Petroleum Engineering 620 — Fluid Flow in Petroleum Reservoirs Petrophysics Lecture 6 — Electrical Properties of Reservoir Rocks Quote du Jour: When I transfer my knowledge, I teach. When I transfer my beliefs, I indoctrinate. — Arthur Danto (1968) Topic: Electrical Properties of Reservoir Rocks Objectives: (things you should know and/or be able to do) Fundamentals: © Be familiar with conductors and non-conductors of electrical current in porous media. © Be able to derive the following expression and define all terms: RwALy Apl Note that ¢= 42, and r= 42 where + is called “tortuosity” and is considered to be a measure of réck cementation and/or the pore throat distribution. r can not be measured directly and must be determined by calculation. © Be familiar with the definition of the formation resistivity factor, F, where F is defined as: paket Ro= Ry ¢ © Be familiar with the definition of the resistivity index, 1, where [is defined as: oko k Ro FRw Clean Sand Relations: © Be familiar with and be able to use the Archie and Humble equations to estimate porosity given the formation resistivity factor, F. 7 Reg © Be familiar with the Archie result for water saturation, S,,, (fraction): a Sy ate = Ben © Be able to determine the cementation factor and saturation exponent for clean (non- shaly) reservoir rocks using the standard log-log plotting approach, as well as the new "type curve" approach. Shaly Sand Relations: ‘© Be familiar with the effect of shale content on the resistivity index/saturation relation. In particular, be familiar with using the Waxman and Smits model and the Schlum- berger total shale model to yield the resistivity index/saturation relation, © Be able to determine the saturation exponent for shaly reservoir rocks using regression analysis and type curves. Be familiar with the effect of shale content on the formation factor/porosity relation. In particular, be familiar with using the Waxman and Smits model and the "parallel resistance” shale model for the formation factor/porosity relation. © Be able to determine the cementation factor for shaly reservoir rocks using regression analysis and type curves,Petroleum Engineering 620 — Fluid Flow in Petroleum Reservoirs Petrophysics Lecture 6 — Electrical Properties of Reservoir Rocks Lecture Outline: ‘© Development of saturated porous media relations m Resistance and resistivity = Water filled cube — Water filled porous media '™ Definition of the formation resistivity factor — Development of the Archie and Humble formation factor relations ™ Definition of the resistivity index ~ Development of the Archie saturation relation ‘© Developments using the Waxman and Smits model and the Schlumberger total shale model to yield a resistivity index/saturation relation. — Graphical analysis using type curves (direct plotting). © Developments using the Waxman and Smits model and the "parallel resistance” shale model to yield a formation factor/porosity relation. = Graphical analysis using type curves (overlay matching). Reading Assignment: © Review attached notes. m Resistive Characteristics of Porous Media Theoretical and Experimental Con- siderations, = Additional Comments on the Resistivity (or Formation) Factor (F). Formation Factor/Porosity Relations for Shaly Sands @ Archie, G.E.: "Electrical Resistivity Log as an Aid in Determining Some Reservoir Characteristics,” Trans. AIME (1942) 146, 54-62. © Archie, G.E.: “Introduction to Petrophysics of Reservoir Rocks," Bull., AAPG (1950) 34, 943-961. © Wyllie, M.R.J. and Gregory, A.R.: ‘Formation Factors of Unconsolidated Porous Media: Influence of Particle Shape and Effect of Cementation,” Trans. AIME (1953) 198, 103-110. © Waxman, M.H. and Smits, L.I.M.: "Blecttical Conductivities of Oil-Bearing Shaly Sands," SPEJ, (June 1968) 107-122. © Clavier, C., Coates, G., and Dumanoir, J.: "Theoretical and Experimental Bases for the Dual-Water Model for Interpretation of Shaly Sands," SPEJ, (April 1984) 153-168. © Argaud, M., Giouse, H., Straley, C., Tomanic, J., and Winkler, K.: "Salinity and Saturation Effects on Shaly Sandstone Conductivity," paper SPE 19577 presented at the 1989 SPE Annual Conference and Technical Exhibition, San Antonio, TX, 8-11 October, 1989, 49-60.Petroleum Engineering 620 — Fluid Flow in Petroleum Reservoirs Petrophysics Lecture 6 — Electrical Properties of Reservoir Rocks ‘Exercises: For your own practice/skills building—do NOT turn in! © In each of these derivations/problems you are to work in complete detail and you ‘must show all work. Using fundamental principles, you are to derive the formation resistivity factor, F, in terms of the "tortuosity," +, where F is defined as: =Root FoR} © You are to provide a critical and detailed review (at least 1 page) for the following paper(s): 1 Archie, G.E.: "Electrical Resistivity Log as an Aid in Determining Some Reservoir Characteristics," Trans. AIME (1942) 146, 54-62. ™ Archie, G.E.: "Introduction to Petrophysics of Reservoir Rocks," Bull., AAPG (1950) 34, 943-961. = Wyllie, M.R.J. and Gregory, A.R.: "Formation Factors of Unconsolidated Porous Media: Influence of Particle Shape and Effect of Cementation,” Trans. AIME (1953) 198, 103-110. = Waxman, MH. and Smits, LJ.M.: "Electrical Conductivities of Oil-Bearing Shaly Sands," SPEJ, (June 1968) 107-122. ™ Clavier, C., Coates, G., and Dumanoir, J.: "Theoretical and Experimental Bases for the Dual-Water Model for Interpretation of Shaly Sands," SPE, (April 1984) 153- 168. = Argaud, M., Giouse, H., Straley, C., Tomanic, J., and Winkler, K.: "Salinity and Saturation Effects on Shaly Sandstone Conductivity," paper SPE 19577 presented at the 1989 SPE Annual Conference and Technical Exhibition, San Antonio, TX, 8-11 October, 1989, 49-60. For each paper you are to address the following questions: (Type or write neatly) © Problem: — What is/are the problem(s) solved? — What are the underlying physical principles used in the solution(s)? © Assumptions and Limitations: — What are the assumptions and limitations of the solutions/results? — How serious are these assumptions and limitations? © Practical Applications: — Whaat are the practical applications of the solutions/results? — If there are no obvious "practical" applications, then how could the solutions/ results be used in practice? © Discussion: — Discuss the author(s)'s view of the solutions/results. ~ Discuss your own view of the solutions/results. Recommendations/Extensions: — How could the solutions/results be extended or improved? — Are there applications other than those given by the author(s) where the solu- tion(s) or the concepts used in the solution(s) could be applied?Resistive Characteristics of Porous Media Theoretical and Experimental Considerations (from Petroleum Engineering 620 Course Notes — 1997) Petroleum Engineering 620 Fluid Flow in Reservoirsocx séuns PonosiT oF Rock. ih R eouns aa as wel Fg. 623—Propapaton ofan deca charge ina contne From: Jorden, IR. and Campbell FL: Well Logging l-Electrie and Acoustic Logging, SPE Monograph Series, 10 (1986), Richardson, TX.FORMATION FACTOR Formation factor is defined as the ratio of the resistivity of com- pletely brine saturated rock to the resistivity of the saturating brine. Tests are conducted on 1" or 1 1/2" cylindrical plugs or whole core samples. Formation factor is, therefore, a function of porosity and the pore geometry of the rock. One expression conmonly observed is: Ro Rw 1 a Where m is called the cementation factor and is equal to the slope of the line relating F to @ on a log-log plot. In many formations a conmon relationship is: = te F im *. Tests are conducted over several day increments until values stabilize. Samples should be selected to cover the porosity range noted in the core. This assures better definition of m and a. From: Keelan, D.: "Special Core Analysis," Core Laboratories Report (1982).7 “XL “uospreysry ‘(9861) OF ‘sug ydesBouopy ‘Ads ‘8us8807 Qnsnosy pun ouioaqq--]1 88307 poms "7 “IWoqduey pur “y’f ‘wapioy “wor (cc@™oL J04e) sweisXs e1od peziteep1 40} Ausoi0d pus JolDe} Auansisau uoNeWHO} YeaMIog diysuONe|oN—B2"9 “Bld ‘AG 981) “ALISOWOE ob 101 +0 10h u : = "4 ONY ® NI SNOILVIEVA _BAID OL G3IMVA ‘6660 LV LNVISNOO @ Wit ‘OWSISAS OMA ONY SNVIE 4 "4. ONY @ NI SNOLLVIEVA ‘ANID OL G3IVA & jou 'UBLVM N3AID ANW HO- 2 YOLOVS ALIALISISSY NOLLVAWOS wWISAS NVI “4 ONY @ NI SNOLLWIVA ‘BAID OL G3IHWA P jor '800.1¥ INV1SNOO "4 ONY @ NI SNOLLWIEVA JAI OL O3IUVA E -| © waisss San iF Toye u B de fa 'v WALSAS DNA ONY 3anL“XL “uospzeyory ‘(9861) OT ‘seuss ydesBouoyy Bas ‘303307 ausnooy pun ousoaqgr-1f Surs3O7 Toy TA TloqduIeD puE YE “topIoy :woLy “syp01 Buywoqueien jo 8 out vo ssoubEna Jo 0H 3—sz"9 ‘Somaonpued te ‘AG 924) ‘ALISOOd ey O41 _oso ozo 10 soo 00 100 HOLOV4 ALIAILSISSH NOLLVAYOS“€¢961) wodoy souows0qe7 2109 t¢heuy 109 esas, “q‘wBe WON a, Weare] Wl uOReUEA BugeRsN| JoyoeY UOREWALI painseay Jo aueY 10} ‘Aysouog SA.10}0e4 uoReULIOY ‘Aysoiog SA JO}9e4 wOQeULOY uonaey: () Aysouog ae 000 1 OSA Ft sy = 4303 (4) soy9e4 woyeun0y SS (4) s0}9e4 woyewn04 oat‘OIT ~€01 ‘S61 (E561) AIT “sung, ,uoNmuatIaD Jo 09j3q pur advys ofonzed Jo eouoNUL EIpapy Shoxog pareprlosuooun Jo sioloey UoTeMION, “YY ‘AOBAID pue TWN MAA “WoL, “survonooy aaivanosno>s “ana aaa ae721 = ¢ “a1voRI09¥ snowed ‘ovanosniooi 40 viva MISOIOs-¥OLD¥4 NOLO! "EO asters es ff‘At 5000 PSI Overhurden ~ 2.23 At Zero taal F= oe Effect of Overburden Pressure on Formation Factor For 6 = 15% Formation Factor Varies from 43.6 to 68.8 For F=45 Formation Factor (F): fo ae We Varies from 15% to 18% 1 \ 0.001 0.01 01 1.0 Porosity (): Fraction Effect of Overburden Pressure on Formation Factor From: Keelan, D.: "Special Core Analysis," Core Laboratories Report (1982).RESISTIVITY INDEX @ 011 and gas are not electrical conductors. Their presence in an element of reservoir or in a core sample will reduce the mean cross-sectional area of the flow path for an electric current and increase the length of the flow path, thus increasing the resistivity. ©@ Resistivity Index is defined as the ratio of rock resistivity at any condition of gas, oil, and water saturation to its resistivity when completely saturated with water: © Thus, the Resistivity Index is a function of water saturation. It is also a function of the pore geometry. The presence of cation-exchangeable clays (montmoritlonites) often cause apparent low Resistivity Index values to be observed, particularly as water saturations decrease. @ The Resistivity Index exponent, n, is also influenced by confining or overburden pressures, and should be determined under overburden con- ditions where the rock is significantly susceptible to the effect. © A minimum of three saturations, and preferably five, should be obtained on each sample. Individual sample I versus S, plots, as well as com posite plots of all samples are typically reported. From: Keelan, D.: "Special Core Analysis," Core Laboratories Report (1982).c “XL ‘wospzeyory ‘(986 1) OT ‘set19g yderSouoy dS “8u18807 ousnosy pun qusoays--17 BMSIOT peas “Tf ‘eaduseD pue “Y’ “wepioy wos “yoo. Bue=q uoaieoo1phy Jo Ayaonpuos uo eyuhd Jo 199}k3—Ze'9 “Bid é ‘90! Buli€oq-vogieaospAy j0 Auaonpuod feoutao}o ‘Ad 984) ‘NOLLVHNLYS H3ALYM uo vopeintes uoqreaosphy pu ueiue9 fe Jo 1/9—06"9 “Bia OL 104 Ob ‘Ad 984) ‘NOLLVHNLVS HAV Orso | 98") tv} ses00 | 91°} sovo ess00 | 0 |voro 2.82 @ |yove anvs 92 Jou wos BIeQ Jose wo ore ] sr oo sera De ‘waLvM ONY }OOU NAAID ANY HO. XSQNI ALIALLSISAY NOBYYDONGAH XSQNI ALIALLSISSH NOBYVOOHGAH yoyem vaAl8 Aue 40} —y“XL. ‘wospueyory ‘(986 1) OF ‘sou9g ydesBouopy Aas ‘318807 susnosy puw surayq~-11 SuI83OT Nem = TA ‘ToadueD pue “Y'f ‘uaploy wor 63 s0y8) soucispues 40 Auwganpuos jeau00}@ Uo AgvioN JO 10013—¥e"9 “Bi “{ges0ujuuar pue Aousans 14) “289 Jo ANFoNPUOD ELIE! UO ANGEHON ‘Ad 201) ‘NOLLWUNLYS HaLVM o1go soso ro co co 'NOLLWUNLVS UBLVM o1 co so co zo 10 L0s00 08 3409 \ damyauven XAQNI ALIAUSISIY NOBUVOOUOAH JLaMeuBLVANON1000 |e = 100 ~< 3 = 2 2 3 Bw oc Effect of “‘n’”’ on Water Saturation AtRI = 30, Sy Varies From12% to 21% Pore Space 1 0.001 0.01 01 Water Saturation (S,) Fraction Resistivity Index vs Water Saturation for Range of Measured Saturation Exponents From: Keelan, D.: "Special Core Analysis," Core Laboratories Report (1982).Saturation my Rw _ & | Measured Porosity 7 Cementation Exponent Rock Properties Influencing Calculated Water Saturations From: Keelan, D.: "Special Core Analysis,” Core Laboratories Report (1982).CONDUCTIVE SOLIDS ©@ The clay minerals present in a natural rock act as a separate conductor and are sometimes referred to as "conductive solids". Actually, the water in the clay and the fons in the clay water act as the conducting materials. The effect of the clay on the resistivity of the rock is dependent upon the amount, type, and manner of distribution of the clay in the rock. A few minerals are conductive in their own right, but do not constitute a significant problem in formation evaluation. Clays result in a variable formation factor dependent on the water salinity (resistivity) used to saturate the core. Fresh water yields low F values and low m values. 1000 é { "4 Tate | 2 — = g Ew 1 sis | aor 01 0 10 Brine Resistivity: Ohm Meters @75°F Effect of Brine Resistivity on Formation Factor From: Keelan, D.: "Special Core Analysis," Core Laboratories Report (1982).Additional Comments on the Resistivity (or Formation) Factor (F) (from Petroleum Engineering 306 Course Notes -- 1991) Petroleum Engineering 620 Fluid Flow in ReservoirsPere 0b -tebbitionk. Commente on the Pedtsthaty Debyuition of berisrtivity,, 2 (*-m) be rA wo where, 4 re resistaute (uhat you ould meanwe), Az eross-s0ctional area J. to curvent flay m2 b= Length of conduetor, m Gohinitin of £ 4th a where, e 42 resistance of cove At lool, water saturation, Ae erese- sectional area. of cove face, m? Lz Lath of cwe,m BA saltlon of Lu hue kw he bere, be by = vesictance of the fomation bring Agz croescsectional area of the container, wt lp = eHective hengplh'sf the contalnerm (3) from the tiapre we vote that 4 fren of Milo ayae te) a fhe available for thuid stoacpe 4o= My fle $b @ yY Factor, F OTHE vofsolar Kos a BRINEFILEO PORE Space eernocrne ‘CONDUCTOR Anca oF Fone sPack PLANE WNOP eauats 4 sbuos| ‘Couonent ensucaton| PonosiTy oF nock Eouats Fig. €:23—Propagaton ofan saci charge a conc Prope charge a conde Bb ned b pel a "olome" alerage crns-sectional aren,O78 elsol4 Lebiution of formation factor, F At 00% saturation, 2 hy, which yiehde Bb ehh, @ 4 Ap Soluinay bye tr Fe kay gles Fe fos Alp @ bu z where F ts debined ar the "rmation reeietivity tater", Comblaing 4s 4,8 aed ? gives Fe kyle lp a 4 a Zz Fe : (#) 4) where we new debine the'tertvority” as t+ CU)" Take apes fet 4) ¢Formation Factor/Porosity Relations for Shaly Sands (from Petroleum Engineering 306 Course Notes — 1991) Petroleum Engineering 620 Fluid Flow in ReservoirsDeve tof Formation Factor (4s) / Poros'try (4) Relations Oulelat & tly Si Wacman- Smits Shele Relation Ueaman- Smits (GPT, Tne 08) ahve the following resht Br the condsctivityy of cheby sands hoz Lita) oO whee 4 Gee Condbetiity of core ak Syl, Cohmem)” 5 + Smale teas thulby, tuctor Ceyltul, Hmencionkese 4 = Lybly* chele parameter, dinenvionless fu ® tamaton water recktivity ohm-m Poel B= Lion emp l-0r3lby 146, correlation varible ~ moh eg. lite? 2 vale concentration of clay exchange von, mob eg, diter Using, the debddlttan of the formatlen vecktivity, factor, 6 py Vgc, gis tL. 4lra) BD o ae 4 = lealé a@ Gecdll the general medal tor F aber ty Archie fee pg” @ Comblining Ege. Sand 4 aives frelon) oe cle e's clita) 6) where ¢'>0, 20 & tod of ty rip tor a shaly sand shovdd 4e abere a2 trend of ty vib bra siniler clean sard, WeOulslu wilh use the data given by Viwman and init to verity or dkpote this revlt. Let Hist we wih develop an alternate yelatlon taced on pore veistance theary propored ty Wyllie, Prrlleh Pesistance Lelation node and Wyllie (Tans AME, 1950) proposed that chany acte Lie a parallel resistor da enjmmetion with the reek and farmation brine. Gecall Ohm's law ev @ 7 Ae zvesistors th parallel the total evnrent, L, és equal te the sum of the currents tn the loop, Also, te voltage across the leap, V, + constant, This alves 2 Bye OE or ueingy by.b YorvVev we on OR Lininatlng, the covstant voltage term Leto @ et Sobstitethng the aggavent reeistivity Lop (0 Sigg » One the dary aud nonchauy recistivitles eyhes ele 4) br lay Woheiply ing throveh G8 by the veristluiny of the 4ymatlon brive, bu, qlves@ uta 2 hla Be a Lebel 2) 4 a where. be ky eo) Play Sling by. for the "agpavent" formation recistiviby Factor; A, ives yo be mS (+bF iB Eley Geeall the general model tor F ghien by Archie Cy(a)] fee g™ Conblaingy Byes #and U ayes eee eae “” se Mate! * Filla gives ¢>0' sndleating that a trend of Ab te a shaby sand chould lay below a tend for a clean aud, This contradicts the Wasman- Sn relatlon, w@ -mO slaty The onliy wavy to vesobve th'r contradiction Lif there is A veclutlon) ts to compare ehean nnd shal tymatibn dts, late trom Waman and Smite ave plotted on the akuched graph, the cand om wbves are es blloss ’ oe Set e mn ‘ clean Saudstone OAS zs 2 Thaliy Sandetone #1 2,42 zt g haley Sandstone #2 “ee eB 4 Wor, Shaliy Saudetone 18 20 s Bocene Shdy Saudetone O88 “8 ¢ Tttiainy Sholry Seocobore ue 20 The only clear concliilon it that ef > el) tor these sanplec, The does not mean that ‘thle conclusion /s te h general, These data. thom Waxman} Gnits suggest tat the Warnwn-smits concgot ts more app leabhe tb tild data than the L resistance concept, The mort important conchoslon i's that the Ayche type model lyse’ 6") appears b be valld for tly sands -- which allows vr» make cabevdatlons regardless of ubore riedele hr e! te eonect, Tt be adso worth voting that, for the Luatnan- Smits dat, , om, CL seqgrertéd th the corse notes,‘Waxman and Smits Data for Clean and Shaly Sands 10° 107 jot 2 4 geree 23 eggzee 23 teeny 10 popes hE Le eresttbetette teeetteetttct 10° fea Sas Rel} He “Clean” Sandstone Prairie HELL E "Shaly” Sandstone #1 7 rm i ‘Stat Sansone #2 / i “Very Shaly” Sandstone —f> Hl Bocce Sly Senatee zeae Hae 10 tee SSeserst as Seer aSPlot of Formation Resistivity Factor/Porosity Relations for Clean Sands (No Shale) [Legend (F=R JR) 9° (Archie, c=1) (Humble, c=-0.62) =F = 141.85(-4)(60.10) =F=G- ee arte tom =F =Fe Fp Fp. 129) = (13219 - 032199779 =F p(Slawinsk) F,(Maxwell) are for equations ‘while the Humble and "The Maxwell and Slawinski unconsolidated spheres, for consolidated It Perez-Rosales equations ae: sandstones, po r To ot|—RJR, versus Porosity($) type curve solution Grid paper for F- ES Hi i ae a Hee BE aH 5 aoe Hina 4 HH HEC a CS 10! feSaturation/Resistivity Index Relations for Shaly Sands (from Petroleum Engineering 306 Course Notes — 1991) Petroleum Engineering 620 Fluid Flow in ReservoirsDevelopments of Saturation [Lesistivity Index belatious © THE u/z/a1 be Gaby Sands hevall. the Archie empirical. resht Jehu" a z Wecouan and Soits (e761, dime #68) developed a semnianaby tical ander catvrattion / reristivtty relation for haly sands, Tolt relation ds Ze 5" A +4 J] a a +4/sy, as ku 8 ay @ but temasion wader vesictivityy, chm =! B= Liab explore eu) 46, correlation vartble, Mita v7 0,2 valeme concentration of chry stchange cadians, mol eguiv hier” Wave ad Melawps G2 ts theoredically vigoran, ib is abeo dfbiult & wale for water saturation, $, Leavranging ard sohulag tr sy will ‘Hortate this pont se liga) 2 lee i sw ce See a eee ite @ e Mote that tor arbitmry n-valves bp, ¢ eanwot be wbved iphicitly and some root ablug method u's reguired,@ 78 up Hower, an exphlat solotion for bt st highly desired, Me vecolia this dilema by whving Gh fr nzt, which gates Sit aSy = lta a zr Apely Ing the guadratie fumble to by S, and atky accepting the Largest root ahes sve Lh ae Miradlr wal 2) 2 S we conehide, tor practical 5, that nee ts wbhicient. A tape come” sate of 2 alka Mote that the ace cases have several n-velues and the nez cases have severel a-values, This plot can be osed to determine n tw the aro cats antler a tor the nz casts, simply by plotting, Dive, Su cate directly on the plete and reading the appropriate a or n-velve. An ccample at te end oF tht develypment Hluctrates these determinations, Other Comidernbous Ht prevent we donot hare a phyrical inderrtandinay of the 2" cockbitient in the Wasmen-smite relation (Bs, 2 and3) Ore yostble concept luokves the use of the “total ct ; equation developed by schbumberger Laearcls, This lon is Le si + Mth Sy @ A fs byli-ty,) shOme ull Moltiplying th b.? by & als Sets Yu 1 se ff BM? (i-th) a since Faly= Slew , Elean= “Gd? j| we bave ca Edy _! Zz + th Sy & Fb St Be eh Sy (s) Febeon C-%eh) Bh Reeeranalaty the tesiman- sits relation (és) to be ln the same form as BiB we base “ 9) zt WA Wa Iguana, cocditleate tu Lys. 8 a= Fibabey / de) Ha kebean Crh) 4: bf w& i) wa hy, Dhulthay by u by Eg '0 aves 25 Lo flean Vo li-Ueh) tz) Lah Febaly where, Eleay = formation revitivity factor tora. chean sand Fibdlay 2 trmaction recstivity factor ter a shal sand b= besistivity at 542! sn the shally sand, ohm-w eye Cesittivity ab iy =! tn a nearby shake, ohm-muj = shale traction tn the shabry save! @ 706 ulelr Uh ts determined tom some total shabe dleaton for dnstavce, a Neutron density beer erenplot. fy, 12 thdieates 2 theoretical Link between the Wexman- nits model and reservoir rack properties. This vastliy imaroves our Gontidence tn the liaman-smite moded. Le cs not clear that the result in Gle can be extended for cage where néz tor the lieeman-tmits mah, &. 