% water pro-ductive. Otherwise, subtract the value of water solutiongas from the total gas and proceed to step 3.

Step 3: Computation of Gas in Oil

The oil within a formation is either above or below bub-ble point pressure. If the pressure at which the oil is foundis above bubble point pressure, it contains the greatestpossible volume of gas in solution. If bubble pointpressure is less than the pressure of the oil in the forma-tion, then free gas cannot be present in the oil and gaswill not be produced. Hence, free gas is only producedif the oil is below bubble point.

To compute the volume of solution gas evolving from therecovered oil, utilize the chart shown in Fig. 35. Enterdownhole formation pressure at the right side of thechart. Proceed upward to the formation temperature,then left to the OAP1 gravity of the recovered oil. Fromthis juncture, move upward to the gas gravity and thenleft to the gas/oil ratio (GOR). Note that this GOR valueis in ft3 per barrel. To determine the volume of gasliberated from the recovered oil,

vosg et31 =GOR (ft3/bbl) x V, (cm3)

159,000(9)

If this volume of gas is greater than the volume of gasremaining after completion of Step 2, then free gas is pre-sent and free gas production is expected with the oil. Ifthis volume is less than or equal to the gas remainingafter Step 2, then liquid oil production with no free gasis expected downhole.

FLUID FLOW IN POROUS MEDIA

Flow regimes in reservoir rocks are either steady state ornonsteady state. In steady-state regimes, flow rates andpressures at any level will adjust instantly to a change inflow rate and pressure at another point in the flowregime. When readjustment time is short, the flowregime may also be assumed to be steady state.

The snorkel probe of the FMT has a theoretical intakeflow response which is similar to a spherical flow model(Fig. 36). The theory implies a situation analogous to awell and pipe string penetrating an infinitely thickporous stratigraphic unit.

Four different types of flow geometries are of interestin the analysis of wireline tests. In addition to sphericalflow, linear, radial, and hemispherical flow patterns arealso considered (Fig. 36). In linear flow, the lines of flowdistribution are parallel and the cross-sectional area ex-posed to the flow is constant. In radial flow, the flow pat-terns converge two-dimensionally to a central point, e. g.,the borehole. In spherical flow, the flow patterns con-verge three-dimensionally toward a central point,whereas in hemispherical flow, the flow patterns con-

27

.18

.16

a .09

5

s .08

&c

.07

3;i3i3

.06

2‘05 1000

%I500 psi

100 150 200 250 300 330TEMPERATURE (OF)

Solubility of natural gas in water vs. temperature and pressure.

verge three-dimensionally from one side toward a com-mon center.

Darcy’s Law

In 1856, a French engineer named Henry Darcy (Henrid’Arcy) performed tests on water filters providing theengineering profession with a method to measure andstudy the ease of fluid flow through porous rock. Dar&sLaw of fluid flow states, “the rate of flow through a givenrock varies directly according to some numerical quan-tity and the pressure applied, and varies inversely accor-ding to the viscosity of the fluid flowing!’ The numerical

quantity is the permeability (k) and is measured in dar-cies. Reservoir rocks seldom have permeabilities as greatas 1.0 darcy, therefore the usual measure of rockpermeability is in millidarcies (md), although reservoirswith several darcies of permeability do exist.

Wide variations in rock permeability frequently existboth horizontally and vertically. Permeability may oc-casionally change by a drastic amount over a shortdistance in reservoir rock that otherwise appears to beuniform. Horizontal permeability, which is measuredparallel to bedding planes, usually exceeds verticalpermeability.

28

’ ““111111

I-I

tl I I I I I IllIll

0.6 250’I0.5 200’

1:0.4 150”I

FIGURE 34

0.1 5

1

.04 .05 .06 .07 .08.09 .l .2 .3 .4 .5 .6

RESISTIVITY OF RECOVERED FLUIDS(AT 70°F)

.7 .8 .9 1.0

Salinity correction for gas solubility in water.

BUBBLE POINT PRESSURE

FIGURE 35Determination of gas/oil ratio at bubble point.

29

\I/-.-‘I’

HEMISPHERICAL FLOW SPHERICAL FLOW

FMT FLOW (QUASI-SPHERICAL)

FIGURE 36Types of flow conditions

The magnitude of rock permeability plays an importantrole in the ability of a reservoir to produce into awellbore. This is illustrated by the fact that a reservoirrock 10 ft thick having 1.0 darcy of effective permeability(k,) will permit 150 barrels of oil per day (BOPD) to flowinto a wellbore if the pressure in the well is 10 psi belowthe reservoir pressures.

In practical oilfield units, Darcy’s Law is expressed as:

(10)

where

9 = flow rate, BPDk = permeability, darcyA = cross sectional area, ft2

:p/dL= viscosity, cp= pressure gradient, psi/ft

Formation permeability may be inferred from FMTpressure records. The basis of such permeabilityestimates is the pressure drawdown and subsequentbuildup phases discussed earlier under FMT operation.

PRESSURE DRAWDOWN -PERMEABILITY ESTIMATE

Permeability can be inferred from the flowing drawdownphase of the FMT pretest. The calculated permeability

is based upon a quasi-spherical flow model, since theprobe size is small relative to reservoir dimensions (Fig.36). Flow is assumed to be steady state. Fluid entry intothe tool is taken as that entering a sphere having adiameter equal to that of the probe.

The flow model factor (C) is typically considered a con-stant, although in reality it may vary from 0.5 for per-fectly spherical flow to 1.0 for hemispherical flow.Several methods have been proposed to derive a quan-titatively precise value for the C factor; however, thederivations are usually based on controlled conditionsin the laboratory, a situation unlike the borehole environ-ment. For practical oilfield use, the C factor for quasi-spherical flow in an 8-in. borehole is approximated by0.75. It should be kept in mind that flow into the snorkelprobe is through a flat disc (not a hemisphere).

A transform is necessary to convert the variousmeasurements in Eq. 11 into common units. The probediameter transform is recommended to avoid the con-fusion of radius vs. diameter, a step which quite oftencreates calculation errors for the user. Considering theabove, the simplified equation to estimate permeabilityfrom pressure drawdown for FMT work is:

k, = 1842 x C x

where

k = permeability (drawdown), md

9 = flow rate, cm3/sec

c1 = viscosity of fluid, cp(should be temperature corrected)

AP = pressure drawdown, psi

C = flow model factor

d = diameter-of snorkel orifice, in.

(11)

The flow rate (q) is determined from the pretest pressurerecord and is equal to the pretest fluid volume (usually10 cm3) divided by the time of flow as indicated on thepretest record. The pressure drawdown (Ap) is the dif-ference between the flowing pressure during the latestages of drawdown and the shut-in formation pressure.Hydrostatic pressure is not relevent to the calculation.The FMT standard probe has an orifice diameter of0.562in.; however, other sizes are occasionally used. Thisorifice diameter is the avenue through which fluid flowsfrom the formation into the tool. The viscosity 01) is thatof the fluid flowing from the formation into the tool dur-ing the pretest. This fluid is usually mud filtrate, typicallysalt water, and viscosity is commonly assumed at 0.5 cp.More accurate viscosity should be determined by the

30

chart in Fig. 24, which illustrates the effects oftemperature and salinity on the filtrate viscosity

It is of utmost importance that thepretest chamber size,probe diameter, downhole temperature, and salinity ofthemudfiltrate berecorded on thelogheading. The in-formation is not only necessary for the foregoing com-putations, but is important for a review of the data at alater time.