2. banple - Btermination of n'coetbiclent for a clean and, Cate tom a dolomite with vigniticant dlagentit tn the form of caltite nagets - thie particular sample aypeare to be homogancors Lith no caberte miagets, gels ke itt mol Siu, Hench ion Eble fe vpete (air-brine) ! / 292 of 1262 281 obese hss Mae oes 2802 mae 0.508 2468 Jone ezi 4083 IA Mottin thee date on the chean sand! plat qlves net, See abtached type corve tr detasls.RVR, 1s) © 76 whey Plot of Saturation/Resistivity Index Relations for Clean Sands (No Shale) : paint heed titi dt ter ii NTE EET at Aether t iNr “pease ete HEHE a ANE tt HH Hee ee EPH ofa od eee Saag NL. PET ae 10? 1 10 Sy Water Sanration From thit plot WalePlot of Saturation/Resistivity Index Relations for Clean Sands (No Shale)Plot of Saturation/Resistivity Index Relations for Clean and Shaly Sands (n=1) a ‘Sy Water SaturationPlot of Saturation/Resistivity Index Relations for Clean and Shaly Sands (n=1.2) evalues as indicated |— "Shay" Sand Curves n=12, a-values as indicated1S)=RR, Plot of Saturation/Resistivity Index Relations for Clean and Shaly Sands (n=1.6) Jo" 10° gered 10° |— “Shaly* Sand Curves n=1.6, values as indicated Tl ‘=f {Waxman-Smits Shale Equation: oo (S,) = RYR, = S,* [(1+a)(i+a/S,)] Ufa case om1.6) cannot be solved a lexplicitly for saturation. Heisei meine 10° 10°. +] where Ve. is obtained from a Neutron- {] [RR at APlot of Saturation/Resistivity Index Relations for Clean and Shaly Sands (n=2) [dsayrsas, >) HEC: Sy, Water SaturationPlot of Saturation/Resistivity Index Relations for Clean and Shaly Sands (n=2.4) tt a U a es, = RVR, =S,* (C+a)CL+8/S 9} Tis case (n=2.4) cannot be solved explicidly for saturation. faa RARE, where Vis obtained from a Neutron Sapafaiit Sam aee SSE SEEReferences — Electrical Properties of Reservoir Rocks: © Archie, G.E.: "Electrical Resistivity Log as an Aid in Determining Some Reservoir Characteristics," Trans. AIME (1942) 146, 54-62. © Archie, G.E.: "Introduction to Petrophysics of Reservoir Rocks," Bull. AAPG (1950) 34, 943-961. © Wyllie, M.R.J. and Gregory, A.R.: "Formation Factors of Unconsolidated Porous Media: Influence of Particle Shape and Effect of Cementation,” Trans. AIME (1953) 198, 103-110. © Waxman, M.H. and Smits, L.J.M.: "Electrical Conductivities of Oil-Bearing Shaly Sands," SPEJ, (June 1968) 107-122. © Clavier, C., Coates, G., and Dumanoir, J.: "Theoretical and Experimental Bases for the Dual-Water Model for Interpretation of Shaly Sands," SPE, (April 1984) 153-168. © Argaud, M., Giouse, H., Straley, C., Tomanic, J., and Winkler, K.: "Salinity and Saturation Effects on Shaly Sandstone Conductivity," paper SPE 19577 presented at the 1989 SPE Annual Conference and Technical Exhibition, San Antonio, TX, 8-11 October, 1989, 49-60. Petroleum Engineering 620 Fluid Flow in ReservoirsThe Electrical Resistivity Log as an Aid in Determining Some characteristics Reservoir By G. E, Ancars* (Dallas Meeting, October 1043) ‘Ta usefulness ofthe electrical resistivity og in determining reservoir characteristics is governed largely by: (1) the accuracy with which the true resistivity of the formation can be determined; (2) the scope of detailed data concerning the relation of resistivity measure- ments to formation characteristics; (3) the available information concerning the conduc- tivity of connate or formation waters; (4) the extent of geologic knowledge regarding proba- ble changes in facies within given horizons, both vertically and laterally, particularly in relation to the resultant effect on the electrical proper- ties of the reservoir. Simple examples are given in the following pages to illustrate the use of zesistvity logs in the solution of some problems dealing with oil and gas reservoirs. From the available information, itis apparent that much care must be exercised in applying to more complicated cases the methods suggested. It should be remembered that the equations given are not precise and represent only approximate relationships. It is believed, however, that under favorable conditions their application falls within useful limits of accuracy. InrRopucri0N ‘The electrical log has been used exten- sively in a qualitative way to correlate formations penetrated by the drill in the exploitation of oil and gas reservoirs and to provide some indication of reservoir content. However, its use in a quantitative way has been limited because of various factors that tend to obscure the significance of the electrical readings obtained. Some of these factors are the borehole size, Sep OTSCeSE BG ope feceeae Peres Pepeocdow Tecusoiocy, january tors. "FSkel Oi Co,, Houston, Texas. 4 the resistivity of the mud in the borehole, the effect of invasion of the mud filtrate into the formation, the relation of the recorded thickness of beds to electrode spacing, the heterogeneity of geologic formations, the salinity or conductivity of connate water, and, perhaps of greatest importance, the lack of data indicating the relationship of the resistivity of a formation in situ to its character and fluid content. On the Gulf Coast it is found that the effects of the size of the borehole and the mud resistivity are generally of little importance, except when dealing with high formational resistivities or extremely low mud resistivities. Fortunately, little practical significance need be attached to the exact values of the higher resistivities recorded. Low mud resistivities are not common, but when this condition is encountered it may be corrected by replacing the mud column. With the present advanced knowledge of mud control, invasion of mud filtrate into sands can be minimized, thereby increasing the dependability of the electrical log. ‘The effect of electrode spacing on the recorded thickness of a bed is often subject to compensation or can be sufficiently accounted for to provide an acceptable approximation of the true resistivity of the formation. As development of a field or area progressively enhances the knowl- edge of the lithologic section, the resistivity values of the electrical log take on greater significance, ultimately affording accept- able interpretations. The salinity, andG. E, ARCHIE 5S therefore the conductivity, of the connate water associated with the various produc- ing horizons may be determined with sufficient accuracy by the usual sampling procedure. Determination of the significance of the resistivity of @ producing formation as recorded by the electrical log appears, for the present at least, to rest largely with the application of empirical relation- ships established in the laboratory between certain of the physical properties of a reservoir rock and what may be termed a formation factor. It should be stressed at this point that numerous detailed laboratory studies of the physical proper- ties of the formations in relation to the electrical measurements in question are essential to a reliable solution of the problems dealing with reservoir content. ‘The purpose of this paper is to present some of these laboratory data and to suggest their application to quantitative studies of the electrical log. It is not in~ tended to attempt to discuss individual resistivity curves and their application. ‘The disturbing factors (borehole, bed thickness, and invasion) are discussed briefly only to indicate instances when they are not likely to affect the usefulness of the observed resistivity. ‘Resistivity oF SANDS WHEN Pores ARE ‘Enrmezy Fue wits Brive A study of the resistivity of formations when all the pores are filled with water is of basic importance in the detection of oil or gas by the use of an electrical log. Unless this value is known, the added resistivity due to oil or gas in a form cannot be determined. ‘The resistivities of large number of brine-saturated cores from various sand formations were determined in the labora- tory; the porosity of the samples ranged from xo to 4o per cent. The salinity of the electrolyte filling the pores ranged from 20,000 to 100,000 milligrams of NaCl per liter. The following simple relation was found to exist for that range of porosities and salinities: Rem FRe Gl where Ry = resistivity of the sand when all the pores were filled with brine, Re = resistivity of the brine, and F =a “for- mation resistivity factor.” In Figs. 1 and 2, F is plotted against the permeabilities and porosities, respec- tively, of the samples investigated. ‘The data presented in Fig. x were obtained from consolidated sandstone cores in which the cementing medium consisted of various amounts of calcareous as well as siliceous materials. The cores had essentially the same permeability, parallel to and perpendicular to the bedding of the layers. All of the cores were from producing zones in the Gulf Coast region. Cores from the following fields were used: Southeast Premont, Tom Graham, Big Dome-Hardin, Magnet-Withers, and Sheri- dan, Texas; also La Pice, and Happytown, La. Fig. 2 presents similar data obtained from cores of a widely different sandstone; that is, one that had extremely low per- meability values compared with those shown in Fig. x for corresponding porosities. ‘These cores were from the Nacatoch sand in the Bellevue area, Louisiana. ‘From Figs. x and 2 it appears that the formation resistivity factor F is a function of the type and character of the formation, and varies, among other properties, with the porosity and permeability of the reser- voir rock; many points depart from the average line. shown, which represents a reasonable relationship. Therefore, indi- vidual determinations from any particular core sample may deviate considerably from the average. This is particularly true for the indicated relationship to permeability. Further, although the varia- tion of F with porosity for the two groups of data taken from sands of widely different character is quite consistent, the effect56 of variations in permeability on this factor is not so evident. Naturally the two relationships could not be held to apply with equal rigor because of the well ELECTRICAL RESISTIVITY LOG AND RESERVOIR CHARACTERISTICS ity. Thus, knowing the porosity of the sand in question, a fair estimate may be made of the proper value to be assigned to F, based upon the indicated empirical 109; 550] o — 55 1 1 310 50100 00 Too 00010030100 Permeability, milidarcys Porosity Fic, RELATION OY POROSITY AND PERMEABILITY 70 FORMATION RESISTIVITY FACTOR FOR CO SOLIDATED SANDSTONE CORES OF THE GULY COAST, $00) 100 2 7 £50 + : 2 & 3 AL E10 I 5 1 or 05 10 50 6100 oir 030109 Permeability, millidareys Porosity Fio, 2-—RELATION OF POROSITY AND PERMEABILITY TO FORMATION RESISTIVITY FACTOR, NACATOCHL ‘SAND, BELLEVUE, LA. Permeabilities elowo.x millidarcy not recorded. established fact that permeability does not bear the same relation to porosity in all sands. From close inspection of these data, and at the present stage of the investiga- tion, it would appear reasonably accurate to accept the indicated relationship between the formation resistivity factor and poros- relationship fa or from Eq. 1, Rea Rom is} where 0 is the porosity fraction of the sand and mis the slope of the line represent- ing the relationship under discussion.G. E. ARCHIE 87 From a study of many groups of data, m has been found to range between 1.8 and 2.0 for consolidated sandstones. For clean unconsolidated sands packed in the laboratory, the value of m appears to be about 1.3. It may be expected, then, that the loosely or partly consolidated sands of the Gulf Coast might have 2 value of m anywhere between 1.3 and 2. Resistivity oF FoRMATIONS WHEN PORES which the oil or gas is distributed in the pores may be so different that these relations derived in the Inboratory might not apply underground. 10 S=Waler saturation ARE Partiy Fitep witn Brive, Rewanane Vows Brie Firiep wire On on Gas Various investigators—Martin,? Jako- sky,? Wyckofi,t and Leverett\—have stud- fed the variation in the resistivity of sands due to the percentage of water contained in the pores. This was done by displacing varying amounts of conducting water from the water-saturated sand with non- conducting fluid. Fig. 3 shows the relation which the various investigators found to exist between S (fraction of the voids filled with water) and R (the resulting resistivity of the sand) plotted on loga- rithmic coordinates. For water saturations down to about 0.15 or 0.20, the following approximate equation appli or = R=RS-* [4] For clean unconsolidated sand and for consolidated sands, the value of m appears to be close to 2, so an approximate relation can be written: VE (s] or from Eq. 1, [FRo sa fe a Since in the laboratory extremely short intervals of time were allowed for the establishment of the equilibrium conditions compared with underground reservoirs, there is a possibility that the manner in 7 References are atthe end of the paper. wD. cd B__Rtihvty ofl eg snd = es Resa pace aS fio, gRetaox or $10 8 Legend ond Dla $3, ] orgs can | oka | gator | Sand | NaCl per [Gas] tion lreerlteee| elec Heese et oe | [SESE HER |ure BEET. | Favosky| Feiabte | 20 appeox.| 01 os 3" Considerable encouragement on this point is established, however. For example, Eq. 4 appears to hold even though gas or cil is the nonconducting phase. Each probably assumes a different distribution in the pores, yet the resulting resistivity is not appreciably changed. Also, no great change is found in the average relation between the formation resistivity factor and porosity for changes in types of con- solidated sandstones. This indicates that even though the oil or gas underground may fill the pore space in a different manner from that in the short-time laboratory experiments, the relationship expressed by Eq. 4 should apply equally well underground. Basic Resistivity VALUES 10 BE OBTAINED. ay Estiwatixe FLum Conrent oF 4 SAND ‘The foregoing discussion indicates that the basic values to be obtained are: (1) the resistivity of the sand in question under-58 ELECTRICAL RESISTIVITY L0G AND RESERVOIR CHARACTERISTICS ground (R), and (2) the resistivity of the same sand when its pores are entirely filled with connate water (R,). ‘The first value can be obtained from the electrical log when all factors can be properly weighed. The latter may also be obtained from the log when a log is avail- able on the same horizon where it is entirely water-bearing. Of course, this is true only when the sand conditions, particularly por- osity, are the same as at the point in ques- tion and when the salinity of the connate ‘or formation water throughout the horizon is the same. In a waterdrive reservoir, or any reservoir where the connate water is in direct contact with the bottom or edge water, there should be no appreciable difference in the salinities through the horizon, at least within the limits set forth for the operation of Eqs. r and 4; that is, when the salinity of the connate water fs over 20,000 mg. NaCl per liter and the connate water is over o.r5. In depletion- type reservoirs, or when connate water is not in direct contact with bottom or edge water, special means may have to be devised to ascertain the salinity of the connate water. When it is not possible to obtain Ry in the manner described above, the value ‘can be approximated from Eq. 3, 6 and m having been determined by core analyses and Ry by regular analyses. Caxcutarion oF ConwaTE WATER, Poxos- ity AND Satinrry oF Forwation WATER FROM THE Execracat Loc ‘The resistivity scale used by the electrical logging companies is calculated assuming the electrodes to be points in a homo- geneous bed. Therefore, the values re- corded must be corrected for the presence of the borehole, thickness of the layers in relation to the electrode spacing, and any other condition different from the ideal assumptions used in calculating the scale, Consider a borehole penetrating a large homogeneous layer, in which case the electrode spacing is small in compatison with the thickness of the layer. If the resistivity of the mud in the hole is the same as the resistivity of the layer, there will be, of course, no correction for the effect of the borehole. If the resistivity of the mud differs from the resistivity of the layer, there will be @ correction. ‘Table + shows approximately how the presence of the borehole changes the observed resistivity for various conditions. ‘The third curve, or long normal, of the Gulf Coast is considered because this arrangement of electrodes gives very nearly a symmetrical picture on passing a resistive layer and has sufficient pene- tration in most instances to be little affected by invasion when the filtrate properties of the mud are suitable. ‘Tamte 1.—Effect of Borehole on Infinitely Large Homogeneous Formation Tg a isc Borate |“ Resiatvieg of Mud in Hoe geben ‘The values in Table x have been cal- culated assuming a point potential “pick- up” electrode 3 ft. away from a point source of current, other electrodes assumed to be at infinity, and it has been found that the table checks reasonably well with field observations. Checks were made by: (1) measuring the resistivity of shale and other cores whose fluid content does not change during the coring operation and extraction from the well; (2) measuring the resistivity of porous cores from water- bearing formations after these cores wereom resaturated with the original formation water, Adjustment due to temperature difference, of course, is necessary before the laboratory measurement is compared with the field measurement, ARCHIE 59 tivities. It is assumed that large shale bodies are present above and below the beds, at the same time neglecting the presence of the borehole and again assuming point electrodes. Resistivity, meters Tame 2—Effect of Formation Thickness, suai, No Borehole Present Seif potential a Trae Rataivty | Obened Resviy “Hag 5 ° GRETA |_ meet 3500 Say of ane & Is20 ; ote mh 7 a} afi 3 cet Tong norival | Ps ‘The correction at the higher resistivities a appears to be appreciable. However, cl in the Gulf Coast when the value of Ry For example, assume a friable oil sand f _ are whose true resistivity is so meter-ohms i eto and whose resistivity when entirely water- bearing is 0.50 meter-ohms; the connate water would occupy about 0.10 of the pore volume (Eq. s). However, if the observed value on the log, 65 meter- ohms, were used without correcting for the borehole, the connate water would be calculated to occupy 0.09 of the pore volume. ‘Therefore, although the effect ‘of the borehole size and mud resistivity on the observed resistivity readings may be appreciable, the resultant effect on the calculated connate-water content of the sand is not important. When the thickness of the formation is very large in comparison with the electrode spacing, there will, of course, be no correction to make for the thickness of the layer. However, when the thickness of the formation approaches the electrode spacing, the observed resistivity may be very different from the true value. Table 2 shows approximately what the third curve (long normal) of the Gulf Coast would read for certain bed thicknesses and resis- Fic. 4—ELECTRICAL L0G ov AN East Texas Diameter of bole, 734 ing mud re 3-4. 85*2.; bottom-hole temperature, appro, mately r35'r. ‘Tables x and 2 assume ideal conditions, 0 if the sand is not uniform, or if invasion affects the third curve, the observed re- sistivity values may deviate farther from the true value. The magnitude of the influencing factors, of course, will limit the usefulness of the observed resistivity value recorded on the log. Invasion of the mud filtrate is probably the most serious factor; however, as previously mentioned, it can ‘often be controlled by conditioning the maud flush for low filtrate loss. Fig. 4 shows a log of an East Texas well ‘The observed resistivity on the long normal curve for the interval 3530 to 3560 ft. is 62 meter-ohms, or, from Table x, approxi- mately so meter-ohms after correcting for the borehole. In this instance the mud resistivity at the bottom-hole temperature of 135°F. is approximately 2.2 meter-ohms.60 ELECTRICAL RESISTIVITY LOG AND RESERVOIR CHARACTERISTICS The interval is thick enough so that there should be no appreciable effect due to clectrode spacing. The formation is more or less a clean friable sandstone, so Eq. § can Resistivity mmafercohs Ee a ~aBe I Pt 5 Gra eure. \ 60 080 en yeaa 100 Vf abies Son core te ers 0 Fic. § Evrcrnicat 100 oF 4 sano iN East ‘Ware Pom yieto, Texas. Diameter of hole, 7% in.; mud resistivity, 1.7 at 80°F; bottom-hole temperature, 38°F. be used to approximate the connate-water content, The formation resistivity factor for this sand is approximately rs, using Eq. 2 where 0 = 0.25 and m= 1.8. The resistivity of the formation water by actual measurement is 0.075 meter-ohms at a bottom-hole temperature of 135°F. There- fore, from Eq. x, Re for this sand is 15 X 0.075 = 1.1 meter-ohms. This value checks reasonably well with the value recorded at 3623 to 3638 ft. on this log as vwell as on the many logs from this pool where the Woodbine sand is water-bearing; i.e., 019 to 1.5 meter-ohmns. The close check obtained between the calculated and re- corded resistivity of the water sand indi- cates that invasion is not serfously affecting the third curve. Solving Eq. s, the connate water of the zone 3530 to 3560 ft. occupies approximately outs of the pore volume. The accepted value assigned for the connate-water content of the East ‘Texas reservoir is 17 per cent. ‘An electrical log of a sand in the East White Point field, Texas, is shown in Fig. 5. The observed resistivity at 4o7s ft. is approximately 5 meter-ohms. The value of P for this sand by laboratory determination is 6. The sand is loosely consolidated, hav- ing 32 per cent porosity average. The resistivity of the formation water by direct measurement is 0.063 meter-ohms at the bottom-hole temperature of 138°F. There- fore, R. = 6 X 0.063 or 0.38 meter-chms. ‘This checks well with the value obtained by the electrical log between the depths of 4100 and 4120 ft, which is 040 (see amplified third curve). Therefore, invasion probably is not seriously affecting the third curve. From Tables 1 and 2 it appears that the borehole and electrode spacing do not seriously affect the observed resistivity ‘The connate water is approxi Other uses of the empirical relations may have occurred to the reader. One would be the possibility of approximating the maxi- mum resistivity that the invaded zone could reach (when formation water has greater salinity than borehole mud) by Eq. 1, where Re would now be the resistiv- ity of the mud filtrate at the temperature of the formation and F the resistivity factor of the formation near the borehole. By knowing the maximum value of resistivity that the invaded zone could reach, the limits of usefulness of the log could be better judged. For example, assume that a porous sand having an F factor of less than 15 was under consideration. If the mud filtrate resistivity were 0. meter-ohms, the resistivity of the invaded zone, if com- pletely flushed, would be 15 X 0.5 = 7.5. ‘Thus the observed resistivity values of this sand up to approximately 7.5 meter-ohms could be due to invasion.DISCUSSION or ACKNOWLEDGMENT Cooperation of the Shell Oil Co., Inc., and permission to publish this paper are gratefully acknowledged. The resistivity ‘measurements on the numerous cores were performed under the supervision of S. H. Rockwood and J. H. McQuown, of the Shell Production Laboratories. ‘REFERENCES 1, Mati. Murray and Giliaghasm: Geophysies Gul 1 yeti Bove Gents Org, sii 8 Era Pen Ai Grae is ae DISCUSSION (H. F. Beardmore presiding) S. W. Wnicox,? Tulsa, Okla —This paper recalls some of my own observations on the correlation of the electrical resistance of earth materials with their other physical properties. While Geophysical Engineer for the Depart- ment of Highways, of the State of Minnesota, from 1933 until 1936, I was primarily engaged in conducting earth-resistivity surveys pros- pecting for and exploring sand and gravel deposits. This work was done by two field parties using equipment of the Gish-Rooney type, and was carried out in every part of the state, both winter and summer. In brief, when a sand or gravel prospect was discovered, in any way, it was detailed by the resistivity survey to outline its extent and to locate test holes for field and laboratory sample analysis. This survey consisted of a grid of “steptraverses”” of one or more electrode separations, and for each an “iso-ohm,” or ‘equal resistance contour plan map, was drawn. Several thousand earth-resistivity readings ‘were taken over more than one hundred prospects. In some instances the test pitting ‘was started before the completion of electrical survey and their findings were soon available for checking any suspected correlation theory and confirming what subsurface factors were being measured and how effectively. From accepted earth-resistivity theory, it follows that within a definite sphere surround- iemograph Service Corporation. ing the electrodes the apparent resistance ‘measurement is uniquely determined from the specific resistance and position of each and all of the particles making up the sphere. Any rational interpretation of these spparent resist- ance measurements is possible only for the simplest combinations of particles and their specific resistances. Fortunately, soils, sub- soils and subsurface rocks, with their embodied fluids and gases, vary greatly in this property among themselves. For example, clay appears to have an average specific resistance of approximately 0 to x50 foot-ohms, whereas for sand and gravel the specific resistance is roughly from 2000 to seco foot-ohms, The Important feature is the great absolute differ- ences in resistance, consequently a resistance profile across a buried lens of sand or gravel sur- rounded by clay produces a striking response. Tn spite of the amount of control available and the freedom for selecting various electrode intervals, no reliable quantitative predictions ‘could be made that were not related to bound- ary surfaces. The probable depth to the first discontinuity—namely, the clay-sand contact could be determined fairly accurately if the thickness of the sand body was considerable. When the depth to the sand was known from independent data, or could be assumed to be constant, it was possible to predict its thick- ness. If both were knowa, a good guess might be made regarding the depth to the water table; and, in addition, if all these were known, 1a surmise could be made about the quality of the sand; ie., whether it contained organic material or was weathered. Perhaps if the degrees of control were suficient the porosity of the sand, its grain size, or even its temperature right be predicted, Tobserved that few of these variables, even the ones that generally contribute to the bulk of the readings, could be quantitatively separated without additional independent data; therefore my interpretation was necessarily empirical and based on experience. Fortun- ately, in sand and gravel prospecting the economically most importent factors contribute their eflects in the same direction. A high apparent resistance indicates either a thin body of highly resistant gravel near the surface, or & thicker one overlain with more clay stripping. Clean gravel is more resistant than weathered, and hard gravel more so than soft.62 ELECTRICAL RESISTIVITY LOG AND RESERVOIR CHARACTERISTICS In practical terms, I found that an apparent resistance reading of soo foot-ohms for a oft. electrode separation recorded over ‘ground or glacial moraines of southern Minne sota reliably suggested a deposit of sand or gravel worth further investigation. As a matter of record, prospecting in the part of the state where these materials are very scarce, less than 3 per cent of the test holes located on the ‘geophysical information failed to yield granular materials of commercial quality and quantity for at least highway subgrade treatment. ‘Varying the electrode interval gave additional confirmation as to the thickness of the deposit and very litte else. In connection with our feld work, we made extensive laboratory studies, attempting to ‘work out the relation between the moisture content of sand and gravel and its specific resistance. These apparently simple experi- ‘ments were not of much help in clearing up my field interpretations. Several variables were very hard to control in the laboratory. ‘The analogy between this type of earth- resistivity mapping and electrologging is close. ‘The frst measures electrical impedance along a surface generally parallel to the bedding planes; the latter, up a borehole more or less perpen- dicular to them. The same general limitations and possibilities appear to be common to both methods. Obviously, controls for checking are easier to obtain for plan mapping than for well logging within the depth of effective penetration, ‘My interpretation problems appeared to be essentially similar to those of electrical well logging where the operator, after observing the character of the resistance and the self- potential curves, tells his client whether pipe should be set. The accuracy of his prediction is based largely on experience and not on slide- rule calculations. Mr. Archie's paper suggests an experimental attack for expanding and improving the interpretation technique of electrical well Jogging. Any contribution of this nature that increase its effectiveness is of great value to the petroleum industry. T offer my own experi- ences and observations to emphasize that he has tackled a dificult research problem and wish him luck. Dr. A. G. Loouis,* Emeryville, Cali —In the laboratory, we take into account the var tions in measured resistivities of sands and tap water by finding out the cause ofthe variations in resistivity. That is, ifthe tap water itself varied from day to day, its electrolyte content aust vary from day to day and chemical analysis would indicate the change. If sands did not give consistent resistivity readings, the character of the sands (in other words, the formation resistivity factor) probably changed or the kind and amount of water contained in the sand must have varied. + Shett De lopment Co.ET OF THe Sines ume eomheneaae, ama INTRODUCTION TO PETROPHYSICS OF RESERVOIR ROCKS‘ 6. E, ARCHIE ‘Houston, Texas ABSTRACT ‘Theres need fora term to expres the pyc of rocks, Itshould be related to peteology much as zope a rated to geology. SPeteophyncs” i eeggsted asthe term pertaining to the payne of Fst rock pes, wheres geoph peas othe pi of ge rok stems compose ‘Fas petropyncof eer oss seed hse, This scjet i study ofthe physical pro- vane ali ett prec al dhe Ge eat ie Bes ea made es, suchas Permeability, epdlary pres ‘Son sturadon, ld prope, cel raat ae or natura poten) and radii” ‘of different types of rocks. Tihese properties have beea vestigated separately and in relation, one to ‘nother, paral as they pertain othe detection and eralustion of hydrocarbon beating inj. OENERAL ‘This paper is concerned with rocks and their fluids in sil, particularly for the detection and evaluation of hydrocarbon deposits penetrated by a bore hole. Fundamentally, therefore, the study is one of pore size distribution and fluid distribution of each phase (oil, gas, water) within the pores of the rock. ‘The subject must not be limited to permeable rocks containing hydrocarbons, but should include the impermeable layers and permeable layers containing water-as well. This must be done in order to distinguish between them. Iscussion Rocks are heterogeneous. Therefore, the pore-size distribution as well as the fiuid distribution within the pores may be complicated, particularly from the microscopic point of view. We are dealing with a heterogeneous material, to- ether with a great many varying conditions within this heterogeneous material. ‘This, no doubt, is the reason for eatly belief that a quantitative approach to the problem might never be attained. When rocks are studied from a macroscopic viewpoint, however, 2 definite continuity #s found. A correlation of rock properties has resulted in the discovery that definite relations or trends exist between rock characteristics. If pieces of rock representing each transition phase of 2 formation are studied, definite trends are noted. No matter how thoroughly a single piece of rock is studied, even on a microscopic scale, it is not possible to predict the properties of a for- mation as a whole. This should not be taken to mean that fundamental research on a microscopic scale is not of great importance in the study of rock porosity. ‘Though permeable rocks ate, by natuze, heterogeneous, their characteristics aa Beate blore Houston Gono Soy, Septenber re, rp. Manet received, Oc 15,190 2 Shell OU Company. The writer expres he appreciation to the Shell Oil Company for isin to publok tls paper and to the aay author bavingpublabed information of the 3 ‘over hete, An extensive bibogaphy would be aceded toler the reader tall pertinent arte. os44 G. E. ARCHIE follow definite trends when considering a formation as 2 whole. Relations between the basic rock pore properties may be indicated somewhat as follows. Type of Rock _ Pore Size Distribution Porosity _—— (Connate) Permeability. “—$—~ capitiary Pressure (@ore Interconnection) ‘Type of rock, as here referred to, is a formation whose parts have been de- posited under similar conditions and have undergone similar processes of later ‘weathering, cementation or re-solution, as, for example, the upper Eocene Wil- cox, lower Frio, Woodbine, or Bartlesville in a particular area. ‘The connecting lines are meant to portray the fact that a specific formation or rock type will have certain effective pore-size distributions which willprodiice a particular family of cepillary pressure curves. The pore-size distribution con- trols the porosity and is related to the permeability and water saturation. Fur- ther, a certain rock will exhibit a relation between porosity and permesbi ROCK TUPR-POROSIIY-FERMEABILITY RELATION A broad relationship exists between porosity and permeability of a formation. Figure x shows a plot of the measured values of these properties of cores from the upper Eocene Wilcox sandstone at Mercy, Texas, and the Nacatoch sand stone at Bellevue, Louisiana. The scattering is great, but it must be remembered that the only reason a trend exists at all is that the formationas a whole wasde- posited under a similar environment; individual parts (local environment) may differ from the whole. The trends shown in Figure x may be represented by 2 line. The average trend for different formations is shown in Figure 2. Note the paralleling of trends of different formations. (Only limited data are available in the high permeability range; therefore the lines have not been extended beyond the values shown.) Rocks indicated on the left, those having low porosity for certain permeability, have relatively large pores, for example, the oblitic lim stone. Those on the right have a high porosity for the permeability, indicating ‘a smaller pore siz, for example, the poorly sorted shaly sands of the Gulf Coast ‘and the poorly sorted, shaly, calearcous Nacatoch sandstone. It is interesting to find that the increase in permeability with increase in porosity is of the same general order for many of the formations. An increase in porosity of about 3 per cent produces a ten-fold increase in permeability. This is striking in view of the fact that the formations are widely different. For extmple,PERMEABILITY ln Wd ET, wlf 6 ob fs al Fics. x and 2—Average relation between porosty and penaeaiity for dierent formations,946 G. B. ARCHIE the Smackover limestone, on the average, has a permeability of x millidarcy when the total porosity is only 7 per cent; the Nacatoch sandstone for the same permeability has an average total porosity of about 28 per cent. It will be noted that some of the treads are not essentially parallel with the 2) Paluxy sandstone is believed to be of dune origin and has grains of uniform size; as a result, the average pore size is larger than for water-deposited sandstones and the permeabilities are higher, particularly in the lower range of ity; (2) lower Wilcox has probably undergone considerable change since deposition (cementation of the fine pores and solution causing larger pores) ard has a steeper trend than the other formations. ‘Ie appears that as diagenesis continues, the porosity-permeability trend moves toward the left. Compare the loose Gulf Coast sand trend with the harder sand- stones where the smaller pores are absent and the larger pores are: possibly en- larged. ‘Figure 2 indicates that an approximation of the probable permeability can be made when the porosity of a formation is known. It must be remembered, however, that the relation is an average of a large number of data from a forma- tion and, when applying the relation to the same formation at a different locality, the assumption is made that the rock structure is similar. If the type of void structure throughout the formation is similar, the relation is, of course, more significant, However, even at the same location, certain parts of a formation may deviate considerably from the average trend because the pore structure or type of rock changes appreciably. An example of this is shown in ‘Table I. This small interval is purposely chosen from several hundred feet of Eocene Wilcox formation, TABLE 1 BE PE on neo os ‘The interval r0,458-64 feet has a porosity of about 14.5 per cent and a per- meability of about 200 millidareys, wheress the interval immediately above, from 10,450-s8 feet, has a porosity of about x6 per cent and only x maillidarcy permeability. The former interval is clean and