The drawdown model for permeability estimation is avery microscopic indicator because the model is affectedby the formation within a very few (2 or 3) centimetersfrom the probe. The inferred permeability is a relativepermeability to the filtrate, which means it is not usual-ly representative of absolute permeability Drawdownpermeabilityestimatesare usuallypessimisticsince theyare calculated from data related to a region near thewellbore where formation damage is likely to occur.Another reason for low permeability numbers is that thespherical flow characteristics may be impaired becauseof shale barriers or reductions in vertical permeability.Reduction of permeability near the borehole is oftenreferred to as “skin effect”, i. e., the zone of reducedpermeability being the “skin!’ An example of pretestdrawdown permeability computation is provided in Ap-pendix A.

Formation permeability may be inferred from either thepretest pressure record or the pressure record from a sam-ple test. The equation for permeability evaluation is iden-tical for both records but the larger volumetrics of thesample test result in improved calculations.

The resolution on pressure recording during sample testdrawdown in high-permeability formations is thereforeimproved substantially. If permeability is 1000 md, thepressure drawdown during pretest would-be about 1 psiwhereas it would take about 7 seconds to fill a 2.75-gallonsample chamber.

PRESSURE BUILDUP -PERMEABILITY ESTIMATE

A number of methods have been described (by variousauthors) to determine an estimate of undamaged forma-tion permeability by pressure buildup analysis. Drillstemtests (DST) and routine production tests assume flowduring buildup to be radial, and reasonably so. Theassumption of radial flow is also valid for wireline testsmade in very thin beds. However, the FMT buildup pat-tern in thick beds is as follows:

(1) The immediate buildup after drawdown (fill up)is one of linear flow. This very short time spanis difficult to observe on FMT buildup plots.

(2) The next stage is quasi-spherical. The time dura-tion for this quasi-spherical buildup period is afunction of bed thickness and formationanisotropy. Vertical permeability (k,) can becomputed from the information derived here.

(3) Radial flow is the tendency at the later stages ofpressure buildup. The flow patterns have a cylin-drical symmetry around the borehole axis.

Spherical Pressure Buildup -Permeability and Formation Pressure

Following the flowing or drawdown period of the pretest,the flow entering the tool stops and a pressure buildupis observed to occur. At the time the flow ceases, apressure gradient exists between the FMT snorkel andthe undisturbed virgin formation. Following drawdown,flow continues toward the FMT probe, causing a recom-pression of the fluid in the vicinity of the probe and arise in pressure back to the original formation pressure.The disturbance propagates spherically into the forma-tion as shown in Fig. 37 and the pressure profile withinthe formation varies during the buildup period as il-lustrated in Fig. 38.

Since the pressure disturbance associated with thebuildup initially propagates spherically, the pressuredetected at the FMT probe can be modelled during thisperiod with the spherical buildup equation:

FIGURE 37Schematic illustrating spherical flow.

31

FIGURE 38Variation of the pressure profile during the buildup period.

Pws = pi - 8.0 X lo4 ~

0S

where

Pi = formation pressure, psi

Pws = pressure at probe after shut in, psi

9 = flow rate during drawdown, cm3/sec

P = viscosity of formation fluid, cp

ct = compressibility of formation fluid, psi-’

kS = permeability of formation (spherical),md

+ = porosity of formation, fraction

t = length of pretest flowing time, set

At = time elapsed after shut in, set

While the foregoing equation assumes an infinite,homogeneous formation, it may be applied before anyboundaries are encountered by the pressure disturbance.Furthermore, theprevious equation may beapplied onlyafter the pressure has built up sufficiently. The dif-ference, pi - pws, must be lo-20% percent of thedrawdown pressure (Ap) encountered during the flow-ing portion of the pretest to assure its validity. A furtherimportant consideration which may raise questionsregarding the valid application of this equation is thepresence of formation boundaries or thin shale barrierswhich effectively block thepropagation of thesphericalpressure disturbance. Their presence may cause thebuildup to become cylindrical rather than spherical incharacter.

To evaluate permeability, the foregoing equation may bewritten:

A plot of

on linear-linear graph paper produces a straight line dur-ing the spherical buildup period as illustrated in Fig. 39.The slope, m,, in psi/set may be used to evaluate perme-ability, provided ct and + are known, by the followingequations:

FIGURE 39A linear-linear plot produces astraight line with slope m dur-ing the spherical buildup period.

32

m, = 8.0 x lo4 (12)

or

(13)

An example computation using this spherical builduptechnique is given in Appendix A.

The computed permeability, k,, is the permeability toflow in a spherical flow regime. This permeability maybe related to the horizontal (k,) and vertical (k,)permeability by the equation

k,3 = kH2 x k, (14)

In most conditions, the vertical permeability is less thanthe horizontal permeability and the degree of anisotropy,

kani, can be defined as

kani = k,/k, (15)

A graphical means of determining k, and k, from theevaluation of k, and degree of anisotropy, if known, isshown in Fig. 40. For example, if k, is determined to be10 md and the anisotropy is 0.1, the horizontalpermeability is found to be about 21.5 md. If theanisotropy is unknown, it may be evaluated fromhorizontal permeability information determined fromthe cylindrical buildup discussed later.

The measurement of permeability from sphericalbuildup, unlike drawdown, is affected by informationcoming from deeper within the formation. This depthis affected by FMT pressure gauge resolution as well asformation and fluid-related parameters. The depth ofinvestigation (cm) or sphere ofinfluence is typically onthe order of a few feet beyond the probe. An equationwhich relates the depth into the formation which most

1000

10

kH (md)

FIGURE 40Chart for determination of kH and k, from the evaluation of k, and degree of anisotropy (d), if known.

33

affects the buildup is given as

‘inv At0.3 (t + At)0.2 (16)

where

rinv = depth into formation affected by buildup, cm

The plot of the pressure data during buildup, shown inFig. 39, may be used to obtain a better estimate of for-mation pressure. As pressure becomes great relative tot, the difference

approaches zero and FMT pressure approaches true for-mation pressure. Formation pressure may be estimatedby extrapolation of the straight line plot of the sphericalbuildup curve to the

point, even though the observed pressure continues toincrease at the time the pretest is terminated.

Cylindrical Buildup -Permeability and Formation Pressure

At some point after shut in, the spherical pressure distur-bance will likely encounter the bed boundaries or otherlocal permeability barriers as shown in Fig. 41. As aresult, the configuration of the disturbance undergoesa transition from a spherical to a cylindrical shape.Following this transition, analysis of pressure builduprelies on the cylindrical pressure buildup equation.Assuming an infinite reservoir, the following equationdescribes the cylindrical pressure buildup at the FMTprobe:

(pws - pi)= - 88.4 ($) log,o(fg) (17)

where

Pi = formation pressure, psi

Pws = pressure at probe after shut in, psi

9 = flow rate during drawdown, cm3/sec

IJ = viscosity of formation fluid, cp

kc = formation (cylindrical) horizontalpermeability, md

h = distance between impermeable barriers, ft

t = length of pretest flowing time, set

At = time elapsed after shut in, set

As with the spherical equation, the cylindrical equationapplies as long as no discon tin uities in the formation areencountered and only during the late stages of buildup.

For evaluation of permeability, Eq. 17 is rewritten to theform

(Pi - Pws) = m, (18)

Schematic illustrating the transition from spherical to cylin-drical flow.

When buildup data is plotted on semi-logarithmic graphpaper, with pi - pwS as the linear y-axis and t + At/Aton a logarithmic x-axis as shown on Fig. 42, the plot isknown as a Horner Plot. This data plots linearly on theHorner Plot during the cylindrical buildup period. Theslope of that plot (m,) is measured in psi/cycle andpermeability (k,) may be computed from the followingequation:

(19)

The permeability (k,) is a measurement of horizontalpermeability beyond the skin-damaged zone, since thedepth of investigation extends several feet from thewellbore.

Where the FMT probe has been set in clean thin zones,the parameter his the zone or bed thickness. In more ex-

34

CYLINDRICAL BUILDUP PLOT(HORNER PLOT)

CYLINDRICALPRESSURE

4500 -DISTURBANCE

%

3 DISTURBANCEY

10 8 6 5 4 3 2 . 1

t + AtAt

IGURE 42Semi-logarithmic plot of buildup data provides a straight lineresponse of slope m for cylindrical buildup data.

tensive zones, especially those where numerous semi-extensive thin shale layers exist as illustrated in Fig. 43,the proper value of h to be used in computations is notreadily apparent. A value of h = 0.5 ft is frequentlytaken and appears to provide reasonable results in manyinstances. However, for best results in development wells,the parameter h should be adjusted to provide the bestmatch between k, and permeability from core or otherreliable data taken in the same reservoir.

FIGURE 43Numerous discontinous shale stringers pose a problem toa proper calculation of bed thickness (h).

To evaluate the formation pressure (p*) from the cylin-drical buildup data, the linear portion of the Horner Plotmust be extrapolated to the point where t + At/Atapproaches 1. This value of p* is valid as long as thepressure pulse has not encountered any discontinuity andthe pressure pulse is cylindrical in character as modelled.This technique is illustrated in Fig. 42. An example com-putation demonstrating the cylindrical buildup methodis carried out in Appendix A.

COMPARISON OF PERMEABILITY ESTIMATES

The three approaches for estimating permeability(drawdown, spherical buildup, and cylindrical buildup)often produce three different permeability values, andat first glance, a very poor agreement with each other.It becomes necessary to review the difference betweenthe threeindividualpermeabilitymeasurements and toestablish thepotential utility of each. The schematic ofFig. 44 illustrates the regions of investigation for eachmethod, assuming a fairly thin zone in which eachbuildup region fully develops for some period of time.

The drawdown test drains 10 cm3 of fluid from the for-mation. During this pretest the converging character ofthe spherical flow assures that the greatest weight is givento the permeability where the greater pressure drop oc-curs, within about 2 cm from the probe. While the flowis spherical in character, it does not converge to a pointsince the flow enters the full diameter of the probeorifice. The computed drawdown permeability willtherefore be closer to the cylindrical permeability resultthan thespherical buildup value. Of primary considera-tion is the fact that this region of formation near theborehole has been invaded and flushed by drilling fluids.As a result, permeability estimates are adversely affecteddue to mud filter cake, skin damage by clay particlehydration, mud solid infiltration, compaction, andrelative permeability effects.

FIGURE 44Schematic illustrating regions of investigation.

35

Assuming no damage or plugging of the pore structure,the effect of relative permeability is illustrated in Fig. 45.kabs is the absolute permeability to one phase (gas, oil,or water) when only that one phase is present, and ef-fective permeability is the permeabiity to an individualfluid phase when two or more phases are present. Ab-solute permeability is the maximum permeability toflow. Effective permeability (k,) to any single phase isalways less than kabs when two phases are present sincethe pore space is occupied to some extent by the secondphase. The flow channels are therefore somewhatrestricted and effective permeability is less than absolutepermeability. The term relative permeability (k,) is simp-ly a fraction of k,, and is a convenient factor to use tocompute k,. For example, if water and oil are present inand fill the pore volume,

keo = kro x kabs (20)

and

kew = krw x kabs (21)

The flushed zone water saturation (S,,) for a water-wetoil reservoir may exhibit flushed zone water saturationsof 0.75 to 0.95. The relative permeability curve to wateris steeply declining over this range from k,, = 1.0 (Fig.45). Therefore the observed permeability is significant-ly reduced even though only water flows into the FMTtool.

FIGURE 45Effects of relative permeability.

If the mud is water based, a water zone will exhibit k,,as its permeability. In an oil-wet reservoir, an oil zoneflushed with oil-based mud filtrate will exhibit k,,. Thisassumes no mechanical damage in either of the abovecases. However, the mud may have been viscosified, gell-ed, or otherwise treated with soluble additives, and sothe character of its filtrate may be quite different frombrine.

Permeabilities determined from buildup pressures areresponsive to formation characteristics deeper into theformation, i. e., the fluid which is mobile deep in the for-mation is the fluid which affects the test. In an oil zone,the permeability measured is less than the absolutepermeability if the formation is water wet. This measurepoint is illustrated on the relative permeability curves ofFig. 45. The spherical model may be affected, in part,by the invaded zone since the spherical disturbance pro-pagates in all directions. It is apparent from Fig. 44 thatfor spherical buildup to measure effective permeabilityto oil deeper in the formation, the distance between bar-riers must be great relative to the invasion depth. If thisis not the case, the transition to cylindrical buildup beginsbefore the spherical disturbance propagates appreciablybeyond the invaded zone. Unless depth of invasion is verylarge, thecylindricalmodelis best suited to evaluate theeffectivepermeability to hydrocarbon, subject to propervalues used for the formation thickness (h).

Spherical permeability, as discussed earlier, may berelated to horizontal permeability if the anisotropy(k/k,) is known. As a result, the permeability in-dicated by spherical buildup may be considerably lowerthan actual horizontalpermeability. Core analysis mayalso indicate anisotropy exists in the horizontal directionwith the difference between the maximum and minimumfrequently exceeding a factor of 10 or more.

A comparison of permeability computed by drawdownwith that from cylindrical buildup is illustrated in Fig.46. While the correlation appears good, the differencesmay be resolved or used to resolve such unknowns as ef-fective distance between permeability barriers, degree ofskin damage, anisotropy, etc., especially when combinedwith spherical buildup and/or core data.

FIGURE 46Comparison of cylindrical buildup permeability computationsto drawdown permeability computations.

36

BED THICKNESSDEFINITION DURING BUILDUP

During the buildup phase impermeable boundaries maybe encountered by the spherical pressure pulse. As aresult, the buildup data begins to deviate upward abovethe linear portion of the spherical buildup plot, as il-lustrated in Fig. 47. The point of deviation signals thetransition from spherical to cylindrical buildup. Thedistance from the probe to this impermeable bed may beestimated by the following equation:

(22)

where

h’

V

Pi

P*

= distance to impermeable layer, cm

= volume of flow, cm3

= initial formation pressure, psi

= pressure determined from extrapolationof linear spherical buildup, psi

+ = formation porosity, %

ct = formation total compressibility

kani = anisotropy, k/k,

FIGURE 47Upward deviation of spherical pressure buildup (upper rightcorner) signals the transition from spherical to cylindricalbuildup.

An alternative method of determining h has beenreported. This technique is based on the time, t*, afterthe beginning of flow, from which the actual pressurebuildup deviates from its linear character during thespherical phase. The equation to determine h on thisbasis is

x 1O-4 (23)

where

t* = time after beginning of flow whenbuildup deviates from straight linespherical plot, set

At* = t* - flowing time, set

SUBSURFACE PRESSURE REGIMES

Hydrostatic Pressure

Hydrostatic pressure is created by the unit weight andvertical height of a fluid column. The size and shape ofthis fluid column have no effect on the magnitude of thispressure. Hydrostatic pressure, pHY, equals themathematical product of the average fluid density andits vertical height such as

pHY=PxGxZ (24)

where

P = average density,

G = gravitational constant

Z = height of the column

In terms of drilling and formation tester operations,

p,,(psi) = C x MW x Z

where

(25)

Z = vertical height of fluid column in feet,

MW = fluid density or mud weight in lb/gal(lb/gal, ppg) or lb/ft3

C = conversion constant (C = 0.052 if MWin lb/gal and C = 0.00695 if MW inlb/ft3).

In the metric system, Eq. 25 becomes

pHY = 0.098 x MW x Z (26)

where Z, the vertical fluid column, is in meters and MW,the mud weight, is in g/cm3.