well sorted and therefore not typical of the lower Wilcox. Therefore, this relation of porosity versus permes-INTRODUCTION TO PETROPHYSICS OF RESERVOIR ROCKS 947 bility is only valid in a general way fora formation as a whole, and may not apply to 2 small integral part. The average data shown in Figure 2 does, however, help one visualize relative pore size of different types of formations. ROCK TYPE-CAPILLARY PRESSURE-CONWATE WATER-PERMEABILITY RELATION Air permeability of a dry rock sample is a measure of the average contributing efiect of potes of all sizes, This average isnot suicient for conpleteaalsis because the pore size must be considered inorder to obtain permeability “si” (cclative permeability) and fluid distribution. Capillary pressures of rocks help to analyze this average figure by obtaining what may be called effective pore- see distbution, Capillary pressures af rocks have been discussed in the literature regarding connate water, and recently the relation between permeability and capillary pressure has been presented. In order to tie the captioned relation more closely to rock type capilary pressure carves of diferent types of permeable formations ae presented; the pore structures are due toa wide varetyof geologic procesey, that is, (a) sedimentation with little alteration, (b) alteration by solution, (c) redeposition or cementation, A comparison of capillary-pressure curves for dif- ferent rock types and for the seme type with varying permenbiities may thea bemade, “The following types of permeable formations are presented. Sree eee eer eres * aya cewee (ar Bae Wee gra Meg eas) PP 1 Fale andiene poy sorte gaa sal and latins; comparatively hgh pores Tor a 2. ponnipoamagpiaaneinanietnd © SEDUEES (Devons chy nesian, Wek Tema) TNSTCY Mab Poesy fs ee In types x, 2, and 3, the pore size shape of pore, and interconnection of the pores are controlled mainly by the original deposition, altered little by later cementation or solution. Therefore, pore structure is due mainly to the manner jn which the fragments were deposited, that is, the sorting action and packing of the grains due to wave action and later compaction. The pore structure in type 4 is due mainly to later cementation and solution; in other words, the pore structure originally due to sedimentation has been altered considerably. The ef- fective pore structure in type s is due almost entirely to solution with some re- deposition. The origin of 6 is controversial. ‘Families of capillary pressure curves for each of these types are shown in 1 W.R, Purell "Capi ny Pres: Tet Macnee USog Mercy andthe Calesatin ot Permeabilty Pheetory eel Techy 39 (949). Ts P25tana’ BULK VOLUME OCCURED BY WERCURY (sonar aunssaua aveTievs Fro, 3--Familes of caplry presure curves for some sandstones,ros as. PERCENT:BULK VOLUME OCCUPIED BY MERCURY io, 4-—Fanilis of capillary presure curves for some limestones; alo aie curve.950 G. B. ARCHIE Figures 3 and 4. ‘These curves depict results obtained by injecting mercury into the cores. In order to show porosity in the graphs, the abscissae are “per cent bulk volume occupied by mercury.” For example, it may be noted in Figure 3 that for sample No. 3379SA, permesbility x625 millidareys, the curve, with in- creasing pressure, approaches the abscissa 27.7, its effective porosity. These charts representing data from suites of cores show how the capillary-pressure curve, permeability, and porosity are related. Certain general conclusions can be drawn by comparing the graphs. Tt can bbe seen that all of the samples having appreciable permeability exhibit « plateau, for seat, and a steep slope (jdealized curve, Fig. 4). Examination of the capillary pressure curves will reveal that two straight lines (representing plateau and steep slope) can be drawn in and define the curves fairly well. The angle formed by extending these characteristic lines is useful. This angle “A” increases as the permeability of each type of rock decreases until one line appears and the plateau disappears. ‘The families of curves exhibit a striking similarity for the same permeabilities, regardless of type of formation or origin of porosity. Of course there are differ- ences; for example, the rocks with high permeability for porosity exhibit mitch steeper steep slope (for example, San Andres limestone) because of aless amount ‘of small pore space. Formations with comparatively low permeability for porosity, however, exhibit a more gentle steep slope because of the many small pores (Nacatoch and Devonian limestone). The curves of the Devonian limestone, which approaches chalk in texture, differ most from the others. It hes a propor- tionally large amount of fine to very fine pores. Photomicrographs of specimens of each type of rock investigated are presented in Plates I and II in order to show the visible difference in rock texture and pore structure. Notice the large pores in one of the San Andres limestone specimens, the medium pores in the sandstones and the very fine pores in the Devonian limestone and shaly sandstones. The rock textures include crystalline limestone, medium to very fine granular sandstones and very fine granular, almost chalky, Devonian limestone. (OMTER ROCK PROPERTIES RELATED TO FORE-SLZE DISTRIBUTION If it were possible to measure the fundamental properties (exact pore size and fluid distribution) én site of formations penetrated by the bore hole, the volume of the hydrocazbon in place and the productivity of the layer could be calculated. However, itis practically impossible as yet to get a direct measure- ‘ment of the factors, porosity, permeability, hydrockrbon saturation, and thick- ness of the layer, in place, by coring or other physical measurements. ‘Complete recovery of cores can not be assured and all permeable cores te- ‘covered are invariably contaminated with the drilling fluid, or Suid conditions hhave changed because of pressure and temperature changes on bringing the core to the surface. Therefore, we must resort to indirect measurements, such as elec-9st INTRODUCTION TO PETROPHYSICS OF RESERVOIR ROCKS XSI NOLLM0S-3Y S7EVYICISNOD CNV G31N3W39 ATIAVZH''SNOLSONYS GUVH SVX3L NVGINSHS ‘XOOTIM 3N3003 Y3MO71 Yod He's! aneo "wos % eur “ami wos se XS| SNOLSONVS ATVHS OL ONIGVYD SNIVYO G3LUOS AT8O0d “BNOLSON S$ 2 SVX3L ADMAW '¥O0TIM 3NBNO3 Y3ddN “an ovzi svivaNAGATOCH SANDSTONE, BELLEVUE LA. I5X HIGH POROSITY FOR GIVEN PERMEABILITY a0 uo, 384% POR, news, 3es% FoR, Lao, er SAN ANDRES LIMESTONE, WEST TEXAS IOX ‘erm, 109 %Pon, 2M, Pa1%PoR. temo, us %ron. DEVONIAN CHERTY LIMESTONE, CROSSETT TEXAS 10X COMPARATIVELY HIGH POROSITY FOR GIVEN PERMEABILITYINTRODUCTION TO PETROPHYSICS OF RESERVOIR ROCKS 953 tical resistivity, sll-potential, and neutron resetion, which ean be recorded in 4 bore hole filled with mud. In order for these indirect measurements to be useful they must be directly related to the physieal properties desired (poresity, per- reabilty, and Mud saturation). Correlation of the indirect with the actual phys- jeal properties from a macroscopic point of view hy actually testing mumerous cores has led to the discovery that definite relationships or trends do exist. ROCK TYPE-SELF POTENTIAL-CKOUND WATER SALINITY-DERMMAMILITY RELATION Discussions in the literature have pointed out that the self potential is com= posed chiefly of two components: (1) the flow potential, and (2) the chemical potential ‘The flow potential is thought to he a smaller part of the total where v Fpesn waren (a) ve FIGURE 5. Zier waren co Freon waren (2) lexturement af shale potent the ground waters are,very saline. Actually, the flow potential in x well ean be measured by recording the S.P. with pressure on the well head. ‘The chemical potential has been expressed by the equation RK SP. = — K log: ~ Re where K depends mainly on the type of impermeable rock, « ratio is that of the mud filtrate Ky and formation water As, ‘The difference be- tween the $.P. recorded opposite an impermeable shale, for instance, and that recorded opposite an infinitely permeable clean sandstone may be expressed by l the resistivity insite to Subauetace Studies hy Meuns uf ec- «schuniorgr a Laan, Nee C men E., Vol. 110, Geuphysical Prospecting (1934), P. trical Measurements in Drill Holes” Tren, 7 +3, R. J. Wyle, “A Quantitative Analysis of the Klcetvacheial Comprnent of the S.0. Curve)" Jour Ped Fed, Wale 1, No.2 Uanuary, 1949), Ps 47954 GR, ARCHIE jon, ‘This di the laboratory by phieing sxlt water at one end of a piece of shale and fresh water at the other end, Actually, in the Inhorstory it is difficult to measure the valtage thus generated by placing the electrodes in the salt and fresh water hecause of electrocte potentials; there~ fore, the set-up shown in Figure § is used. 14 will he seen that each of the elec- trodes is now in the same solution, thus cancelling the clectrade potential so ference ean be measured Vothersome in laboratory measurements. Note further that the potential re- cording at 1 hale potential” plas the liquid contact potential (2/8). ‘The recorded potential is the result of: BIC fresh water-shale contact C/D shate-salt water contact DB salt water-feesh water liquid junet Shale cores obtained from different formations have heen tested hy using df= erent amounts of dissolved salts in the salt water, [Cappers that A in the fore- sanox ie. Qualitative eat betes tential aad sande svt ple, shales restuallow fon takes on different values for different shales. Far ex ‘of the order of bo, while that of going equa dof the Eocene Wileas have a Av Pennsylvanian shales may When the permeability of the formation is not high, however, the S.t corded apposite it ina hore hole is somewhat less Ln the ease of a snd and shale section, the relation between type of formation ancl self potential may pressed by Figure 6, placed qualitatively, at least, with x permeability scale. 7-A shows the 5.1. curve recorded opposite a section of the formation. The interval £0,800 to 11,800 feet is a sandstone al shale section. TheINTRODUCTION TO PETROPHYSICS OF RESERVOIR ROCKS — 955 elation between auf potential and permeability of Bocene Wilenx sandstone, Mercy, Texas. sandstone is more or less the same type throughout, being poorly sorted, well ce- mented, and hard. Increasing amounts of argillaceous material are noted decreasing permeabilities. ‘The formation water is the sume throughout this, section, the mud in the bore hole at the time of the survey was uniform, and all the formations are water-bearing. The formation was extensively cored and the average permeability is indicated opposite the layers from which suffi were recovered and analyzed. “Phe magnitude of the selfpotential for the various layers is plotted versus the permeability in Figure 7-B. Note the apparent semi-logarithmic relation (simular to that shown in Figure 6).956 G. B. ARCHIE It may be noted that there is considerable scattering of points and the rela- Letween S.P. and permeability is qualitative in nature. [ROCK ZYPE-ELECTRICAL RESISTIVITY-TOROSITY-WATER SATURATION RELATION ivity of rocks when the pores are saturated with brine ‘The electrical resi may be expressed by: Kom PRe ‘where Rois the resistivity of the rock when saturated with brine (over about x0 grams per liter dissolved salts): F is the formation resistivity factor; and Ru is the resistivity of the brine. F is found to be related to the porosity and the type of rock. Fm frorRe= Rf where fis the porosity of the rock and m is related to the type of rock. ‘Again it must be remembered that this equation represents x trend or an average line through a number of measured values (Fig. 8 and 9). RELATION OF POROSITY TO FORMATION RESISTIVITY FACTOR ‘Fo, 8--Kesstvty factors of various sandstones serous porsity. ‘The electrical resistivity of a rock when hydrocarbon-bearing may be ex- pressed by: R= Res where S is the fraction of the voids filled with bring; X is the resistivity of the hydrocarbon-bearing rock; and s depends, apparently, on the type of rock. 4G, E, Archie, “The Electrical Resistivity Log as an Aid in Determining sine Reervoit Char- acteristics” Trane, A.TALE, VOL. 148 (943), P- 34.INTRODUCTION TO PETROPHYSICS OF RESERVOIR ROCKS 957 RELATION OF PERMEABILITY TO FORMATION RESISTIVITY FACTOR sestsrnary excron rc ue. g-—Resstivity factors seus permeability. Laboratory results on artificially saturated loose sand packs indicate n=2, (Figure 10-2), whereas measurements on consolidated sandstones naturally saturated (as withdrawn from the well) indicate m=1.9 (Fig. 10-b). The fore- going equation holds for water saturations down to about xo per cent. Below this value there is some indication that the interconnection between the water is no958 G. B. ARCHIE [RELATION OF WATER SATURATION TO ELECTRICAL. RESISTIVITY nomen, RELATION CETERMEO 8 USING WATURAL RESISTIVITY OF GORES “ Fis, 20-—Relation of water saturation to electrical resistivity. longer uniformly continuous and the resistivity increases more rapidly than in- dicated by the equation. ‘Actually, therefore, it may be said that the electrical resistivity of « hydro- carbon-bearing rock depends on the porosity, brine saturation, salinity of the brine, and type of rock. [ROCK TYPE-NEUIRON REACHION-IYDROGEN CONTENT-POROSITY RELATION Several papers have been written on this subject, and it appears that the rea tion from neutron bombardment of formations depends to a large extent on theirINTRODUCTION TO PETROPHYSICS OF RESERVOIR ROCKS 959 hydrogen content.? The effect of rock type is apparently not known. ‘The porosity rust be considered, for this determines the actual void space available to con- tain hydrogen-bearing fluids. The relative relation between total porosity and response to neutron bombardmeit, using present field methods, is illustrated in Figure rx. The limestone zones are oil-bearing, Each point represents an average ‘of several feet where the porosity was relatively uniform. The relation would be difierent if these zones were dry gas-bearing, for the hydrogen density of dry gas in the pores is much lower. eLATION Of TOTAL, ponosiTY neurnow ResPotse Nin mas eae Devnet ResroRseanartnan, wT io, 1.—Response to neutron bombardment. SUMOUARY OF BEDRRECE PROPERTIES Figure x2 illustrates diagrammatically how the various indirect properties are related to the desired reservoir properties. Some relationships are rigid and quantitative, while others are not easily predicted and are considered only quali- tative. After a qualitative relation is studied further, other factors may be dis- covered which, when incorporated in the equation, put the relationship o a more quantitative basis. For example, the qualitative relation between porosity and permeability is made more quantitative by introducing pore size or the capillary Pressure curve. The solid lines in the chart indicate what are now known to be quantitative relationships, while the broken lines indicate qualitative relation- ships. Figure x2 shows how intricately the various properties are related; all relations 1 RE. Fearoa, “Neutron Bombardment of Focmations,” Oil Weekly, Vol. 118, No. (1948), P styalso“Nucleonies" Vol No.6 June, 1949), p. 30.960 G. B, ARCHIE INTERRELATION OF PIYSICAL, PROPERTIES OF ROX _~ = =a a a Fro, 2 PETROPHTSICAL SYSTEM Sedimentary Leper copinety % Pressure Percenteges ! . fi Proper interfactet ey tenstons) content istivity ‘eetor Ground Water | ve ———— eet Resistivity Fro.INTRODUCTION TO PETIOPHYSICS OF RESERVOIR ROCKS 96x are tied to the type of rock. The diagram shows the many possibilities that can arie in attempting to unravel the interrelationships in order to evaluate a hydro- carbon accumulation éx sifu, In actual practice, further complications arise due to practical difficulties, economic considerations, and the personal equation. For example, the presence of the bore hole itself, its geometry, and the fact that the bore hole must be filled with mud in order to drill the hole brings up many prob- lems.* The layers penetrated are not infinitely thick; therefore, boundary effects (thin layer effect) must be calculated and applied to all indirect measurements.? Also, sud filtrate contaminates the permeable layers near the bore hole and it too must be considered.!* In fact, these practical complications are commonly ‘the most difficult to interpret in attempting to detect and evaluate deposits penetrated in a bore hole conctusrons In conclusion, # tentative petrophysical system, from the macroscopic view point, is presented in Figure x3 for illustrative purposes. The system revolves mainly around poresize distribution which defines the capillary-pressure curve, permeability, and porosity. The pore-size distribution does not necessarily define the type of rock, for actually several types of rock may have essentially the same poresize distribution. It is not meant that the mineral composition of the rock should be neglected in a study of these relationships. It must be recognized that the type of clay minerals present, for exemple, will no doubt play a greater role in future study. It should also be mentioned that lithologic description of rocks is important. In act, it should be broadened to express pore-size distribution as well as mineral distribution, It is felt that this is very important because drill-cuttings sample- logging (Formation type) really és an integral part of this outline. ‘The relations between rock characteristics should be thought of as trends. Actually, these may be expressed by mathematical formulae; however, the for- mulse can not be applied ina rigid manner, as is done when considering the prop- ‘erties of homogencous materials. It must be kept in mind that appreciable de- ations from the average trend may occur. The less uniform the date, the less igid will be the average relation, for some permeable rocks are more heterogene- ous than others. The generally uniform types of permeable rocks are sandstones, oilitic limestones, and the so-called granuler-eppearing dolomites. The less uni- form types are the so-called vesicular, vugular, or cavernous, or even fractured limestones and dolomites. * seklumbecgts Company pamphlet, Resitily Deperure Cures (May, 194). arg GpDal The SE, Loge Testi Ansys and Paces of Interpretation," Tron Sclumbergec Company pamptlet, op1p, 3541 FORMATION FACTORS OF UNCONSOLIDATED POROUS MEDIA: INFLUENCE OF PARTICLE SHAPE AND EFFECT OF CEMENTATION MR. J. WYLLIE AND A. R. GREGORY, GULF RESEARCH AND DEVELOPMENT CO,, PITTSBURGH, PA., MEMBERS AIME ABSTRACT. ‘The liv the systematic experimental determination of the formation actors of unconsolidated porous media. No experiments appear to have been made on the effect of increasing cementation on the formation facior of an intially unconsolidated. porous ‘Measurements have accordingly been made of formation fac- tore as function of porosity for aggregates of epheres In the porosity range 1256 per cent and of cubes, cylinders, duct find triangular prisms In the porosity range 3048 per cen ‘The reslts are examined inthe light ofthe theoretical equ Hons of Clerk Maxwell, Fricke and Slawinsk. Agsregates of unconsolidated spheres and beach sends have been artificially cemented with slice and the formation factor 7 relationships determined. A theory is outlined which lt obtained and which postulates ‘were derived. I follows also that for consolidated sandstones Co" where F is formation factor, ¢ fraction] porosity fand C and k are contants. INTRODUCTION AIL schemes of electic log fterpretation iplicly or ex: plicitly require the use of formation res they are now generally eae, formation factors. Theo both of the concept of formation factor and of the {formation factor data in elect log interpret ‘More recently it hasbeen suggested th . ‘of value in desribing fuid Aow in porous media. A full di ‘union ofthis problem together with adequate references has been given by Wyllie and Spangler? In spite of the theoretical and practical importance of the subject there appear to be rematkably few besie formation Hong inae SEE SELES rare eae Vol. 198, 1953 PETROLEUM TRANSACTIONS, AIME [At various times over the last century, attempts have been rade to compute the theoretical relationship between the for- tration factor and porosity of simple systems of unconsolidated pattcles. However the formulae derived are not all Identical ‘tnd the experimental evidence adduced in support of each Is ‘not wholly coavinelng. Most of the experimental work has been” carzied out on. sytem of unconsolidated spheres, although some attention has been given to systems of spheroids tnd evlindes. Other simple shapes have received nether the: frtiel nor experimental investigation. ‘The writers fel that 2 better understanding of the formation factors of complex systems of practical importance requires « frm foundation of Feliable ‘experimental data applicable to simple systems. Accordingly, they have reviewed the relevent erature. and have determined experimentally the relationship betwoen the formation factor and porosity of unconsolidated aggregates of particles of regular geometrical shapes. Some experiments have Also been carried ott to investigate the efiact of cementation ‘onthe formation factor-porsiy relationship. PREVIOUS WORK Although the term formation factor was not used, the rela tionship “between the resistivity of ealuion containing Aicleetic solids and the resistivity of a similer solution not containing solide fs an old and important problem, ‘The formation factor of a eystem of dispersed spheres as a fonction of porosity was considered by Clerk Maxwell. Max: well arrived theoretically atthe following expression: snd # fractional porosity. also examined the problem of the format factor of sysems of apheres and cylinders and concluded that for spheres Cleck Maxwell's expression gave a value of F ‘which, at low porosities particularly, was too small. However, 1031. 