37

The hydrostatic pressure gradient is affected by the con-centration of dissolved solids (i. e., salts) and gases in thefluid column and different or varying temperature gra-dients. In other words, an increase in dissolved solids(i. e., higher salt concentration) tends to increase the nor-mal pressure gradient whereas increasing amounts ofgases in solution and higher temperatures woulddecrease the normal hydrostatic pressure gradient. Forexample, a pressure gradient of 0.465 psi/ft assumes awater salinity of 80,000 parts per million (ppm) NaCl ata temperature of 77’F (25OC).

Typical average hydrostatic gradients which may be en-countered during drilling for oil and gas are shownbelow:

Hydrostatic Equivalent Total BasinGradient Mud Wt. (ppg) Chlorides (ppm) Location

0.433 8.33 fresh Rockywater Mountains,

Beaufort,Brunei,

0.442 8.5 20,000 Malay,Sverdrup,N. Slope inAlaska(most ofworld’s basins)

0.452 8.7 40,000

0.465 9 . 0 80 ,000

0 .478 9 . 2 95 ,000

North Sea,Delaware (olderportion -Pre Penn.)

Gulf Coast

Portions ofGulf Coast

In general then, the hydrostatic pressure gradient (gfp)can be defined in psi/ft from:

gfp = 0.433 x yw (27)

where yw is the specific gravity of a representative col-umn of water.

Overburden Pressure

This pressure originates from the combined weight of theformation matrix (rock) and the fluids (water, oil, gas)in the pore space overlying the formation of interest.Mathematically, the overburden pressure (p,) can be ex-pressed as:

Weight (Rock Matrix + Fluid)PO = (28)

Area

- +) prna + +Pf G1

where

z =

+ =

P ma =

Pf =

Generally,

vertical height of geologic column

porosity of formation expressed as afraction

density of rock matrix

density of fluid

it is assumed that overburden pressure in-creases uniformly with depth. For example, average Ter-tiary deposits on the U. S. Gulf Coast and elsewhere ex-ert an overburden pressure gradient of 1.0 psi/ft of depth.This corresponds to a force exerted by a formation withan average bulk density of 2.31 g/cm3. Experience alsoindicates that the probable maximum overburden gra-dient in clastic rocks may be as high as 1.35 psi/ft.

Worldwide observations over the last few years haveresulted in the concept of a varying overburden gradientfor fracture pressure gradient predictions used in drill-ing and completion operations.

Formation Pressure

Formation pressure (pf) is the pressure acting upon thefluids (formation water, oil, gas) in the pore space of theformation. Normal formation pressures in any geologicsetting will equal the hydrostatic head (i. e., hydrostaticpressure) of water from the surface to the subsurface for-mation. Abnormal formation pressures, by definition,are then characterized by any departure from the nor-mal trend line.

Formation pressures exceeding hydrostatic pressure(pf > p,,) in a specific geologic environment are defin-ed as abnormally high formation pressures (super-pressures), whereas formation pressures less thanhydrostatic are called subnormal (subpressures).

Figure 48 and Eq. 27 both illustrate how these subsur-face pressures and stress concepts are related:

PO = Pf

where

PO =

Pf =

u =

+C7 (29)

overburden pressure (total vertical stress,lithostatic pressure)

formation pressure (pore fluid pressure,pore pressure)

grain-to-grain presssure (matrix stress,effective stress, vertical rock-framestress).

38

FIGURE 48Subsurface pressure concepts.

In normal pressure environments (pr = pHY) the matrixstress supports the overburden load due to grain-to-graincontacts. Any reduction in this direct grain-to-grainstress (o-0) will cause the pore fluid to support part ofthe overburden, the result being abnormal formationpressures (pf > p,,). In other words, the overburdenmay effectively be buoyed by high formation pressures.

There are numerous factors that can cause abnormal for-mation pressures such as surpressures and subpressures.Frequently, a combination of several superimposedcauses prevail in a given basin and as such is related tothe stratigraphic, tectonic, and geochemical history ofthe area. This has been discussed in detail (Hawkins,1956).

Generally speaking, any subsurface fluid pressure (pf)is a function of the fluid pressure gradient (gfp) and truevertical depth (D), such as

pf (Psi) = gfp x Dgauge pressure units

(30)

pf @ia) = gfpx D + 15absolute pressure units

(31)

gfr, (psi/ft) = pf (psi)/D (ft) (32)

In subsurface water pressure regimes, the typical average It is also veryimportant that all measured pressure datapressure gradients for fresh and brackish water are 0.433 be evaluated at the true vertical depth (TVD) regardlesspsi/ft and for salt water, 0.465 psi/ft. These values cor- of the borehole drift angle. This is illustrated by the ex-respond to fluid density values of 1.0 g/cm3 and 1.07 ample in Fig. 50, where vertical Well A was drilled tog/cm3. Figure 19 shows water density as a function of 10,000 ft and the measured depth of deviated Well B wassalinity, temperature, and pressure. 12,000 ft, although the true vertical depth of the target

Overpressures are defined by:

pf (Psi 1 = gfp x D + C (33)

Pf @ia) = gfp x D + 15 + C (34)

whereas subpressures (underpressures) are described by

pf (Psi) = gfp x D - C (35)

pf @ia) = gfp x D + 15 - C (36)

Hydrocarbon pressure regimes depart from subsurfacewater regimes in that the densities of oil and/or gas areless than that of water. Consequently, hydrocarbonpressure gradients are smaller, typical values being

Gas Density (g/cm3) Pressure Gradient (psi/ft)

0.25 0.110.18 0.08

Oil Density (g/cm3) Pressure Gradient (psi/ft)

0.85 0.370.80 0.35

where

g/cm3 + 2.31 = psi/ft (37)

APPLICATIONS OFFMT PRESSURE MEASUREMENTS

The most important feature of the FMT is its ability toperform pretest pressure measurements with reasonableaccuracy at numerous selected depth intervals. Pretestformation pressures are typically determined followingthe observation of a stable buildup to formation shut-in pressure. It is essential that this formation shut-inpressure reading be taken as long as safely possible afterthe flowing portion of the pretest in order to allow ade-quate time for the pressure to build up and approach theactual formation pressure. A typical formation pressurereading is illustrated in Fig. 49. If the pressure test is ter-minated too early, the formation shut-in pressure readingwill be too low since sufficient buildup did not occur.

Measured Depth vs. True Vertical Depth

39

ANALOG(Psi)

-L-l-2

RECORDED DIGITAL SAMPLING PRESSUREB (Psi)

i

FIGURE 49Adequate time for pressure buildup must be allowed.

40

L

FIGURE 50Vertical borehole vs. measured depths and TVD in directionalboreholes.

zone in Well B was also 10,000 feet. Both wells were drill-ed with similar mud systems, corresponding to ahydrostatic gradient of 0.465 psi/ft, or 4650 psi at TVDin both wells. Serious interpretive errors would haveresulted if measured depth of Well B had been used tocalculate hydrostatic pressure (12,000 x 0.465 = 5580psi), in which case the resultant value would be 930 psitoo high.

Pressure Regimes in Water-Bearing Reservoirs

Subsurface aquifers can have normal (hydrostatic)pressures or they may be either overpressured or under-pressured. If a well penetrates a sequence of permeablewater sands, FMT pretest pressure measurements can beused to identify the normal hydrostatic gradient andlocate those strata which are either overpressured orunderpressured.

The plot of depth vs. formation pressure in Fig. 51 istaken from five FMT pretest pressures: 660 psi at 2000ft, 2325 psi at 5000 ft, 4650 psi at 10,000 ft, 5580 psi at12,000 ft, and 8150 psi at 12,500 ft.