3541 FORMATION FACTORS OF UNCONSOLIDATED POROUS MEDIA: INFLUENCE OF PARTICLE SHAPE AND EFFECT OF CEMENTATION Lord Rayleigh's expressions were extremely complex and were bot supported by any experimental das che! generalized Clerk Maxwell's treatments he Incident. iy confamed Equation (1) for spheres His more general, ‘expeesion, applicable not oaly to spheres but abv to oblate nd prolate spheroids, was st0-9 # for spheres and 2<2 for spherolds. The greser ‘of the spheroid, the smaller Fricke computed I treatment appears to indicate F ee Spherical. Some experimental d find by Frike and Morse vo support bi tons. While the data appenr to substantiate the approximate securacy of Equation (2), they eannot be held to confm it to any high degree of precision. Most of the data presented by Fricke and Fricke and Morse lead only tothe conclusion that Equation (2) refects rather well the relative change is ie a change la the porosty of « randomly ‘Slawinskt" has ao discusced the relationship between F and 4 for aggregates of spheres both in contact and dispersed. The ‘contribution of Slavinski appears to have been rather neg. Teted, poribly because the mathematical approach be used iso simple and crude as to be supertcilly unconvincing. Nevertheless, Slawinsk presente more experimental evidence ‘than other workers to support his theoretical deductions For an aggregate of spheres which are in contact with one another Slawiaskl derives (13219-08219 6)" Boog ® Itolated measurements ofthe formation factors of uneonsol- dated systems of spheres and silica porosities have been repofted in the literature, Notable among these are those givea by Archie! by Klinkenberg’ and by Schofield and Dakshinarnurt* EXPERIMENTAL Preparation of Particles of Regular Sh “lcs or plese spheres which have high spbeiiy cbuained with a wide range of diameters. The seperation of felaively”nonspberial particles trom particles of high {hericiy may be achieved by making use of the greater ability Pike Inter pails to wll down aight nsined plane ‘The procurement of accurately sized particles of non ca shape presents» problem of reser diet. In thin work the problem was largely solved by atsing an lngenone taslon made tothe wnters by Forvet . Vera of the eng Stering division of thveleborstoriex Plate (Lute) rod af ‘Square circular, angular and other regular crowesetons tray be purchsed. Sie these rode re manufectared by an {xttason proses, their croe-sections ae extomely nfm peep for example, eat i then nesesary to cata fod {the agate erossetion Int lengths egua tothe length oa Side ofthe crosscetin. At the suggestion of Versa this was ‘xpediowly achieved inthe fllowing manner. A da was pre faved cootanig six aquare holes of «ze Just suliiet to pas the plate ods. At plate was mounted ebind the die its dence Irom i approximately equal tothe length of 2 Side ofthe required cube. Six rds were then posed through 104 PETROLEUM TRANSACTIONS, AIME the holes in the die and a Imounted ‘cecenteeally. Ti Insured that eutting could wceur di ing only approximately Lal the tine required by the sae ta make one complete rev0- lation. During the noncuting portion of the eyele, the rods ‘were agaia puthed agnint the plat, this action Teading to ae more eubes being ent asthe sav rotated. A jt of sir was used to blow the ext particles from between the die and plate ito 1 container. Uslixing this technique, i wat easly posble, by applying steady pressure to the six rods, to cut cubes at the rate of rome 20) a minute, The eut lengths, after inl djustments of the saw blade and back plate had been made, ‘were found to be reproducible to within 0.00] tn. Feathering Of the cut edges was small and was removed by tumbling the particle in belli! containing sawdust. ‘Cubes, cylinders and triangular prisms were prepared by this method using appropriate dies, Discs were obtained by stamping from plastic sheet ‘The dimensions and other properties of all regular particles used, together with information bearing on the properties of Some natural sands which were aso investignted, are given fa Table | Method of Measuring Formation Factor Formation factor measurements were carred out by packing the particles in plastic or glass cylindrical cells, Cells varied In see from e five in. diameter cylinder with a volume of 2818 cu em to a six mm diameter glass tabe with a volume of 69 feu em. The electrical resistance of the eslumn filled with slectriyte only was measured frst, followed by the resistance of the column packed with parices and saturated with the same electrolyte, The electrolyte used was either 0.1 N or 05.8 potassium chloride. The ratio ofthe two measured reitences? Eves the formation factor dicety ‘The majority of the formation factor data for the uncon- solidated: particles were obtained using a ve in. internal iameter Gylindrica cell designed to util ethers two- electrode or 2 fourclectrode resistivity messering system. Te ‘was found (Fig 1) that the mesturement oa tensive fot ‘tio of the diameter ofthe cell to that ofthe particle. A mi ‘um ratio of 25 wa found to be necessary. This messureieat was easily satisfied when the five in. dlameter cell was used. Poroities were determined (comm the weights aad densities of particles used and the knowa volumes ofthe cell, Tir rattan were Ast eel te 2% ie eine ane ane 16.1 — EFF OF CONTAINER/FARTCLE SIE RATIO ON FORMATION. Pxcroas measunes. Vol. 198, 1953MR J. WYLUE A ‘Table 1—Summary of Properties of Parti IND A. R. GREGORY mom eri a Deny Saher Ch Tee dO ent 283 “a0 Bitorene FE Sat Se ie le Claw O71 mm di, 158x104 emf 11D em! 250 1990 aise SHES BH ie ie es Sites MSS ie gts Geiss laisse ie ba cs Mee igs 2S ra cube Tate Ciagroias —Sezi0* int Rardin Las os sis? ides taste Oizo din, 200010 FLA? TSS Caen et bee : Plane OUSE- dy 492 x10%in? 60810 mn? = AM tt a aae A angola Poe cite O25 ngh 11262107 in? aba 10* in? LITT Teanga P L Rees ane te 26 ea Gall Cou Beach Send Sion SaneSind as 38 oa ear EA is os Control of Porosity When the forwation factor porosity reationsh consolidated aggregate is determined. itis desirable to cover the widest posible porosity range compatible withthe rete of contacts between particles. Craton and Frazee established Sx types of stale packing arrangements for spheres. These {asled from a cubic packing with a porosity of 47.6 per cent to thombohedral pecking with a pores Other definite geometrical arcangements yield poo Tig between thee two extremes. However, regular packings are fanisotropic and in this work it was particularly desired thet the packings achieved should slways be isotropic. ‘When three mum glass spheres were sed. it was found pos- sible to cover a porosity range of only four per cent (36. 40 per eent) by dint of jarring or vibrating the container in arhih the spheres were picked. By mizing spheres of diferent tines, @ wide range of porosities could be covered. Furnas” ‘method for computing the proportion of parses sf various szes needed to produce a mixture of minimum poros: iy! In general. he found that minimum voids were obtained the ratio between the sizes of particles used increased and, trithin lit as the numberof sizes increaed, although the ormalae developed by Furnas are not rigorous, they are approached quite closely in actual measurements made with phere. In thi work, the einizum ratio of the diameters of used war OIL. When a two-component mixture com: aed of 7L3 per cent by volume three mm diameter spheres 14285 per cent by volume 0.033 mm diameter spheres was ‘eed a mnimum porosity of 1625 per cent was achieved. This Figure ts in good agreement with the poswlater of Furnas However, when third component, spheres of diameter O11 rm were added tothe mixture (Ja the volumetric rato tree mim 561 percent, O7IL mm = 285 per eeat and 0033 mm BA per cent) the porosity was decreased further to 124 per cent: This i notin accordance with the views of Furnas ‘who suggested that when the ratio of diameters fs greater than 0.005 only two components are required to achieve & ‘minimum pores Th the ease of particles other than spheres, oly onesie was lable ad ths fact necessarily reduced the range of pore ity over which it was found possible to make measurements of formation factor. Nevertheles, as is noted below, the tread ofthe results for nomspherical particles parallels comparsble Vol. 198, 1953 PETROLEUM TRA\ for spheres. There seems no reason to believe that an increase in the range of porosities of the packings of any of the nonsphercal particles wed would lead to anomalous data. Cementation It was not possible to reproduce any naturel methods of cementation in the course of thls work. Nevertheless, while realiing the limitations of the method used, It was considered worthwhile to cement atifellly unconsolidated porous media 4nd to determine the relationship between formetion factor ind poresty in such cemented media. Two types of uncon: solideted packings were choren for the experiments; fl heres and a beach sand. By controlling the particle size Astribaton, it war posible to vary the original porosity of the uscontolidated packings of spheres before cementation ‘was commenced. The cementing material was introduced by frst saturating the unconsolidated percles with « hydrolyzed ‘olution of tetraathylorthsiicate. A steam of air was then Blown through the satursted aggregate. The evaporation of alcohol and the formation of hydrated sli gel between the particles. This gel served to bind the particles together. The porous aggregate containing the hydrated slica fel wae then heated to drive off the water: a. pure silica Aeport remained. A. repetition of thi series of ‘operations served to decrease the porosity to any desired gure. In prac: Ue, it was necessary to confine the unconsolidated material in 1 Lacite cylinder of one in internal diameter. Since the evapo- ‘ation of alcohol takes place preferentially atthe ends of the sore the silica deposit builds up rapidly at these two points. ‘To achieve entirely homogeneous cementing does not stem possible, bt «reatonable approximation to homogeneity may fe attained if a core about 10 in. log ie cemented and this core is et into sections each of which ts about 1% fe lone. Te'was found that sections which were taken from near the center ofthe original ong core were reesonably homogeneous. Formation factor measurements were made on the artical, cemented cores after they had been salurated with 3 05 8 KCI olution. The cores were clamped between platnized platinum ture electrodes which were backed with cotton compresces fturated with OS N KCI solution, Porosities were esleulated Ihy weighing cores oth dey and after they hed heen saturated swith KCI solation. The resistivity of the sturating KCL sol tion was obtained from messurements made with a calibrated INSACTIONS, AIME 105set * ponasitt oe Faarici SOGREGATES 10m p = 080 PERCENT dip cel, All measurements of cemented cores were made using the two-lectrode method, and all revsivtien were corected wise. Spheri In order to classify the shape of the particles, sphericity eal. ‘ulations were made for all the non apherieal particles used, ‘The spherilty of the lerge plastic particles of definite geo. tmetrical form wat determined according to the dofnition of ‘Wadell for true sphetiiy. For sand grains of iregular shape, Wadell's method” for determining the sphericity of particles was used. Fifty grains ‘were choten at random from representative sand samples for These mearorements. The results of the sphericity mesure fuente are shown in Table 1 and Fig. 4 ity Determinations Uneonsolidated Packings Fig. 2 shows s comparzon of the formation factors of sll unconsolidated particle aggregates examined for the porosity range 0.30050. Fig, 3 shows the data for spheres over the ‘wider porosity range 0.12056, In each figure curves corre Sponding to Equation (1) and Equation (3) azo showa, to tether with a curve for the relationship F =". The later txpression, based on Archie's erginl work, Is often quoted 1 the formation factor porosity relationship for unconsolidated ‘oroes media Considering first the data of Fig. 2, it is evident that the Clerk Maxwell equation, (1), underestimates the formation factors of spheres in contact. The same is true, but to leser ‘extent, of the Archie relationship F= 9" The Slawinki ‘equation, (3), appears to fir the experimental data. extremely ‘vel Inthe poroity range shown. ‘The data appear qualitatively to satisy the general predic. tions of the Fricke equation, (2). That i, at any porosity the formation factor appeats to be @ minlmumn for apheres. The dliference at any porosity hetween the formation factor of an faggregate of spheres and an aggregate of non-spherieal pi Hieles may be as much as 20 per cent. The shape of the ticle is clerly of Importance in affecting the magnitude of 108 PETROLEUM TRANSACTIONS, AIME FORMATION FACTORS OF UNCONSOLIDATED POROUS MEDIA: INFLUENCE OF PARTICLE SHAPE AND EFFECT OF CEMENTATION this diference. An attempt to correlate the experimental data ‘the Wadellsphercties is made below. In Fig, 3 it will be seen that the experimental date for spheres indicate that the Slawinski formula ives formation Eetors at porosities less then 0.20 which are too low. The Archie relationship fis the experimental data very well for o2s>e>a10. Fig. 4 shows a plot of Wadell sphericity aguist formation factor for all particles used. The porosity choven for plotting the figure is 0.35. Although the trend of the plotied carve indicate an increase of formation factor with decrease in ‘Wadell sphericity, there is 4 considerable scatter of pols. ‘The scatter is Teast for the sands which are, of courte, not composed of identical particles and which might, therefore, be presumed to have a tendency to ft an average curv. Consolidated Packings Fig. 5 shows the elect of comentation on the formation factor of packings of varying initial porosities and particle shapes. The very rapid Increase in formation factor which eo sues as porosity is deereaved by the presence of cement fs obvious. 4 was found virtually impossible to produce an entirely homogeneous distribution of cement in the cores. In conse. (quence the accuracy of the data in Fig. § is not as high as ted at Pe S . * tt. HO, 3— FORMATION FACTOR: POROSITY DATA OF UNCONSOUDATED POROUS ACOREOATE, @ == 12486 PER CENT Vol. 198, 1953,MR J. WYLLIE AND A. R, GREGORY i ° i sass i (No, 4—WaDele, SPHENCTY FORMATION FACTOR AT 9=35, UN ESieuiDateo PARTICLE AGGREGATES the best relationship between formation factor and poroity iss siaigt line on double loparthmle paper. The line passes through a nila point defined by the formation factor and porosity of the unconsolidated aggregate which was cemented in particular series of experiments. The data appest to ind fate that for the particular type of cementing ured the rls tionship between formation factor and porosity hes the form Fece o where k= 42 and C is a constant controled by the porosity Und formation factor ofthe original unconsolidted aggregate before It wat cemented, DISCUSSION ‘The data obtained forthe formation factor-poronty elation shipe of nconsolideted aggregates of particles have been thown in Figs 2 and 3. From a practical point of view, espe- tially in the application of formation factor data to electri fog interpreution, the following points seem worthy of em phasis. The formation facor of en unconsolidated aggregate Df particles is unquestionably a function of the average fof the particles constitting the aggtegste. At any_ porosity. In an jeotropic packing, the formation factor ie mininum for 1 packing of spheres. For aggregates of sand grains auch as fre encountered in nature, the formation factor-porosity rla- tionship will probably Ie somewhat above the Tine shown on Fig. 3 for spheres. The actuel positon of the line will be dependent upon the sphericity of che sand grains but the data ren forthe beach sand ray be regarded as reasonably typleal. The fact thatthe pots for triangular prisms do not pirallel those for particles of other shapes is almoet certainly Treonsequence of the tendency of aggregates of tviangl prisms to orientate as thelr porosity ts decreased No existing formula relating the formation factor and poos ity of unconsolidated packings of particles fis all the exper: rental dta found. For spheres Slawinsk's formula. Equation (G) is excellent when 0.25<9 0.05 tho cart) and one fo two points in the curved region, but these vsually. were obtained at ‘equilibrating solatica conductivities greater thaa 7 tambo ca! (solotion resistivities leas than about 1.4 cba m). Group 1 refers o all the above samples. While 2 large variecy of rock types is cepresented a Group 1, the distibution of Q, values is weighted heavily’ in che cange 0 < Qy £1 equiv/lier; this group contains about 94 percect of the sotal aumber ‘of samples. Another set of shaly sands (Group 2) was selected on the basis of a wider distribution of Q, values; with these samples, conductivity measurements at low equilitsatiag solution conduc vies (down to about 2.1 m sho cart) were ‘emphasized. Conductivity data for Group 2 cores fre probably the mose accorate and complete with Fespect to range of Q, and C,, values examined. Independent Q, determinations, together with GoGy measurements, were described for 2 selected set of sands by Hill and Nilburn.} Petrophysical and conductivity data for this set ace presented in Table 1, Similar mensuremencs were cattied out on 54 additional cores. A description of these samples ix given in Table 2; peerophysical characteristics and conductivity data are summarized in Table 3. Table 4 gives general characteristics and petro- physical data for che rensiniog samples used in the conductivity studies. Expetimencal procedures employed ia these conductivity measurements were described by Hill and Milburn. 0, Determinations Cation-exchange capacities (CEC) ax deterained by Hill and Milburn (Table 1) were measured cheomatographically using ammonium acetate solutions. Further CEC determinations (Table 3) were made using a procedure suggested by Mortland nd Mellor! This method “requires. cepeated equilibration of the crushed rock sample with Concentrated BaCl; solution and washing to remove excess barium ions, followed by coaductometciclard MgSO, solution, Results are ‘equivalear to the ammoniua acetate method. ‘Core porosities were calculated from the following measorements nade prior to the CEC determination (bulk volumes measured by mercury displacemeat and (2) grain yolunes determined by buoyancy wreighings under brine. croup 2 Shaly Sand Samples Shaly sand samples were selected which varied widely both ia CEC and in manner of clay distribu tion. The range of Q, values extended from 00 1.5, meq/nl. Cores where 0.3 Jo 9eNEINq eTQeTTAX 3OU posepsev0> soreusuazed Yoonseq seNTEA sro Orroroo —za0'0 Seat = "a9 “pas Ce ~ sBezony : - : 9670 y00"1 szs0"0 caco'o erro 100 eo 69670 25070 1c9"0 wrt at : : 1960 ozs0"o nzc0"0 wt st wero ese'0 su6"0 8790°0 9z50°0 woe * z90'0 S260 or ee ecto 9ce"0 956°0 wo 062°0 sze'0 S80" eo |e £190 ccero cs6"0 eso oz. o08"0 706"0 Cy zo 25670 eso at 959°0 £96"0 feo ut tz90 496°0 0 ot coco 298° fee 1 sr ays'o ee"0 so sis'o 606". azo ct ad 1960 isco, ero 00" so 980"0 eso ot 86"0 sooo. osc" esoro 8 082"0 esoro z26"0 scoro 9 596°0 s0z0°0 z1700°0 ors 20" ort0o L#200°0 gto e62'0 stso"0 '89200°0 00 sicto ava"0 1050°0 +79200°0 woot (ez1"0) (zero) e08"0 (er"0) 2200°0 GOD $#0200'0 svow00"O _zoRLO0"O _ze¥LOTOwzHR0"O SYESO'O Goany uo oye) (wa/oms) —_(uo/Seu) “ON uous oy S30) p30 esa oaE «COUT IN UOT DBE > as00 ALIALLONGNOD WBLYK 4O NoLLDNN Y SY oun {¥) BONYADNGNOD ANBIVAINOS MANIKYH JO (@) NOILDVad — @ 3T0¥LWinsauer ef al,23 can be used to relate Fe to porosity. ‘The two major tests of the proposed conductivity nodel, as discussed previously, are met. Conductivity data obtained for Geoup 2 cores are quiee detailed in the curved portion of the Cy-Cy, ploee (2.085 0, We assune that the mobility of the exchange ions is unaffected by the partial replace- ment of water. Then the conductivity of the counterions is given by BQ,/Sy (aho ea), Tr assumpeion was based initially on field evidence ‘which indicates that the SP deflection opposite Oil-bearing rocks is reduced when compared with the SP response opposite water-bearing sections of the sane ceservoir. Reduction of the total electrical potential is ausibuted roan increase in pemnselective membrane efficiency of the oil-bearing seads relative to the efficiency of the same sands at Sy, = 1. The diffusion potential across these sands comprises one of che components of the total SP response. This increase in the membrane efficiency fof the sands is associated with aa increased effective clay content which must be due to 2 decrease ia the volume of pore space filled with water. A similar suggestion war made earlier by Hill and Milbum® and later by de Witte)? The ‘equivalent Hill Milbura assumption is eee ay tecalling that the & parameter is approximately proportional t9 Qj. LABORATORY EVIDENCE FOR Q/ ASSUMPTION Laboratory data obtained by McLaughlin2 support the assumptica stated in Eq, 20, These data consist of diffusion potential measurements using shaly sands at conditions of water saturation equal t0 tnd less than unity. The respective membrane clficiencies are interpeeted from the measured em/'s using Fig. 13 of Ref. 3, and are expressed in cerms of the & and 5” parameters, {A group of plugs were cut from cores taken from forntion inthe Goble field, Texas Dillion oratns tate samples were messed Ero Nacl scloios hating concenwetone Se oglu! wal’ Hane ‘and frauen water, ftopectvely, These pocntals were measured bol SEE TT atd peeetideal al condigonse 25 ladieaced sores talues of bere deterniaed sg fem the en) mensucnens Silay, B tatued were dexertned fr res sande at 0° 1 Kiplee of bys bobs ta given tn Figs 7 Conaleriog EE eiperinnel ploblees neces ine eter trio Hf Sos Comparison afb values wih seucs of B8e Shot tcmmably good optcoent The assunption of Eq. 20 rules out the use of the Hill-Milburn equation to describe the conductance behavior of oil-bearing shaly sands. With decrease in Sy and the accompanying increase in Qj oF B the Minima in the calculated conductivity curves ate shifted to quite high values of Gy, well within the range of practical field values. Thus, depending fon the original values of band the particular values ‘curves will indicate increasing sand conductivities with decreasing Cy'e'(at constant S,, values) and will be physically meaningless. We further assume that che conductivity of an oil-bearing shaly saad can be described by aa ‘analogous to Eq, 9 for water-saturated EG FW) aD Here G, is the specific conductance of a partially water-saturated sand and Gis a geometic factor, being a function of porosity, water saturation and pore” geometry, but independent of clay content (Q,). Ge is expected to increase with decrensing Sy #ince the waterdfilled pore space is decreasing, faparc from complications due ¢o distribution of the two phases in the pore network of the rock. For clay-free or "clean" sandstones, F* aad G* reduce to os FIG, 7 — EFFECT OF OIL, SATURATION ON DIFFU- GION POTENTIALS OF SHALY SANDS.F = Glo = R/Ra = @3) and CH GIG = RIK. +h “The cesisivty ratio I is defined as T=R/R = G/G = G/F. +09) and according co Archie's second empitical elation ® T= S* (= G/F) + + 26) ‘The value of m may differ from the value of m ia Eq. “a By analogy, we define GAIA BM ee -@ where at fs he exponent for u hypothetical sock Thing te same pore geomeny aa the shaly sand, ber whoce clay cotene ls lenctive, Le at ie Indepeodene of Q,. For the limiting cases Qy = 0 or Cyn, Eg. of reduces to By. 26. CSabinng Ea, 9, 22, 25 aod 27, we obtain for the resiavey eco G+8q, e+ BQ 78)” Eq, 28 can be expressed in terms of water resistivity ef DERI Te SSRIS, where the units of water resistivity Ry and the term BQ, sre ohm m and (oha my, respectively. For , expressed as meq/ml or equiv/liter, the numeri~ cca value of B in Eq. 28 is + (23). Tes = 09) B= [1 - 0.6 exp (-0.77/%)14.6 + ++ Go) Figs, § through 11 demonstrate the effects of cernence inti FIG, § — WATER SATURATION AS A FUNCTION OF RESISTIVITY INDEX, WITH VARIABLE WATER RESIS- Tire. vatiation of clay content, water salinity aad pore geometry on the I'S, relationship, az predicted by Es, 29, We oservedac evn sal anne of have a masked effect on the tesistivicy Significant deparsices fiom the simple IS tela (Eg. 26) are noted with increasing water resistivity ac constant Q,; farther, log 5, is nonlinear with respect to log f. Asa final example, an illustration ofthe influence of clay in a reservoit sand on electric log interpre tation is givea, We assume s resistivity ratio of F's 4 obtained from log readings, and estinate the FIG, 9 — WATER SATURATION AS A FUNCTION OF [RESISTIVITY INDEX, WITH VARIABLE WATER RESIS- TwirY. i 3 f 5 if adedon Sos iawn BF ft atu eesti ta FIG, 10 — WATER SATURATION AS A FUNCTION OF [RESISTIVITY INDEX, WITH VARIABLE Q,. FIG. 11 — WATER SATURATION AS A FUNCTION OF RESISTIVITY INDEX, WITH VARIABLEvalue of n* to be 2. For a clean sand, these values Tead to an oil saturation S, = 0.50 by the Archie felation (Eq. 26). The influence of various degrees of shaliness associated with diffeceat water resistivities for this example is shown in Fig. 12. Oil-in-place estimates are increased considerably from 33 to 47 percent with increasing values of Q, above the clean sand estimate in certain fresh CONCLUSIONS ‘An equatioa has been developed, based oa a phys- ical model, char relates the electrical conductiviey of a water-saturated shaly sand to the water conductivity and the clay content of the sand. Clay contents. are expressed by the cation-exchange capacities per unit pore volume of the rocks which can be deterained by independent analyses. Various criteria established to test the model ae satisfied by shaly sand conductivity data. These data furcher perm determination of sand conductivities and clay Counterion mobilities down co very low equilibrsting clectolyce solution conductivities. ‘The model is extended to describe the conduc- tivigy of oil-beating shaly sands. An expression is developed which relates the resistivity ratio to water saturation, water resistivity and CEC per unit pore volume of rock. This equation generally predicts higher oil-incplace estimates than sre ‘beained from the usval clean sand relations. NOMENCLATURE 2 = dimensionless constant, Eqs. 10 and 11 Ap = constant in de Witte's equation! (Eq, 12 in chis paper Bp = constant in de Wieee's equationt9 (Eq. 12 in this paper b = Hill-Milbuea pacamecer® relaced to eff: clay content of shaly sand at Sy =1 Hill-Milburn parameter3 related wo effective ‘lay content ia oil-bearing shaly sand B = equivalent conductance of clay exchange PO on on 0,0, 08 6.08 mace cetiares FIO, 12—O1LAN-PLACE ESTIMATES AS A FUNCTION OF CLAY CONTENT OF SAND_AND FORMATION WATER RESISTIVITY (ASSUMED CONDITIONS, I= 4, mee cations (sodiun) as a function of Cy at 25C (Eq. 10), aho aq em meq? conductance contribution of clay exchange ‘cations to sbaly sand conductivity, ‘ohm! oho + specific conductance of clay exchange ‘ations, mho em? oaductance contribution of bulk electro ‘yee solution to conductivity of shaly sand, ohm of mh ecifie conductance of sand, 100 percent Saturated with aqueous salt solution, aaho eat vpothetical specific conductance of shaly ‘sand ac low values of C,, defined in Eq. 17, mho ca specific surface conductance of la ohn! o¢ abo Cy = specific conductance of a partially water- saturated sand, mho cm! pecitic conductance of aqueous elect Iyee solution, mbo ea"! D = self-dittusion coefficient, sq em see“! Foy = Hill Milbura formation resistivigy factor for shaly sands F = formation resistivity facto for clean send + « formation cesistiviy facto for shaly sand (Chis paper) F = taraday 2 = weight of clay per unit of rock pore volume, ‘ge clay/eu eo G = formation resistivity factor for partially Wwatercsaturated clean sand, Eq. 24 G* « formation resistivity factor for partially Wwateressturated shaly sand 1 = resiativiey index, Eq. 25 i = porosity exponent of lithology factor, Eq. 14 myaci ~ wolal concentration of NaCl solution, mol ‘NaCl/1,000 gm water N= aqueous electrolyte solution concenteation, ‘Nocwalicy, moles solute liter solution 1 = Archie saturation exponent for clean sands, 4. 26 int = saturation exponent for shaly sand Q, = surface concentration of clay exchange ‘cations, equiv en-? , = volune concentration of clay exchange cations, meq ml~) or equiv liter“ 0, volume concentration of clay exchange ‘cations in oll-besring shaly sand (Eq. 20), eq ml“! or equiv liter"? = ratio of measured to hypothetical shaly ‘sand conductivities at identical C's, Ey.17 Ry = sesistiviy of sand 100 percent sarurated With aqueous sale solution, cha =Ry ~ eesistivity of = partially water-sacucated ‘sand, cham S = specific surface area of clay, 29 ca/em clay = fesetion So, = residual oil saturation = fractional water saturation = geomettic cell constant for shaly sand, 9.2 y = geometric cell constant for shaly sand, Eq. 2 = constant defined by Eq. 10, mho em-t = ratio of equivalent conductance of clay ‘exchange cations to the maxiaum value of this equivalent conductance at higher sale concentrations, Eqs. 15 and 16 cil saturation mmarinum equivalent ionic coadvctance of sodium exchange cations associated with elay (25C), 29 €m equiv! ohm! (afea)’ = equivalene ionic conductance of sodium ‘exchange cations associated with clay (30) at Cy = 0, sq em equiv! ohm! Wf = mazinun sodium exchange ion mobility (250), 5q em vole! see"t 6 = porosity ACKNOWLEDGMENTS ‘The authors thank che managements of Shell Development Co. and Shell Internationale Research Maatschappij N-V. for permission to prepare aod present this paper. The assistance of F. Chow and PH. Massze and the valuable suggestions of other ‘Shell colleagues are gratefully acknowledged, He 1, Nyllie, M. RJ “Log Interpretation In Sandstone [Reservoira", Ceophysies (Avg., 1960) Vol. 25, 748 2, Lynch, E. J: Formation Evaluation, Heper and Row, New York 1982) 212. ‘4. HIM HJ. and Milbom, J. Dit “Btect of Clay and Water Selinity on Electrochemical Behavior of Reser: volr Rocks”, Trans, AIME (1986) Vol. 207, 65-72. 4. Mayer, Ke H. and Sievers, J. Fr La Pemeabilisd det embranes,. Lv Tatorte deta" Permeabilite Tonique™ Hels. Chem. Acta (1936) Ve. 19, 89. 5. Teorell, Ts: "An Attempt to Formulate « Quantitative ‘Theory of Membrane Permeability”, Proc. Soe: Expl Blot Med. (1938) Vol 33,282, 6 Archie, G. Eu: "The Electrical Resiatvity Log ax Sn: Aid in Determining Some Reservoir Charect ew", Tran, ADIE (1942) Vo 146, 56-57. A. Lal, T. Me and Morland, MM “SelCDitfton of Exchangeable los in Bentonite", Proc, Minth Not. Cont, Clays and Clay Mineral, Pergamon Pres Lid. ‘The Macmlan Co, Now York (1960) 229, 9 Gast, RG: “Applicability of Models to Predict Rates of Cation Movement in’ Clays Proc Sell Sek Boe. of America (1966) Vl. 30, 48, : USP Log Interpretation ia Shaly Soe, Pet. Ene. J Uune, 1968) 123136. 12, Saves, M. Cx, Jn Southwick, P. F., Splegler, K-§ fed Wylie, Ba Rhus Electrica! Condactenee of Porous Plage lon Exchange Resin-Soltion Syatema” Ind. and Eng, Chem. (1995) Val. 47, 2189, 13, Cremer, A. and Laudetoat, H.t “On the Inocond ue of Clay Gel, Soll Selence (965) Vol, 1% conductive Leudetout, Eecrique dos Gel Argilesx et Antsonetrie de leurs Eldments J. Chim: Phys. (1965) Vol, €2, 1185. ry Av and Laudelout, Hs USurtace Mobilition ff Catlons In Clays", Proc. Soil Sek. Soc of Amarin (1966) Vol. 30, 570, 16, Burge, H.C: “Daa Leltvermigen verdinnter iach ‘Physik. Z. (919) Vol thomaticnl Trestment ofthe Els ‘Capecityof Disperse Syston tele Condsetioty a Phys. Rev. (1924) Vol 24,535. 18, ‘A. and Laudelot, H.: Personal communica: tion Gay, 1967) 19, de Witte, A. Electric Lo 1957 89. 120, Patnode, H. W. and Wyllie, M. R. J. "The Pres of Gondactve Solids in Reservoir Rocks te a Factor tele Log Interpretation", Trans AMIE (1950) Vol. 185, 47-52. 21, Mortend, Me M. aod Melle, J. Le Conductonstric ‘Titeatlon of Soile for Cation Exchange Capacity”, Proc, Soll Si. Soc. of America (1954) Vol. 18, 363. 22. Grim, R. Et Cley Mineralogy, MeCeawHl Book Ce,, New York (1983), 23, Winrauer, WO. nd Witlans, i Sunde in Reletion to Pore Geonetry”, Ball, APC (1982) Vo. 36, No.2, 235. 24, Van Olpen, Hi and Waxman, M, Hus “Surface Cone Asctence of Sadism Bentonite tn Water", Proc, Fifth Nati. "Cont, Clays snd Clay’ Mincrats, NAS NRE (1958) Pub. 566, 61- 25, MeLaughli, W. A.t Pergonal communication. ‘saturation and Porosity trom st 1", Oil and Gas J. (Match 4,Theoretical and Experimental Bases for the Dual-Water Model for Interpretation of Shaly Sands C. Clavier, see, sehlmberger Techni Services 6G. Coates, Pe, ehlumberger Well Series J. Dumancir,* Pe, Sttmberger Wel Services Abstract ‘A simple petrophysical model proposed by Waxman and ‘Smits (WS)' in 1968 and Waxman and Thomas (WT)? in 1972 accounts for the results of an extensi perimental study onthe effects of clays onthe resistivity of shaly sands. This model has been well accepted by the industry despite a few inconsistencies with experimental results Is proposed that these inconsistencies resulted from the ‘unaccounted presence of salt-ree water atthe clay/water inerface. Electrochemistry indicates that this water ‘should exist, but is there enough to influence the results? ‘Both a theoretical study and reinterpretation of Waxman- ‘Smits-Thomas data show that there i. "The corresponding new model starts from the Waxman and Smits concept of supplementing the water conduc tivity with a conductivity from the elay counterions. The crucial step, however, is equating each ofthese conduc tivity terms toa particular type of water, each occupying. ‘a representative volume of the total porosity. This ap- ‘roach has been named the ‘“dual-water” (DW) model because of these two water types—the conductivity and volume fraction of each being predicted by the model “The DW model has been tested on most of the core data reported in Refs, | and 2. The DW concept is also ‘supported by log dts? and has been successfully applied {tw the imerpreation of thousands of wells. However, the scope ofthis paper remains limited tothe theoretical and ‘experimental bates of the DW model. ‘The Petrophysical DW Model ‘The purpose ofthis model is to account forthe resistivity bbchavior of clayey sands. For petrophysical considers- tions, a clayey formation is characterized by its total porosity, ¢,; its formation factor, Fo; its water satura- tion, Siri bulk conductivity, Cy; and jts concentra tioa pet unit PV of clay counterions, Q,. The formation ‘behaves like a clean formation with identical parameters 1» Fo, and Syy but containing a water whose conduc te + Ces differs from the Bulk formation water.** [Neither the type of clays nor their distribution influences the results, ‘Since the formation obeys Archie's laws, The clayey sand equivalent water conductivity, Cues 620 be considered a mixture of two water. 1A clay water surounds te clay pales bu as a condacivigy independent of he type and amount of lay. us conductivity, Coys comes excsively from the clay counterions. The volume faction of clay water, Vows is diealy propotional to the counterion concentration, Vow =VQQiB rn cen .@ where vg is the amount of clay water associated with 1 unit (meq) of elay counterions. 