The formation pressure gradient (gf,) for each zone iscalculated as follows:

660 psi Q 2000 ft grr, = 660/2000 = 0.33 psi/ft

2325 psi @ 5000 ft gr, = 2325/5000 = 0.465 psi/ft

4650 psi @ 10,000 ft gr, = 4650/10,000 = 0.465 psi/ft

5580 ps1 @I 12,000 ft gg = 5580/12,000 = 0.465 psi/ft

8150 psi @ 12,500 ft gr, = 8150/12,500 = 0.65 psi/ft

L

FIGURE 51Formation pressure gradient.

The shallow zone is slightly underpressured and thedeepest zone is considerably overpressured. This type ofinformation can be invaluable to drilling plans for off-set wells and in optimizing completion practices.

Supercharging

Formation pressure measurements can be affected by aset of conditions known as supercharging. Supercharg-ingis thenaturalresult oftheradialflowofinvadingmudfiltrateinto the formation during theprocess of buildingup a filter cake over a permeable depth interval, as il-lustrated in Fig. 52. The supercharging effect causes theobserved formation pressure (near the wellbore) to begreater than the actual formation pressure. Supercharg-ing should not be confused with intrinsic formationoverpressures. Two mud-related factors which affect thefiltration rate are (1) the degree of pressure differential(or overpressure) between the mud and the formationand (2) the extent of mud cake buildup and its effec-tiveness in preventing further filtrate fluid loss into theformation. The second factor tends to mitigate the ef-fects of supercharging with time if the zone has adequatepermeability to allow the pressure to bleed off anddissipate. Supercharging can be quite large in very tightformations (< 0.5 md) as illustrated by the data inFig. 53. Plots of pressure vs. depth from several pretestreadings will usually reveal these zones which areanomalous because of supercharging as shown inFig. 53.

41

FIGURE 52Supercharging results from radial flow of the mud filtrate intothe formation during filter cake buildup.

FIGURE 53Large supercharging effects are most common in tightformations.

Selection of Test Intervals

Proper analysis of openholelogs should allow selectionof themorepermeablezones forpressuremeasurements.Good log interpretation practices will help the FMT useravoid testing strata where supercharging is likely to oc-cur. In any case, the higher credibility should be giventhose pressure measurements taken from zones ofhighest permeability. Very long pretests are indicative ofextremely low permeability and likely to be supercharg-ed. Effects of supercharging can be further minimizedby running the FMT service as long as possible after mudcirculation, which would allow for maximum mud cakebuildup and pressure dissipation.

Pressure Gradients and Particular Pressure Regimes

When an adequate number of formation pressuremeasurements are acquired in a borehole, a plot of those

pressures vs. depth (TVD) presents a quantitative pro-file of each individual horizon’s ability to drive its pro-duced fluid to the surface. A typical plot of pressures vs.depth (TVD) compared to bulk volume analysis fromopenhole logs across three potentially productivehydrocarbon zones is shown in Fig. 54. A hydrostaticmud coIumn pressure gradient is also plotted.

Maximum advantage of pretest formation pressure datais attained if the pressures used on the plot are derivedfrom the extrapolation of the appropriatepressure andbuildup plots. As discussed earlier, buildup pressure datais a truer representation of formation fluid pressures,especially when rock permeabilities are low. The presen-tation of the mud column pressuregradient serves as acheck to verify proper tool operation during thedownhole pressure survey.

If a particular stratigraphic unit is relatively thick andundisturbed by prior depletion, a formation pressureprofile across that zone may indicate the type of mov-ingpore fluid. Equation 27 applies in this circumstancejust as it did with hydrostatic gradients.

gfp (psi/ft) = 0.433 x Reservoir Fluid Density (cm3)

Formation water densities generally vary in gradientfrom 0.433 psi/ft (fresh water) to 0.465 psi/ft and greaterfor salty waters. Gas zones generally exhibit gradientsless than 0.1 psi/ft. Liquid hydrocarbons will vary from0.25 to 0.34 psi/ft or greater depending on oil gravity andgas/oil ratio (GOR). A key to gradient (or slope) is givenin the lower right-hand corner of Fig. 54. FMT fluidpressure gradients therefore play an important role inverifying, or identifying, the presence of water, gas, orliquid hydrocarbons in a formation.

Determination of Movable Formation Fluid Densityin Zones with High Connate Water Resistivity

In depth intervals where the reservoir connate waterresistivity (R,) is high, and the traditional Archiemethod of log analysis allows for some uncertainty ofpore fluid type, a crossplot of pHY versus p* is recom-mended. Pressures derived from the Hewlett-Packardgauges should be utilized because of their superiorresolution. With several data points available, a best-fitline or slope can be established. The resultant slope isproportional to the in-situ density of the formation fluid(Pf). Multiplying the slope value by the mud density(P,,d) yields the product Pf. The above assumes statichydraulic equilibrium over the designated depth interval.

Defining Gas/Liquid and Oil/Water Contacts

Pressure gradients derived from FMT data have alsofound significant usage in defining gas/liquid and

42

50 0 0 FMT 1.19 g/cm3

JOB SUMMARY - PRESSURE

\ HYDROSTATIC

OHYDROSTATIC

0 500 1000 1500

PRESSURE (psi)

2000 2500

FIGURE 54Comparison of bulk volume analysis from open hole logs to a typical FMT pressure versus TVD depth plot.

oil/water contacts. The free water level indicated in Fig.55 represents the depth where capillarypressure equalszero. A series of FMT pressure measurements across theoil and water zones were plotted vs. depth. A saturationprofile from log analysis is provided on the right side ofFig. 55 for comparison. Note that the free water levelpoint occurs where the oil gradient and water gradientintersect.

FIGURE 55Comparison of FMT pressure data to a saturation profile fromlog analysis.

An oil or gas reservoir, under virgin conditions, exhibitstwo fluid phases near the wellbore, i. e., mud filtrate andeither oil or gas. Thepressures of the twointerfaces dif-fer because of the effects of capillary pressure. Theoil/water contact lies above the free water level by adistance determined by capillary pressures, grain size,permeability, etc. In the transition zone, the capillarypressure is a function of the wetting phase saturation:

PC = PO - pw

where

PC

PO

PVf

= capillary pressure

= oil capillary pressure

= water capillary pressure

The height (Z) above the free water level is a function ofcapillary pressures, i. e., differences between permeabili-ty, fluid densities, and the rock fluid interfaces. It mightalso be noted that the oil/water contact is indicated asbeing several feet above the free water level in Fig. 55. Theoil/water contact represents the depth where oil satura-tion begins to increase from zero. A transition zone is in-dicated where oil saturation continues to increase untilirreducible water saturation (Si,) is attained. Transitionzones may exist well above both the free water level andthe oil/water contact due to poor vertical permeability,water saturations greater than irreducible, etc. Comple-tion in the transition zone often results in some waterproduction. There are also occasions where “hydrocar-bon shows” are observed in well cuttings and/or cores,but the FMT water gradient verifies that the hydrocar-bon is only present in negligible amounts. Keep in mindthat the pretest shut-in pressures are derived from thecylindrical portion of the buildup data, affected by theformation fluids some distance from the borehole.

Zone Isolation or Communication

When multiple potentially productive zones are en-countered in the same borehole, it is possible to use I?MTpressure data to determine whether or not hydrostaticcommunication exists between the zones. As shown inFig. 56, connected zones differ from each other by theamount ofhydrostaticpressure head between thezones.When it is considered that the communicating reservoirfluids may be some distance from the wellbore, it maybe uncertain whether the connecting fluid is water, oil,or a mixture of the two.

If a hydrostatic envelope is drawn from a pressure plotin each zone as shown by the arcs A, B, and C for pointsa, b, and c of Fig. 56, any overlap or contact of theenvelopes (shaded areas) corresponds to a depth at whichthe zones may be connected. The common overlapregion defines a point (depth) at which the apparentseparate zones may have a common pressure from whicha hydrostatic gradient yielding the individual zone’spressures is possible. The overlap only indicates thepossibility of communication.

Zones A and B of Fig. 56 overlap below point a andabove point b and have a common contact along the oilgradient line between points a and b. The close proximityof the overlapping envelopes would lead one to strong-ly suspect vertical communication between zones A andB.

The probability of zone C being connected to either zoneA or zone B is less likely. Note that the overlap ofhydrostatic gradient from zone C overlaps with the oilgradient from zones A and B at point d in Fig. 56. Point

44

0 FMT FORMATION PRESSURE

O I L G R A D I E N T - 1

POSSIBLE POINT OFCOMMON PRESSURE TO

ZONES A AND C, BAND C

FIGURE 56FMT pressure data is useful to determine whether or not hydrostatic communication exists between multiple zones.

45

d is far removed from the three zones underconsideration.

Furthermore, if the oil/water contact occurs in zone Bas indicated by bulk volume log analysis, pressure wouldonly move down to zone C along the water gradient. Itmust therefore be concluded that zones B and C are notin communication.

It is possible that zones A and C are connected, althoughexcluding zone B from such a vertical communicationwould appear unlikely. However, no water contact isnoted in zone A so a remote possibility of connecting tozone C must be considered.

It is extremely important that interpretations, such asthat given in Fig. 56, be made from pressure data takenfrom virgin reservoirs, i.e, where production has not yetbegun. Reservoir depletion from offset wells causesdramatic pressure changes in reservoirs.

In the example in Fig. 57, the pressures in zone 1 aresignificantly lower than the pressures in zone 2.Although the oil gradients in zone 1 and the upper por-tion of zone 2 are similar, the two zones are not con-nected because of the significant difference in thehydrostatic gradient.

Zones 3 and 4 of Fig. 57 are in all likelihood part of thesame reservoir, as indicated by the schematic. FMT

FIGURE 57Determining zone isolation from FMT pressure data.

pressure data therefore plays an important role in iden-tifyingzoneisolation or communication between zones.The scenario in the figure could be enhanced with well-to-well log correlations, comparison to seismic inter-pretations, and detailed stratigraphic analysis from dipdata, curve shape studies, and other electrofaciesfingerprints.

Impermeable layers within a reservoir can also be iden-tified from the pretest pressure recordings. The recogni-tion of non-permeable streaks is especially important inmanycarbonatereservoirs where the better permeabili-ty and higher formation pressures are fundamental tohydrocarbon production.

Determination of Oil/Water ContactBelow Total Depth of the Borehole

FMT pressure test data can be combined with theanalysis of well logs and used to calculate an approx-imate depth of the oil/water contact even though theborehole has not penetrated the contact. Such informa-tion is obviouslyimportant to the developmentgeologistin order toproperlyselect the geographical location foroffset wells. It also provides the reservoir engineer withneeded data for estimating reserves.

The FMT data from the well in Fig. 58 showed a forma-tion pressure of 3280 psi at 7000 feet. The recovered oilhas an ‘API gravity of 24’ and agas/oil ratio (GOR) of200. Using the chart in Fig. 20, a GOR of 200 exhibitsa density of 0.85 g/cm3 (or a pressure gradient of 0.37psi/ft) and assuming a hydrostatic gradient of 0.465psi/ft for water:

p0 (psi) = gfpo x D + C (oil)

3280 = 0.37 x 7000 + c

or, C (oil) = 3280 - 2590 = 690

and

pW (psi)

3280

or, C (water)

= gf,, x D + C (water)

= 0.465 x 7000 + C (water)

= 3280 - 3255 = 15

Knowing that p0 = pW (p, = 0) at free water level, then

0.37 x D + 690 = 0.465 x D + 15

0.095D = 675

D = 7105 ft, the estimated depthof free water level

46

3280 OS, li, 700011

DEEPEST POSSIBLEOIL/WATER CONTACT

FIGURE 58Oil/water contact below TD - oil reservoir.

Reservoir and Zonal Depletion

When several wells in a reservoir are produced, newlydrilled offset in-field wells usually detect changes in theformation pressure profile as a result of production. Ifnumerous thin zones are produced, the pressure changesin the offset wells provide a clue as to which zones arebeing depleted. When a single thick sand is produced,the changes in the pressure profile of the reservoir froma linear fluid gradient indicate that certain parts of thereservoir are preferentially produced over others asshown in Fig. 59. This is often due to higher permeabilityreservoir sections being depleted more rapidly while thetighter sections maintain their pressure, or to permeabili-ty barriers separating various portions of the interval.Detection of these production anomalies may indicatethat some changes in the completion practice should bemade in order to optimally produce the reservoir dur-ing primary production.

Monitor Injection Program in In-Field Wells

A closely related application is to monitor reservoirpressure from newly drilled in-field wells during secon-dary recovery operations. This technique verifies the ef-fectiveness of the injection wells and the pressuremaintenance program. A pressure contour mapdeveloped from wireline formation tester data is shownin Fig. 60. It is apparent that the high pressure ridges lineup with the bank of injection wells.

Fracture Detection

Naturally fractured formations, where interconnectedfractures form a high permeability network amongotherwise low permeability blocks may, under the pro-

FIGURE 59Pressure profiles can illustrate the parts of a reservoir whichshow a preference to produce.

per circumstances, be detectable by the FMT. Somematrix blocks are initially water saturated but later ingeologic time the fracture permeability is filled with li-quid hydrocarbon. The matrix blocks become partiallysaturated with the hydrocarbon. If the blocks are largeenough, the lower portion of the block remains watersaturated until the pressure differential due to hydrostaticand capillary effects is sufficient to displace the water.Above this point, hydrocarbon saturation increasestoward irreducible water saturation, which is achievedonly in sufficiently large blocks. The FMT response isshown in Fig. 61. The apparent oil gradient correspondsto the overall gradient of the fluid in the fracture, whiledeviations toward lower pressure are indicated where theFMT was set on the water-saturated portion of the block.The FMT pretest buildup plot deviates from building upto a stationary pressure, indicating that the pressure tran-sient was controlled by the pressure within the fracturevolume as shown in Fig. 62.

An estimate of fracture block size can be made on thebasis of the deviation from spherical buildup as il-lustrated in Fig. 62. The following two equations havebeen reported, where h is the block size in cubiccentimeters.

Based on Pressure Deviation

The average block size may be estimated with the follow-ing quadratic equation:

I.

\* l \

..

..

726

.

.

.

2517

f

INJE

CT

ION

W

EL

L

l

PR

OD

UC

ING

W

ELL

@

WE

LL

TE

ST

ED

W

ITH

FO

RM

AT

ION

TE

ST

ER

/-

WE

LL

NU

MB

ER

220

AV

ER

AG

E M

EA

SU

RE

D21

84 -

PR

ES

SU

RE

FIG

UR

E 6

0P

ress

ure

cont

our

map

de

velo

ped

from

w

irelin

e

test

da

ta.

FIGURE 61FMT pretest pressures through a series of matrix blocks,some of which contain a permeable fracture network.

(2301 x D) h, + (C - 115.1 x D) h - 0.3 = 0 (39)

where

C

and

D

hb

P*

Pi

+

Ct

cl

t

= average block size, cm3

= extrapolated pressure, psi

= reservoir pressure, psi

= porosity, %

= matrix fluid compressibility, psi ml

= flow rate, cm3/sec

= flow time, set

Based on Time of Deviation

(41)

An alternate approach based upon time of deviation isreported as follows:

(42)

where

t* = total time elapsed between beginning offlow to deviation from linear buildup, set

At* = t* - length of flowing time, set

l-

FIGURE 62FMT pressure buildup in a fractured reservoir.