2. The water further away ffom the clay is called far water. Its conductivity, Cy, and ionic concentration cor- respond to the salinity of bulk-formation water. The volume fraction of this water, Vey is the balance be- ‘ween the total water content and the clay water. @) ‘The implicit assumption is thatthe far water is displaced preferemially by hydrocarbons. "The two water conductivities are combined as their volumetic weighted averages to give the equivalent ‘uid conductivity ofthe formation a Vpp—Ve—Vou (Sur ¥QQ0)- - Sur _ we 2) ‘rom which Str fe. , Y02 i on [cc =e Fo [ Sa 7CLAY FLAKE | ORY CLAY | CLAY IN WATER ® ° ° e ® ° ° atthe wget ® WAXMAN-SMITS MODEL Qy Cue "Cw +0 Se Fig. 1—Schemate of ly pal. In water-saturated formations, in which Sy: ‘expressions are simplified to Cur U=¥QQ Cu +¥QOLCoe + © and 1 Com Flt ¥QQICw+HOOsConle ooo ‘Water saturation in Eg. 5 is computed asa faction of oval porosity. Because Syr includes the clay water, sha- {ly sands may have high water saturation and still produce water free hydrocarbons. A better clue to the nature of, production may be given by Sy the fraction of porosity pe Filled by far water: % Spm tt op ® ‘pun comparable to effective porosity, is given by $p76/- "2006 so that Sur-v00r 5 7 1=¥@0r +++(10) Therefore, the values of vg and Cow must be known at formation temperature forthe interpretation model to be solved. The determination of these two parameters is iscussed Inter. ‘Theory of the DW Model very large surface compared to their volume (Fig. 12). ‘These clay-sheets are made by the stacking of two oF 154 three layers of aluminum octahedra or sca tetrahedra latices. There is often an excess of negative electrical charges inside the clayaheet* frequently caused by the ‘bstition of Mg** for some’ AI*** in the oc- {ahedral lace. Compensation for this loa electca unbalance is necessary to maintain the eletial netai- 'y of the cay particle. ‘The compensating agents are positive ions, of counterions, clinging tothe surface of the clay-shects in the dry sate (Fig. 1b). The resulting Postve surface charge, equal tothe intemal negative Gefct isa constant characteristic ofthe ely called fon exchange capacity (CEC). I is expressed in mi equivalent (eq) per 100 grams [3.5 02] of dry rmateial* "When the clay panicles are immersed inan electrolytic solution, the Coulomb forces Keeping the. postive ‘counterions at the elay surface are drastically reduced by the high dielectric permitivty of water. The counterions leave the clay surface and move relatively freely in the water (Fig. 16), contributing tothe electrical conduetivi- {y ofthe water and thereby increasing the rock conduc- Livy. This added conductiviy i proportional tothe on- centration of counterions in wate. I is therefre logical to express the surface charge in terms of counterion con enirtion inthe PV or Oy Its expresed in ey of| total porosity. Because of the mobility of clay counterions in water, they may be replaced charge to charge by any ober a= jon in the electolyee, provided the electcl balance is maintained. Refs. and 2 dal exclusively with rocks ‘where counterions and cations are Na*, as does this paper. Despite their mobility, clay counterions do. not distribute themselves uniformly inthe water. They re- tain auracted near the clay surface by the negative Charge they neutralize, Du they are kept at sore distance by water molecules adsorbed a the clay surface or it~ rounding the counterions. As a result they ae separated from the clay by a thin water layer that fons cannot peneite, “The existence ofthis jon-fee water around clay par ticles should not come as a surprise. The fact that the ‘conductivity orchloriity of wate in some shales is ess than that ia suroundingsands*® could be related tot although onrich shale waters have also been reported? More conclusively, it has been shown experimentally that fresh waters ar¢ squeezed out from saltwater cays at ‘ery high pressure. In addition, the "swelling clays" ae the most fequent marifesations ofthe expansion of the fon-fre layer under a change of wate alin ‘The Diffuse Layer and Outer Helmholtz Plane (OHP) If one accepts as a first approximation that fons and water molecules have negligible dimensions, the distribution of Na* and C1~ ions in the vicinity ofthe clay surface would be as predicted by the Gouy model, described in electrochemisty text books, and illustrated in Fig. 2. The solid and dashed lines represent the local concentrations of Na* and CI~ ions. respectively, at distance x from the clay surface. “The zone near the clay surface where Na* exceeds CI constitutes the “diffuse layer." Its thickness is ‘SOCIETY OF PETROLEUM ENGINEERS JOURNALFig. 2-Gouy's cits tayer model characterized by the parameter xy related theoretically 1 the salt concentation. <> in the bulk water at 23°C UrFiby en yen where 4 is im angstroms (A), _ in mol/ém? Imol/mL], and y isthe NaCl activity coefficient. ‘According to the Gouy model, the diffuse layer is equivalent (1) a layer of thickness, x, which contains the Na* cations necessary to balance the charge defi cleney of the lay bet none of the Na or CI ions of the salt dissolved in the water: and (2) the zone beyond £4, withthe salt concentration ofthe bulk water. “The definition of the clay-water zone would be com- plete if it were not for the frite dimensions of ions and Water molecules. Acully, the Na counterions of the diffuse layer are kept some distance away from the clay surface by the layer of water molecules adsorbed on the clay surface and a shell of hydration water molecules ‘surounding the sodium ion (Fig. 3). ‘At their closest approach, the Na* ions are located ‘with their centers lying on what is called the OP? at a distance xy from the surface (Fig. 3) From Fig. 3 and Ref. 9. the distance of the OHP is ay He Det Vir try 618 A, where re, the radius of water molecule, equals 1.4 A, and Pay the radius of the Na ion, is 0.96 A at room temperature. ‘The water salinity at which the Gouy layer thickness, xg, matches the OHP distance, «1, may be computed from Eq. II to be =035 molidm? [0.35 ‘mol/mL]. The corresponding value for 7. the activity coefficient. at this salinity is 0.71. Two situations are then possible. 1. The water salinity exceeds . Then all the counterions are located on the OHP. The zone of salt ex- clusion i limited by the OHP and its thickness, xy, is bout 6.2 A at room temperature, and is independent of bulk water salinity: xyery. APRIL 196 ‘ADSORBED SODIUM TON WATER HYDRATION WATER SCHEMATIC WATER OUTER MOLECULE HELMHOLTZ PLANE Fig, 5~Outerhelmbot plane ‘The salinity is less than . Then, there is tru ly adifuse layer. The thickness of the zone of salt exclu- ion, xg, is given by the Gouy diffuse layer model and is ‘now salinity-dependent. Tn general, we may write where a, derived from Eq, 11, is = 2) OF coure, a=1 when the water salinity exceeds Volume of Clay-Water To determine the volume occupied by sal-fiee water, ‘one needs to know the clay surface area in contact with the water. This surface ofthe clay is actualy related to the clay charg. Tchas been recognized for some time that the surface charge density of most clay minerals is relatively con- Stant.® This point has been investigated" with somewhat diverging results. For example, the average of 25 results from Refs. 13 through 16 leads to an area of {615 m?/meq (7.266 sq fuimeq) when specific area is greater than 100 m?/g {1,076 sq fug] and 580 m?/meq 16,243 sq fvmeq] when specific area is Tess than 100 ‘m¥/g (1,076 sq fUg). On the other hand, Patchet® finds for 109 shale samples an area of $80 m¥/meq (6,243 sq filmeg) when specific area is greater than 100 m?/g 11,076 5q fg and 800 m*/mea (8,611 sq Ven] when specific area i less than 100 m*/g {1,076 sq fg). Pat- ‘chet results do not apply too well here. They deal with shale and the measured area includes the contribution of the silt fraction, particularly significant when the specific area is small, ass‘The data of Diamond and Kinter® have been used in this sudy. They cover nine samples ofthe three main Iypes of clays (montmorillonite, ite, and kaolinite) plotted on Fig. 4. These clays are API standart andthe consistency of laboratory procedures permits « mean- ingfl analysis of results. These data were also used as reference of comparison by Patchet. “The experimental data of Fig. 4 plot on a single sinight line having a slope of one, indiatng direct pro- poronaliy beeen specific area (mi/g) and specific CEC (mee) independent ofthe type of ely. Ag =XCEO ‘where the constant » may be read from Fig. 4: 450 m?/meg [4.848 sq fUmegh. Normalizing to porosity, rather than density the clay surface per unit of BV is Aem1Q0 ‘The thickness of clay-water layer being xg, the volume of elay water, per unit bulk volume, is Vou Aves = A)Q28, = 008 QO1- Setting dfx 450 enPlmeg (6,844 £2 /mea) 618A =0.28 drm? /meq (0.28 mL/meq) . 13) and vara. 8) one derives Eq. I sated previously: Voy =a¥ 60.6, =¥90r0: 16 Conductivity of Clay-Water ‘The water near the clay contains only the clay counterions with pore space concentration Qy. Their concentration in clay-water is then See a9, Let B be the counterion equivalent conductivity. The clay water conductivity is then a6 which is independent ofthe amount and type of clay. In a NaC] environment where sodium CEC predominates, 8 and Coy are universal parameters that depend mainly on temperate and somewhat on salt ‘concentration, With nonsodium CEC, encountered in some shales,” they also depend on the pe of ‘counterion. ‘The value of is derived from core ‘measurements Ite. Effect of Temperature ‘Counterion Equivalent Conductivity B. Temperature ocreases the water viscosity, which affect, inthe same ‘manner, the mobility of any ion in the water. Iti also likely o lower the weak bond of counterions withthe clays. Accordingly, @ should increase as fast as or somewhat faster than the conductivity of a NaCl solution Clay-Water ‘Thickness. For salinity. greater than “<> theclay-watar thickness is.xy distance of OHP. Temperature decease the average residence time ofthe adsorbed water molecules onthe clay by decreasing the Strength of dipolar bond, thus decreasing the effective {Bickness ofthe adsorbed water layer. Consequently, ¥q should dcrease wth temperature. Assuming the average Tesdene time to be inversely related tothe fequency of Jumps of adsorbed water molecules, one may use the rate process theory? to predict the shape ofthe temperature eect 1_e7sGoler T a7 where Ts absolute temperature and AG is free energy change from adsorbed to free water molecule stat. For slinity les than << > the clay water has, the thickness ofthe diffuse layer, shown by its theory? t0 vary with T™, (being the absolute temperature). 18) vorevon[Z]" From 25°C to 200°C [71°F to 392°F], vg should in- crease by 26%. "This discussion has presented the physical foundation (of the petrophysical model. The existence of clay water SOCIETY OF PETROLEUM ENGINEERS JOURNAL,WATER-SATURATED FORMATION ‘conDuCTIVITY co. WATER conueriviry. Fig. 5p vs. C, schematic. ‘has been justified and its petrophysical parameters have been identified. The actual values remain t be deter- ‘mined experimentally. ‘The Experimental Approach ‘The experimental data come primarily from the pub- lished work of Waxman and Smits! and Waxman and ‘Thomas.? In addition, Shell Development has allowed sto use some unpublished data concerning the effect of temperature and saturation studied in Ref. 2. Quantification of vff, B, and Ce» Waxman and Smits’ experiments on water-saturated clayey sands established that the relation between Co (Ghaly sand conductiviy) and Cy (bulk water conduc tivity) has the characteristic shape shown in Fig. 5. For bulk water NaCI salinity in excess of 0.40 smot/dm? [0.40 mol/L} (25°C (77°F), corresponding 10 C,>C, the shaly sand conductivity Cy follows a staight line of slopes. This lin intercepts the Cy axis st a value (~C,), Thus, for value of greater than 0.40, +09) In this high-salinty range, a=1 and the DW model con- ductivity, Co, given by Eq. 7, can be writen in a form similar to Eq. 19: (v8) vO. Fo ( “Tyo, In this form Eqs. 19 and 20 may be compared term-by- tem, Co=s(Cy +00), Intercept. v0, Qy v0, 1-060, ‘The righthand term is obtained by application of Ea. 16. ‘The intercept magnitude, C,, is a quadratic function of G -@n) ‘APRIL 1984 t (oy INTERCEPT Oss ss Fig. Pret otG, v2.0, 4 with parameters vf! and 8. These parameter can be Acermined by euve fiting Eq. 21 t experimental core de For doings, the Cintercept for he 86 samples of Group 1 of Ret. 1 were recomputed from the base Cy conductivity dita. The ret te preseated on Fig. 6 “The vertical bers represent the uncertain cased by departures of the Co data from a straight line at high Sslnis, Curves | nd 2 are bes fs of Bq. 21 for the ory ofthe data and forthe mos southeast dats, repectivey, From cure | ¥6f=0.30 dm?/meq (0.30 mL/meq) and B=2.05 (Sim)/meg/em?) {G.05 (mho/my/(meq/em?)) From curve 2 vf =0.225 em¥/meg (0.225 mLlmeq} and B=2.50 (Siml(megiem?) [2.50 (eaho/m)/(meq/taL)) (Curve Fit 1; which is statistically the mos significant, gives value of v0.30 close tothe theoretical 0.28 from Eq. 13. Another independent estimate of vff is siven under Temperature Effet. It confirms tis choice. Cex is computed from Eq. 16, leading 10 Coy =6.8 Sim 6.8 hoi ‘Some comments should be made, at this point on the values found for 8 and vi for Na* is found to be 2.05 (Simy/(meg/em*) [2.05 (bom) mea/mL)]}. This is between the conductivity of Na“ of montmorionite in distilled water! 1.8 (S/m)lmeqlem?) [1.8 (tuho/n)/(aeqimL)] and the one of Na“ ions in saline Solution of same. ionic strength as clay-water 2.75 (im)mealem?) (2.75 (mbo/m) medi). 1sFORMATION FACTOR Porosity FORMATION FACTOR. PorosiTy Fg, 7—Formaton factor vs tl presi: (a) Waxman Smits) Dual Wales. CConceming vf soul be noted tat impos a tint be eal of tbe epee nse tons, Ascordngt 6 2, and with the ovis cond tion Ore! Qe. - <9 Thus, Q, shouldbe less than 3.3 in formations at 25°C (77°F) and 5.5 in formations st 200°C [392"F]. Mest , reportcd inthe literature respect hese limits although these truly apply only to fully wet clays. When part of the clay surface loses its water (desiccation, grain (0 iain contact, ete), the corresponding counterions become fixed on the dry surface (See Theory of Dual Water) and lose their electrical effect. However, these ions ae still counted inthe CEC measured chemically. ‘This leads to apparent Q, that are excessively high and 6 that seem abnormally lve ‘The exceptional values measured by Kem et al. on samples from low-porosity gas. sands!® (Q,=20, #80.) ilustate an extreme cate ofthis station. Sil, it may happen to a lesser degre on usual rock samples AA\way around this problem is 19 measure an effective Q,, comespoading ‘only to the mobile (conductive) counterions. Q, js then obtained from C, through Eq. 21 sng 6 and vg! at the C, temperature. Actually this clectial” Qy has been preferred to the chemical one most computations inthis sty. Additonal justifies. Hions for this practice ae given inthe last section ofthis paper. 18 Slope. From Eqs. 19 and 20 Lio, see cn F Fo Nea othe inves form Fo=Fut-v oo. 2) In Eq. 22, (1—v4/Q,) comets for the presence ofthe Clayswater and convents the measured formation for, F*, to the idealized one, Fo. ‘Accortingly, Fy shovid Beis isto by ely con- tent, and therfore more closely related Archie's raion Foon = @3) ‘This deduction can be verified on the 136 samples of Refs. Land 3 as shown on Figs. 7 and 8. Figs. Ta and 7b are plots of the formation factor against, for WS and DW models, respectively. It is seen tha the large seater ‘of the WS piot (Fig. 7a) has been reduced by a factor of {two by the clay-water correction (Fig, 70).* Figs. 8a and Bb plot the values of Archie's exponent, m, computed from Eq. 23, against the shaliness factor ¥=Q,6,I(1—6,) or clay counterions per unit volume of “Rovere we onaned oy corecing wey ae bye Ca? SOCIETY OF PETROLEUM ENGINEERS JOURNAL05 y Fig, Archie's m exponent vs, shalynes: (a) Waxman-Siis, (b) Ous! Wate. solid, The m of WS (Fig. 8a) appears more clay- . Clay-water-layer expansion at lower salinities isplaces some of the far water without ionic compensation. Data from Group 2 of Ref. 1 and Fig. 9, showing a typical sample from this group, show that a curvature ex- ints below =0.8 mol/dm? [0.6 mol/L]. This is close enough t0 the theoretical value = 0.35 rmol/dm? [0.35 mol/L] to assume that this curvature ‘could be the predicted one and thatthe actual value of ‘ should be 440 mol/ém? (0.40 mol}. Quantitatvely, the formation conductivities Coe com- puted from Eqs. 7, 12, and 14 were compared to the ‘measured conductivities Coy forthe 15 mos significant samples of group 2, Ref. 1 Gee Appendix). The average (of ratios Com/Coe are shown in Table 1. I is seen that cay-water expansion explains quan- titatively the Co curvature to any Cy down to 1 Sim {1 ‘mho/m}. This i illustrated on Fig. 9. This limit covers ‘most cases of interpretation. The additional dp of Com APRIL 1986 é me ey hos Fig. 8G curvature at igh 0: Sample 26 below 1 S/m [1 mholm) must be attributed to other causes, probably 10 the decrease of counterion mobility in dilute solution as suggested by Waxman.” Then, ‘Table I would lead to approximately Bag =B(t~0.407*n), ‘Temperature Effect “This study is based on the conductivity the nine shaly sands described in Table 4 of Ref. 2. The procedure used to transform it into intercept C, and Slope I/P* at each temperature is described in the Ap- pendix. F* and C, values are listed in Tables A-I and & measurements on 19‘TABLE CURVATURE CORRECTION FACTOR, C,, 20; eh COUNTERION EQ. CONDUCTIVITY {mho/m), (meq/ce) 10 ca ‘TEMPERATURE —> oFhrCUr SLL © 100 200 300 400°F Fig. 10 vs. temperate. 0.35) x Ne 0.25] tw sy ec/meq - Temp °C —m ous 100° Vii vs Temperature 200° Fig. 11g ve. trperatue, ‘Clay-Watet Volume Factor. It was predicted from the theory that vf should decrease with temperature; conse ‘quently F* should decrease with temperature since TABLE 2—Vf vs. TEMPERATURE (*0) 2m 0 200 Vin B50 Tas Das B09 vatT) 099 0254 0219 0.103, ‘The effect of temperature on Ft can be verified qualitatively in Table A-l or on Fig. 10, where the changes in F* between two abirary temperatures are plotted against the shaliness Q, ofthe samples. Despite some lange dispersion from sample to sample, Fig. 10 shows clearly that Fi essentially temperture- independent in nonshaly formation (Q, ~0) and thatthe ‘change die to temperature increases with shaliess; hs is in agreement with Eq, 22. Similar results were also repored by Ker," in which a diferent approach was ‘sed to quantify the variations in F*. The variations of [7fQy] with temperature were com- puted from Table A-l using Eq. 22. under the assump- fions that Fo is invariant and that ¥f(22) is 0.30. ‘The average values of vf e,, computed from Ee-Q7EQ., ar listed in Table 2 and displayed in Fig. I. 295425 Tet25" vffer=0.30% es) Where Tk represents absolute (Kelvin) temperature, CE its g wit ny 2% er. Nei te close dependence of Vf onthe reciprocal of Ts ex pected from the rate process theory. Counterion Conductivity. The increase of intercept Cy (Table A-2) reflects the increase of @ with temperature BQ, om ; 90. A yalue of (60.) was computed for each C,, using the 17 values determined in the previous subsection. This was the base for a weighted average By computed at each temperature. Resulis are normalized in Table 3 and Fig. 12. -026) ‘where Tec is temperature in °C. Itcan be compared with the conductivity C4, of Na fon in salt solution of same onic strength as clay-water listed on the final lin ofthe table. Clearly Na fons from clays or saline solutions ex- hibit similar dependence on temperature above 80°C [176°F} (water viscosity), with the counterion condue- tivity growing slighty’ faster at lower temperatures SOCIETY OF PETROLEUM ENGINEERS JOURNAL‘TABLE3—f.vs. TEMPERATURE °C) 22 0 20 be Bos Ses wa iar an 20s $88 100 40 ch an 657 105 135 “These results are in fair agreement with those reported by Kem et al.,! who found the counterion conductivity t0 follow the same variation as NaCl solution Clay-Water Conductivity, Cov. Entering Eqs. 25 and 26 into Eq. 16 leads to Coe #7104 Pac +8.5)T ag +298) en ‘As is apparent in Fig. 13, Coy increases faster than x, with temperature. ‘The decrease of the clay-water volume increases the concentration in counterions. New Estimate of vff at 22°C [72°F]. Previous ‘estimates of vf] were derived from either chemistry prin- ciples or C, intercept data. A new estimate will be deriv- fed now from an independent source: the F* values (Table AD), ‘According to Eqs. 22 and 26, G)-G)* Tet 29295425) .22)(295+20, en Fo with appropsiate substations Zi=BitAayXs ir Te 425), Boies Ailb,=t$ 223320103 29) ‘A, and B; determined from regression analysis, together With the raw data, are listed in Table A-I for 11 core A eximate of vf (22) could be made foreach coe from 4/8) 32005" ed v2) ‘To obtain a meaningful average value for vf{(22) we need to do better than simply take the average value of. 2), since such an average will reflect the con- ble uncertainty in computing (4,/B,) for samples APRIL 1964 ond 185°C snz 15 Change in F Méue to temperature Nereus Oy. Fig. 12-Chango in F* duo to terperature v8. Qy- ott NaCl SOLUTION so_gS EMEP RES] Gas abo aa Fig, 13—C,_ vs. temperature. with low valves of Qy; instead we use the form Eay/B) __ 916 32000, 320x108 ¥§@2)=——— =0.27 meg/ém? {meq/mL}. en ‘This new and independent estimate compares well with the previous ones of 0.28 (chemisty) and 0.30 erp). Comparison With Core Data. The conductivity at 200°C [382°F] ofthe samples studied by WT have been ‘computed according to the dual-water temperature cor- ‘eetion model. The results are presented in Table 4 as‘TABLE 