Extremely Tight Formations

If formation permeability is extremely low, the pretestpiston will draw a near-vacuum as the formation is essen-tially drawn down by its full pressure. The FMTdrawdown is force limited to 7500 psi below hydrostaticpressure. Once the pretest piston completes its stroke, theformation continues to feed fluid into the pretest systemuntil 10 cm3 is accumulated. (In geographical areaswhere such formations are common a 5-cm3 plug isoften used, limiting the pretest to a 5-cm3 volume.)

The lowest possible pressure during a tight pretest is thevapor pressure of the fluid (usually mud filtrate) fillingthe pretest system. Vapor pressure is a function oftemperature, e. g., vapor pressure for water at 300’F is67 psi. Any drawdown pressure records below the vaporpressure should therefore be caused by temperature ef-fect on the pressure gauge and deviation from the gaugecalibration. Newer FMT tools and current software cor-rect thepressuregaugefor temperature effect. If suffi-cient time is allowed, the pressure will slowly build upto a shut-in formation pressure.

Grain Size Effects

Studies of grain size and sorting have shown that a cor-relation exists to permeability and particular en-vironments of deposition. Studies of log curve shapesand their comparison to full core petrographic analysishave shown that characteristic features of fining upward,coarsening upward, etc. can often provide clues to helpidentifyparticular sedimentary environments. It is alsogenerally accepted that grain size and sorting affect thenature of permeability, with finer grain and/or poorersorting correlating to lower permeability.

49

A profile of numerous FMT-derived permeabilities vs.depth across a particular formation might also providesuch an inference to the original environment of deposi-tion. In a deltaic distributary mouth bar, for example,a permeability profile would be expected to show an in-crease in permeability upward vs. depth, whereas thespontaneous potential, gamma ray, or other log curvessensitive to grain size change would tend to show acoarsening upward trend. This idealized comparison isshown in Fig. 63.

FIGURE 63Grain size studies from logs can be compared to pretestpressure permeability profiles.

FMT Pulse Testing

Pulse testing techniques are widely used to determine thereservoir properties between the adjacent wells involv-ed in the test. Test procedures involve one pulsing well(production or injection) and an observation well toobserve pressure response. In order to utilize the FMT,the observation well must be uncased across the reser-voir being tested.

A series of flow disturbances are created in the pulsingwell by alternating production (or injection) with a shut-in period. The pressure response to those pulses ismeasured in the observation well utilizing the downholepressure gauge. The Hewlett-Packard quartz gaugeshould be used because the pressure responses are verysmall, occasionally less than 0.1 psi. Pulse periods areusually of short duration.

The purpose of pulse testing is to provide estimates ofaverage transmissibility (kh/p) and storage (Qcth) in thereservoir between the wells being tested. Conventionalpulse tests cannot usually provide the horizontal and ver-tical permeabilities of each layer of strata, informationwhich is critical for optimal design of reservoir manage-ment procedures. The FMT can provide the permeabilitydata with the necessary detail.

Optimal management of stratified reservoirs requires aknowledge of the transmissibility and storage values ofeach layer as well as vertical permeabilities across theboundaries between the layers. This is necessary infor-mation if the reservoir engineer is to reliably predict howinjected fluid will travel through the reservoir during awaterflood, CO, flood, etc.

With conventionalpulse testing, it is near impossible toestimate these properties in a stratified reservoir. TheFMT can provide the needed information.

The FMT procedure requires a minimum of two surveysof the observation well. The first survey is conducted se-quentially with the initial suite of openhole logs. Im-mediately following, a disturbance is created in the ad-jacent well by alternating flow rates. Following the flowdisturbance, a second FMT survey is made in the obser-vation well. The second FMT survey should indicate adifferent pressure profile than the first survey. From thisdifference, the degree of vertical and area communica-tion between the two wells in the reservoir can bedetermined.

A numerical reservoir simulator is commonly used toanalyze the data. The pressure profiles and pulse ratesfrom the two FMT surveys are history-matched, allow-ing an estimate of both the horizontal (k,) and vertical(k,) permeabilities of each layer.

Saturation changes are usually negligible during theFMT pulse test and are not usually simulated. The shorttest period virtually eliminates the need to consider otherreservoir influences such as production decline, pressuredecline, well history, field history, etc.

FMT REALITY

The primary goals of formation pressure testing are toquantify the effective permeability of the reservoir andto evaluate the efficiency of the well. Pressure buildupand pressure drawdown are two of the more popular testvariations which are used to evaluate a reservoir.

Formation Multi-Tester tools provide an avenue for welloperators to approach these goals in a quick, relativelyinexpensive way. Other wireline services (e. g., produc-

50

LIST OF SYMBOLS, INCLUDING SUBSCRIPTS

A‘APICC

ctD

AP

Appt

APS

At

DST

Ef

FMT

ffw

g

Y

GRGORGWRhH-P

HYkkabs

kani

kc

kcl

ke

kH

keo

kew

kro

krw

ks

k

m

Area, ft2API units of oil gravityConversion factorCompressibility, psi-’Compressibility of formation fluid, psi-tDepth, ft or mPressure differential, psiDrawdown during pretest(P formation - Pflowing), PsiDrawdown during sampling(Pformation - Pflowing), PsiTime increment, min or setDrillstem testFlow efficiencyFormation Multi-TesterFormation water fraction, percentAcceleration due to gravity, cm/sec2or ft/sec2specific gravity, g/cm3Gamma ray logGas/oil ratio, ft3/bblGas/water ratio, ft3/bblEffective formation thickness, ftHewlett-Packard quartz pressure gaugeHydrostaticPermeability, mdAbsolute permeability, mdAnisotropy (k,/k,)Cylindrical buildup permeability, mdDrawdown permeability, mdEffective permeability, mdHorizontal permeability, mdEffective permeability to oil, mdEffective permeability to water, mdRelative permeability to oil, mdRelative permeability to water, mdSpherical buildup permeability, mdVertical permeability, mdSlope of a pressure buildup curve,psi/cycleSlope of a cylindrical pressure buildupcurve, psi/cycleSlope of a spherical pressure buildupcurve, psi/cycleViscosity of gas

PO

PW

MWCJ

mm

gfP

+

P

PC

Pf

pg

Pi

PO

pw

PWS

P*

PI

9

qpt

r

rinv

Rm f

Rrf

Rt

RW

P

P ma

p f

%J

SG

siw

S O

SR

S W

Sx0

SPtVVPC

Viscosity of oilViscosity of waterMud weight, lb/gal or lb/ft3Matrix stress, psiParts per millionFluid pressure gradient, psi/ftPorosity, percentPressure, psiCapillary pressure, psiFlowing pressure, psiGas pressure, psiFormation pressure, psiOil pressure, psiWater pressure, psiPressure at probe after shut in, psiFormation pressure extrapolated fromHorner Plot, psiProductivity indexFlow rate, cm3/sec or bbl/dayFlow rate during pretest (chambersize/time to fill), cm3/secProbe radius, in.Depth into formation affected bybuildup, cmResistivity of mud filtrate, ohm-m2/mResistivity of recovered fluid,ohm-m2/mTrue resistivity of the formation,ohm-m2/mResistivity of the connate water,ohm-m2/mDensity, g/cm3Matrix density, g/cm3Fluid density, g/cm3Gas saturation, percentGas solubilityIrreducible water saturation, percentOil saturation, percentSolubility ratioWater saturation, percentWater saturation of the flushed zone,percentSpontaneous potential curve, mVTime, min or setVolume of liquid or gas, cm3 or ft3Variable Pressure Control

55

VPC-FMT

WCZ

Z

Formation Multi-Tester with VariablePressure ControlWater cut, percentCompressibility factorVertical height

BIBLIOGRAPHY

Beal, C.: “The Viscosity of Air, Water, Natural Gas,Crude Oil and Its Associated Gases at OilfieldTemperature and Pressure:’ Trans. AIME (1946).

Bonham, L.C.: “Solubility of Methane in Water atElevated Temperatures and Pressures:’ Bull. AAPG(1978).

Brown, K.E.: The Technology of Artificial Lift Methods,Vol. I, The Petroleum Publishing Co., Tulsa, Okla.(1977).

Chew, J.N. and Connally, C.A.: “A Viscosity Correla-tion for Gas-Saturated Crude Oil:’ J. Pet. Tech. (1959).

Craft, B.C. and Hawkins, M.F.: Applied PetroleumReservoir Engineering, Prentice-Hall, Inc., EnglewoodCliffs, N.J. (1959).

Log Review I, Dresser Atlas Publication (1974).

Log Interpretation Charts, Dresser Atlas Publication(1983).

Fertl, W.H.: Abnormal Formation Pressures, ElsevierScientific Publishing Co., New York-Amsterdam (1976).

Frick and Tayler: Petroleum Production Handbook,McGraw-Hill Book Company (1962).

Gunter, J.M. and Moore, C.V.: Improved Use of WirelineTesters for Reservoir Evaluation, SPE 14063 presentedat SPE International Meeting on Petroleum Engi-neering, Beijing, China, March, 1986.

Hawkins, M.F., Jr.: “A Note on the Skin Effect:’ Trans.AIME (1956).

Horner, D.R.: “Pressure Buildup in Wells:’ Proc. ThirdWorld Petroleum Congress, Leiden (1951).

Katz, D.L., Cornell, D., Kobayashi, R., Poetmann, F.H.,Vary, J.A., Elenbaas, J.R., and Weinaug, C.F.: Hand-book of Natural Gas Engineering, McGraw-Hill BookCompany (1959).

Mathews, C.S. and Russell, D.G.: Pressure Buildup andFlow Tests in Wells, SPE Monograph (1967).

Milburn, J.D. and Howell, J.C.: “Formation Evaluationwith the Wireline Tester - Merits and Shortcomings:’J. Pet. Tech. (October 1961).

Moran, J.H. and Finklea, E.E.: “Theoretical Analysisof Pressure Phenomena Associated with the WirelineFormation Tester’ J. Pet. Tech. (August 1962).

Odeh, A.S. and Selig, F.: “Pressure Buildup Analysis,Variable Rate Case:’ J. Pet. Tech. (July 1963).

Pirson, S.J.: Handbook of Well Log Analysis, Prentice-Hall, Inc., Englewood Cliffs, N.J. (1963).

Schowalter, T.T.: “Mechanics of Secondary Hydro-carbon Migration and Entrapment:’ Bull. AAPG (1979).

Sethi, D.K., Vercellino, W.C., and Fertl, W.H.: The For-mation Multi-Tester - Its Basic Principles and PracticalField Applications, SPWLA Twenty-First AnnualLogging Symposium (1980).

Slider, H.C.: Practical Petroleum Reservoir EngineeringMethods, The Petroleum Publishing Co., Tulsa, Okla.(1977).

Standing, M.B.: Volumetric and Phase Behaviour of OilField Hydrocarbon Systems, Reinhold PublishingCorp., New York (1952).

Van Everdinger, A.F.: “The Skin Effect and Its Influenceon the Production Capacity of a Well:’ Trans. AIME(1953).

APPENDIX A

An example problem utilizing FMT-measured pressuresfor many of the computations discussed in earlier sec-tions is presented in this Appendix. The FMT pressurerecord in Fig. A-l will be used through the followingcomputation sequences.

A pretest volume of 10 cm3 with a 0.562-in. diameterprobe was used during the pretest. The following deriva-tions will be made from this pretest record:

l drawdown permeability, k,

l spherical buildup permeability, k,

l effective bed thickness from spherical buildup, h

l cylindrical buildup permeability, k,

l time estimate for retrieving a 10-liter sample

56

Permeability from Drawdown

The flowing period begins at recorder time of 31 seconds,indicated as t=O on the log of Fig. A-l, and ends at 39seconds, indicated as tl =8 seconds. The fluid with-drawn is a filtrate having a resistivity of 0.027 ohm-m at170’F (76OC) or 120,000 ppm NaCl equivalent. Usingthe chart of Fig. 24, a viscosity of approximately 0.5 cpwould be estimated. The minimum steady-state flowingpressure during drawdown is approximately 900 psi andthe pressure builds to about 3930 psi. Drawdownpermeability, k,, is determined using the followingequation (11):

k, = 1842 x C x

From the information above,

C = 0.75

q = 10 cm3/8 set = 1.25 cm3/sec

c1 = 0.5 cp

d = 0.562 in.

AP = 3930 - 900 = 3030 psi

The drawdown permeability, k,, is therefore

k, = 1842 x 0.75 x (o.25~~~o) = 0.51md

Permeability from Spherical Buildup

The raw data taken from the log of Fig. A-l is tabulatedas At, t + At, and the spherical buildup parameter is

&-v&A plot of the pressure recording versus the sphericalbuildup parameter is given in Fig. A-2. This plot showsthe spherical buildup pressure estimate to be 3938 psi andthe slope, m,, to be 930 psi/set%. Buildup permeabili-ty is given as: %

x Q+>”

Using the FMT pretest data in question,

q = 1.25 cm3/sec

+ = 0.16

ct = 3 x 10-s

P = 0.5 cp

ms = 930 psi/set

The computation is

x (0.16 x 3 x 10m5)” = 0.19 md

FIGURE A.2Spherical buildup plot.

Effective Bed Thickness Computation

For this computation, assume that the anisotropyk/k, = 1. From Fig. A-2, the extrapolation ofspherical pressure buildup, p*, is 3938 psi, whereas thedata deviates toward a higher value of formationpressure, p, of approximately 3940.5 psi (see dashed line).The effective thickness, based on the pressure matchcriterion, is given by Eq. 22:

Taking the following values for this FMT test,

k,/kn = 1

Pi - Ps* = 2.5 psi

V = 10 cm3

The bed thickness is calculated to be

10 x 1l-l’ = 1.2

4rr(2.5)(0.16) x 3 x 1O-51 %

= 48.57 cm = 1.59 m

58

Permeability from Cylindrical Buildup

The time parameter, t + At/At, for this FMT test wastabulated. The cylindrical buildup (Horner) plot for thistest is plotted on Fig. A-3. Extrapolation of the lineardata indicates a formation pressure, p* = 3947.6 psi. Theslope of the linear portion of the data is m, = 198 psi&cle. For permeability from cylindrical buildup, Eq. 19 is

which, for the data of this FMT test, becomes

k, = 88.4 = 0.18 md

Time Estimate for Sampling

An estimate of the time required to retrieve a lo-liter(2.64-gallon) sample may be obtained by using Eq. 2 toestimate the time per gallon.

t =63.1 x Appt

qpt x 4

For the FMT test of Fig. A-l,

Appt = 3930 - 900 = 3030 psi

APS = 3930 psi (sample is taken against anair cushion chamber)

qPt = 1.25 cm3/sec

and hence, the time in minutes required per gallon isestimated as

63.1 x 3030t = = 38.9 min/gal

1.25 x 3930

and

FIGURE A-3Cylindrical buildup plot.

2.64 gal x 38.9 min = 102.7 min (or 1 hr, 42.7 min) tofill a lo-liter (2.64-gal) tank.

59