4—-TEMPERATURE CORRECTION TEST ore (Cy) (22a 4.74 Moat Say (Gq Waxman Dual Water Experimental zmc Us an ame gra 0S0 4s} 520 gros 099485829 8.00, moon Sas 400 Gas Smet 20 S16 Ba 880 Shae? 21 Sip Tae B30 ratios of conductivity at 200°C [392°F] to conductivity ‘a 22°C [72°F]. These ratios may be compared to the ex- Derimental ones. The values computed acecording to ‘WT model are also presented for reference. Comparison of DW and WS Models The nal data released to us include the measurements of the conductivities ofthe core samples tused by WT jn their saturation experiments.? These data, given in Tables A-3 and A-S, offer the opportunity to make a comparison between the DW and WS models Variance of Models From Cy. (Water-Saturated) ‘Measurements. The test covers 52 Cp measurements distributed over 12 rock samples. The model parameters hhave been determined by best fitting to experimental conductivities on the basis of percentage accuracy. The basic conditions have been maintained thatthe sum of the percentage erors for any fit sil and the variance is tminimum. For the DW model, Fo and Q, were opti- ‘mized simultaneously fora best ft with conductivities. For the WS model, a fist fit WS (1) was obtained op- timizing F* but Keeping the Q, determined by WT from the Cp conductivity data (Ref. 2, Table 2). Because of the poor quality of results, another fit, WS (2) was aso made, optimizing both F* and Q, simultaneously as was done forthe DW model. Best estimates of Q,, Fo, and F are given in Table ‘Acf for the 12 rock samples. They define the rock model ‘parameters used in the saturation study discussed later. ‘The variance between the corresponding models and measured conductivities has been used to quantify the quality of fit with experimental Co. The smaller the ‘eriance, the better the ft. Variance computations, ex- pressed in percentages, are summarized in Table 5. ‘WS (1) exhibits the greatest variance. Apparently the x determined by WT does not fit the Co data well. The ‘DW model gives the best fit; better than WS 2. Archie's Formation Factor and Saturation Ex ‘ponents. Ii interesting to compare the values of m and ‘obtained forthe 12 rock samples according tothe three models, m and n are computed from ‘The derivation of ns more complex because each rock sample is characterized by many computed m values, tach one coming from an experiment where either the saturation of C,, has been changed. The average fi ob- tained for each Sample, and the variance from 7, othe coresponding standard deviation, (n) will characterize the rock sample saturation exponent. ‘mf, and standard deviation from fare listed forthe thee models of the 12 rock samples in Table A-6. This ‘sample-by-sample comparison of m and Zi may be of in- terest tothe log analyst. It is true, as suggested by Waxman, thatthe stbilty ‘of min relation to 7s a good criterion of the quality of the model, then the DW model seems better, as seen from Table 6. Is variance is one-tenth ofthe WS (1) and one-third of the WS Q). ‘The average of m and Fi over the 12 rock samples are given in Table 7. One may try and express m a8 a func tion of m. An expression of this type is given under Ref (m) in Table 7. The standard deviation of m and Fim) ace rather large particularly for Modet WS (1). “There is apparently no significant advantage i using af (n) instead of (Fy ‘TABLE 5—VARIANCES FROM EXPERIMENTAL Ce Number of Measurements save 6 8 ‘ora Varince Deviation) 19 ws) ws@ DWaodet _Model__ Modal ae RATT 12 boat 4300 Ey 14 038 204 338 257 20 13 (1088 208 778 20 {00 ma Er 00 24 00 oa 151188 19 4a 28 SOCIETY OF PETROLEUM ENGINEERS JOURNAL‘TABLE7—AVERAGE OF m AND Tonia mock SAMPLES DualWater _WS() _ WS) DualWater W511) _ WS —— = he va 198 (teense) a nos m8 oy? se “ee TN ‘Standard deviation (m) ont 014 nda devon ‘Stancard ovation ex a0 ‘roma ae 8 a7 Standard Sevan ft ox O20 ‘TABLES-0, COMPARATIVESTUDY 2, (Ekceal (0) Eletieat recent WB) Model __ WS.) Modet_ sec Osi aco ame ose ost an oa oer “anc oa ose sie oz 2s aes oz 023 See oa4 rd see eae os aaa oa on mae oie oa oe 208 Sou B04 om oh ‘verse change ‘romenomiat ares) ry +20 ve ‘The Q, Evaluation Problem. Table 8 compares the lected! Qy of the thre modes to the chemical Q,- 1, of WS (1), plished in Re. 2, is remarkably Similar'o chemical Oy; however, this sath expense af por fit Cp measurements and lage dispersion of exponent. Q, of WS (2) and DW agree very wel in vew te lage diferenes inthe basi parameters of two tmodes. They both read significantly in excess of chemical Q, (425%). A similar trend was noticed with the sample set of the temperure sly. These difeences cannot be ex Plaine by clyy dehydration, but they flostate the Timitaons of th comelation been chemical and lee. teal Qy- They sugges that Qy infeed from eleceal Ineasurements, whether condctvity, clectochemical potential, °2" o dielectric constant'® are to be Imor appropriate othe interpretation of resistivity logs for satuation. Conetusi ‘The conductivity model of shaly formations proposed by ‘Waxman and Smits has been modified to take into 3c~ count the exclusion of salt around clay panitles. Tt was shown that, conductivity-wise, a shaly formation behaves as iit were clean, but contains a water that i a mixture of formation and so-called “‘lay"" waters. The conductivity of the clay-water is quasiuniversal and ‘depends mainly on temperature. The WS data were reinterpreted in tis light to obtain the parameters ofthis APRIL 1984 model, Improved coherence and fit with experimental data were obtained with this DW inodel. Perhaps more significanly, the DW model has brought unification and simplification to log interpreta- tion by putting the evaluation of shaly formations into the usual framework where resistivity is analyzed in terms of total porosity and saturation. ‘Nomenclature Ay = WG @22)-3200,18, Aug = specific aea of clay surface, m"/g (sa flor} ‘A, = clay surface area per unit of PY, mem? [sq fvem?} £B = equivalent conductivity of compen ‘ating counterions, WS model (Sim(meqiem?) (eaho/myeeq/mL By Fo ‘Bo = constant value of Bat high water salinities, By=Cx/Qy, (Simy/(megiem*) Umho/my(meq/mt)] Cow * conductivity of clay-water, Sim Imboem) C, = tne conductivity of (hydrocarbon ‘bearing formation, S/m {mho/m] 16Cr = tol conductivity, Sim (mho/) Cz = (lormation) wate conductivity, Sn Thofm] Ce = effective conductivity of water in a shaly sand, Sm enho/m) Cz = intercept of linear portion of Co vs Cy cuve extrapolated 10 Cy axis (C.=BoQ,) Sim {mhosm) Co = conditivity of 100% water-saturated formation, Sim [ho/m) oe = computed formation conductivities, Sim (hon) Com = measured formation conductivities, ‘Sh [hon Cy = the C, value above which Cy vaies linearly wth C., Sim {mho/m) CEC = cation-exchange capacity (ry roel), smeq/100 g [mea/3.5 02} F © resistivity formation factor (for 8 clean sand), F=Ro/Ry Fy = resistivity formation factor fora shaly sand as used in DW shaly-sand ‘model = resistivity formation facto fora shaly sand as used in WS shaly-sand model, F*=Ro/Rne free energy change exponent in Archie F—6 relationship or lean sands 1x = sathration exponent salt concentration in water, mola? {ol} salt concentration above which =, mmol? [moV/L] , = concentration of clay counterions per ‘unit PV, mea/em? {oeq/mL} value of Q, computed from condue- tivity measurements, meg/cm? {eq/mL} radius of Na* ion, A = radius of water molecule, A = resistivity of formation 100% saturated with water (Ro=1/Co), om tev resistivity of (hydroearbon- ‘bearng) formation, +m (Cormation) water resistivity, > slope of Co vs. Cy curve for Cy, seater than Cy Spe = faction of “far-water” pore space that is water filled ‘water saturation in volume faction of total porosity = water saturation to nth power Tec = temperature in °C Te, = temperature in K Tp = reference temperature, K fr = volume Fat of clay water Vp = volume fection of fr water Vg = wry =A,tH/0,, cm? /meq [mLmeq] vf} = volume of clay-watr per counterion 22°C [72°F] when a=, em? meg (L/mea] er = E(VJQ,YEQ,, em? meg (Lime) (= same a temperature Ty v8 (22) = (A1B,¥320 Qu, (Simmegfom?) Uenhofm/(meqyL)] xg = thickness of diffuse layer, 24 = doce of OP from ay soc, Tg 425 Y= amount of clay normal ‘of solids volume, Qvér(t-$7) factor for difise layer {6 = equivalent condociviy of sium ‘ounteone, DW model at 22°C {72°F}, (Sm)megfem?) {mnkofm/(meq/mt3) Boy = BBQ,/EQ,, (Sim(nealem?) (Gnho/n) meg/mL)} ay = valve 0f8 for low-salt soluions, (Sin)imeaiem?) (bom mega} Brg ~ values of 8 a temperatures 7, CA +7 = NaCl actviy coeticient specifi ly area coefficient, 1m? /meq (sq fi/meq] 4, = ttl porsity, fection Acknowledgments ‘We thank Shell Development Co. for making available additional experimental data concerning the dependence Of conductivity on saturation and temperature. We acknowledge the contribution of Peter Day in optimizing the Fv. 6 relationship. We thank Emie Finklea for his patient help in the preperation of this paper for publication References 1. Waaman, MH. an Sis, LJ.M- “Betiat Condi in (i Beatg Shly Sais” Se, Pet Eng J. ane 1968 107-2; Tran. AIME. 28, 2, Wermaa, MH nd Thomas, Clea Condcvies s ‘Shuly Sune“, Toe Relation Between Hysnenten Satin nd esi Ine, The Penge Costin of Elec ‘Bn Teh Fe. 192) 213-23; Ton, 3. Contr, G.. Bouteny, Yad Cleve, “A Say of Dual Water Mode! Based ov Log Du." J. Pt Tech. Gan. 1983) 15856. 4, Yan Olen HA farducsion o Cay Call Chemistry, Stn ‘Wiley & Soe ine, New Yor Cy (1963 30-44; 251-7. 5, Rich, Hand Peseta, = Relonship Beez the Con ‘Soci of Tertary Water Bearing Sams ad Naty Sales Fer presented we 1978 SPWLA Symposium, I Pao, one Bie SOCIETY OF PETROLEUM ENGINEERS JOURNAL6, Heer. WAL: “Pov Water Chains of Sobre Stale, PRD daseaton, U. of Wiseonln Maison (1960. 1, Weaver, CE. and Beck, "Cy Wa Duce ‘Bun: How Ma Becomes Gwe.” Special Paper 134 GSA, Bower (97D Tot 4, som Expelad, W.V. an aid, KL: “Concemation Changes {for Solon During Cnpain of Cay Sediment". Se. ‘evotory Ox. 196) 33, Ne 4919-30, 9, Bockre, and Rasy AK Modem Ape of Elec robes, Penn Pes, New Yor iy 190) Cap 7, Ss qare. (0, Paste 1G. “An lvsigation of Shale Conduct.” paper ‘V pesca te 1975 SPWLA Aud Logging Spot, ew ova, ine 47 11, Dimond, 5 and Kier, EB. “Soave Aras of Clay Mies 2 Dedvel rom Menurneatsof Goel Retention" lye nd Chey ‘Minerals, A. Swielod ) Na. Acalemy of ‘SSoeaNat Resch Cour, Watingon, D.C. (959) Ben, 12. Oveteck, 1.7.0. “The Interaction Between Coie Par tc” Cad Since, Elsevier Pub. Co. New York Cy (G93) 25-7 13, Fog 2! "Os te Quetion of Constancy of Cay Mine Caton Exchange Capacy per Unt Are,” Clay end Clay Minera, Proc Le Nl Cay Confrence, Perunon Press, New York Cy (1968, 16, Gree Kell, Re "Cham Denies a Hat of nivenion of ‘Sone Chy Minas" Clay Minerals Bull 5, Mineropet Sorey, Lando (1982) 18. 1, Nomi, Kad Rath Clam, L.A: “Low Ange X-Ray Dit facton Sides ofthe Sweling of Montmatloie and Ver ticle," Clay and Cly Mier, Pro 10 Nal. Clay Cam. ene: Machin Pres, New Yor Gy (96) 12308. 16, Gray. WoC. Saas, Mand Wooe, KH "Exchange ‘etn of Kali of Varying Deg of Cosy.” Am. Gonait Soe (1962) 48, 3-66 17, Yn Often, Hand Wasean, MAL: “Sue Conder of ‘Soum Beni in Wats Clay and Clay Minerals, Proc. 5 Nat. Conewmce Chaps and Clay Minera, Nat Acaemy of 1, Kem J. Wr day WA. and pat, MEM!" Port Gas Sid Anges Using Caton Exchange and Dice Cont Das,” piper preset athe T97G SPWLA Ano! Logging Synod, Dem, use 9-2 19, Kem JM, Hoyer, WA and Spann, MLM: “High ‘Tenpersne Elecial Coatctiy of Siiy Son” paper Presta he 1977 SPWLA.Anmal Lepsing Sympostn, Foes, ce 58 20, Sm USM: "SP Lag Ieee in tly Ste, See. Pet ag. Gane 196) 125-36: Tran, AME, 2. 21, Tas, EC? Tee Deterrnaticn 6 G, From Membrane Panta Meare's on Sal Suns.” So. Pet Bag J. Qe 1916 130-36 rn, AIME, 26, APPENDIX Correlation of F vs. @, and m vs. Shaliness ‘The origin of the sample data is as follows. ‘Table 1 of Ref. 1, all samples. 31 ‘Table 1 of Ref. 3, all samples 12 Table 4 of Ref. 3, all samples 9 ‘Fable 5 of Ref. 3, all samples 2 Table 7 of Ref. 1, all samples 27 Table 3 of Ref. 1, all samples_55 136 Curvature of Cp Line ‘This study is based on the samples of Table 7, Ref. 1, ‘which satisfy the erteion 0.50,<(Qn)a<20, Namely, these are 4, 5, 11, 12, 16, 18, 19, 20,21, 22, 23, 24, 25, 26, 27 ‘Temperature Study ‘The raw data forthe nine samples of Table 4 of Ref. 2 consisted of 35 Co measurements per sample, at seven temperatures and’ five salinities. These data were smoothed by fitting a monotonic curve over the seven data points fora given salinity and the on ‘The next step was to define the (Co. Cw) straight ine from these data by linear regression onthe three points of highest conductivity a each temperature. Each line has ‘en characterized by is intercept C, and F* (slope), ‘which constitute the basic material fortis study (Tables Acland A-2). "The two shales of Table A-2 are described in Teble 5 of Ref. 2. Their F" have been computed from their cor- uctivity at 22°C [72°F] and the ratio of conductivity vs. temperature of Fig. 8, Ref. 2. G)-« Gees)” Comparison with WS Model ‘Table A-3 Experimental values of Roy Table A-4 Bestfit models of Co=1/Ro, ‘Table A-5 Experimental resistivity saturation, ‘Table A~6 Formation factors and saturation exponents. Tables A~3 and A-5 are experimental data provided by Shell Development Co. index and TABLE A‘—F* VS. TEMPERATURE. Temperature (26) oo 10 880 850 — 058 gesG 133° 192 232 Boor haa 2h Gioe 373 88 Ss hk for 258 sae? gc 5 3 fo See gare 75 SB wee 21 ms ms m2 boa Ta 638 G1 610 Sra 17. 159) 42129 Seale? ‘423 ‘372 ‘se ‘tas ‘APRIL 938 Linear Fit 1 A, 958 855 869 0288 O17 wan 131 31 Zt0e" over BS 246 248 O82 O.05B6 Ba Ss S37 -271' Ones Bi ko 239 “Bar Oona Sa Sts S39 “hve oan Bo Se Ses 245 Oss 2a a8 28 a3 Oost fs S89 S84 “aan 0006 119, 100, $37 “a07 oo1et are ‘ara 237-183" Ortsec Bre saa 296 sae sane Soe size arr 2208 Sos Suoa () cteticat Mo 7 20 pasar ~ 0x8 178 208 225, does O48 tse 227 356 wor Os 268 Sat 416 Ghee 038 S30 452 Sor poor Oe ear ss 110 gsc 10s S10 550 Bao re tas sea ar gra 13) 415 Tse 120 62 202 230 288 gro iz 4 Tes 7 feo 12 29 38 peo 132 499 853 123 SB tao 27 ‘TABLE A.9~-EXPERIMENTAL VALUES OF Ri (@-m) 0.0495 o.0808 0.098 0.0987 0.125 0.165 Os27 0985 1.06 300 «4m oar og oR mn tt Sie eam rd ee I'S = ih 75 2 — am ame 4m Ge sep 32 = fos 4% she ts dee 2g SDS SD hoe moo SS = ar os Sor = 455 ss) os 120) aa Ss’ OL Yee CS 25s “tes 27 = = one et ey am en = ft se 5 OS gh H com SB 5 5 rt OS ao ‘TABLE A2-INTERCEPTS C, AT Tuc TABLE A-t—REST FIT MODELS OF Cy us Water wis-1 ws-2 No.of Gy wor Varate or Vaiable morass Moasuroments 0, ue CO ee ON a ON Oy cet ON Oe @ o8 Hoy =01 “Sag Oa wa as “Baa =34 oy < O49 fer “oe | “2 oss tess Ok dons 02 5. ¢O8¢ 2560 OY) 43 Oat moss 03 Sas oe & 38 3901-02 ts4 O28 aoe 0 a8 02 te8 8 03 SB os) ms Oss ago ya oe S 028 2800 O27 0m S199 O88 os 80 3 038 eae 03 NB O25 788 Oe 1080 035s 8 oy ae 03 ms O82 440-02 (78 02 332 5 G10 aba =o «20011 tes? 08 100 08 100 5 Gao ‘927 01 «45013 1000-02245 “ors. AraC te ay ter te ty 2 ons tar 08 Ok a7 0973 on 0 ‘SOCIETY OF PETROLEUM ENGINEERS JOURNALaac une = 9c eh sasca 0870 3220 5M s12¢ Oss 27k 2298 sore 31908 Heivvnnn ceerenerianng APRIL 1988 BERBERS 1! a RUE ER DEER OOOO OREO = +179 reitiie "TABLE A-5—EXPERIMENTAL RESISTIVITY INDEX ANO SATURATION 5 8460 ose oes BQcecciecereeieeceni enc pS VOVUIBRRBBUEEEH TTT ireeri erie rsa gge: 8 g oats oe USgR eccrine Ae EEUU OCC PTEECE ETUDE EETE CHEE EEE eno gBE fe URTERER UST OR EROS OR ORATOR ESTER R TORENT OT NTO 78 ons st oar 368 Sass 200 omar aaa ona 178 oan 134 oso 2: ees 189 Ona? 15 Oats hee oar a0 D510 204 Oss 235 Sass 180 Oana et oss 272 basa 158 0928 115 oa a7 base 325, Osa 263 set 207 ons 150 ay ost ruven BBBSE re BB vei ' Bugis reins ay ons 0213 are VO coer ble e8 eee Rae BF VCO PCEECHEeEE rennet iieettreceecereeeericrat fNo.of 5, Doviation o ‘TABLE A.6-FORMATION FACTORS AND SATURATION EXPONENTS Dual Water ‘Sandard $a58ee |? BeBEaE! SI Metric Conversion Factors °F CR-3.8 = ec fx 308" EOL =m mL x 10° E400 = cm? mho x 10+ E400 = S oz x 284952 EOL = 8 “coments SPEJ Sebel tenn Sn oe vad cern eae et {Sx Perec Siero ne orev coe SOCIETY OF PETROLEUM ENGINEERS JOURNALSPE 19577 SPE Socket of Petroleum Engraors Salinity and Saturation Effects on Shaly Sandstone Conductivity M. Argaud, Eit Aquitaing; H. Giouse, Gaz de France; and C. Staley, J. Tomanic, ‘Schlumberger-Doll Research Shoe ISP ce pvr pee Ptasos apr SPP. Bx 6 ‘ABSTRACT Aithough a considerable amount of effort has been ex= eaded onthe subject, the interpretation of shaly sandstone re- Eisivites often proves problematic, Since the amount of high ‘quality experimental data is somewhat limited, a joint research ‘project was cooducted 10 acquire more laboraiory daa tnd t9 ‘Evaluate the data in terms of new and existing theeredeal models, ‘This project consisted of two sets of experiments: + conductivity measurements on fully brine saturated rock samples invesgating the effet of changes in salinity andor ‘ype ofeatons + resistivity index measurements at decreasing water satura- ‘on and varios. schieved by capillary drainage. ‘ee sep st of xpinen willbe presen is ue. consist of mere % ni andy as eatued by chemical itadon, mewbrane postal snd Cos Cy inercept Oa etsy ne at cy ty sno elo Va clsEem ea ee ere een ae oe en a eng re Beer aerate eae ici st page lap init eee eae aes Eee mmRopucTON ‘The imerpreation of shaly sandstone resistivities has Jong bee problem in fog analysis Asa resol, a large numberof resistivity salinity and reistvty-sarraion {onmlas have beza developed by log analysts good review {s given by Worhington. Tnterpreation becomes paculaly troublesome whenever clay conductivity Becomes a significant References and Wastatons at end of paper. ergot jot ck conduct, ‘Ths can oct ines ving large clay content or in rocks having low tine conduc. ivi Ste vr few batry data ek ean fo bese tons, we present inthis paper new ‘mess ‘nts of resintviy index vs water saturation Gy 2 Sw)s We ‘sted coply qulriam wchnlge sma tat of Wesman and Thomas, exzep forthe adion of a4 terminal ressivty ‘eazurement) We hued rocks having various cay coments and Fan experiments using ferent bene sala [Rather that anempr to evaluate the 30+ models described ‘by Worthington! we will instead focus upon the three most widely used conductivity models (Archic}, Waxman Smits, Dual-Water’) and two new models recently proposed by Schwartz and Sen® and by Giouse and Argaud (called the DC model). The WS and DW models are especialy concered wit the elecrocherical effects of clays, while the S$ and DC models Hholetede geomesied elec oe ceca conueson paths ‘The well-known Archie expression generally applies only w clay. sands ands given by: a Gash ey were Cui the rok conducive the wate sao, @ 1s porosity and Gy ise ine coach. mand etry sored te ntanon exponen ind contin Ejeet ropetne mb ly died by Ace oe expeton a FaSeeg™ one @ her Fish formato factor an Cle the rock contucviy visa! ‘The Wexman-Smits equation includes a corection for ‘lay conductivity ands given by:‘SALINITY AND SATURATION EFFECTS ON SHALY SANDSTONE CONDUCTIVITY SPE19577 seeheE) where Qy i he cain exchange capacity (CEC) per unit eles ut Ss opel netics ot Seay omar: Beasli-o6e} i ‘he Du. Water adel as or complied cay or recon erm given steer cere 8} where Bis the conductivity of the clay bound water and avg Beididees fe bound eer yer The Schwar-Sen mde sunesdifen omosteg for the surface conduction path andthe pore condaction path and arrives at geste [oy + 1228 Joe Free “The constants in the $S model were derived by fitting a more ener form of hs equaton vid Sen woe Wasa Sats ‘Bee ua ws re nett an sre MEDS a rt ey iouehverparieerm eres Shabiecratetnnanere es Seana eee hese Say. aie ome ng echoes ee ENG nie yranwyiume name forSyi| |-=ss — oc + ep. points a 1.0 Fig3 ~ Resistivity Index curve L3+1 * Cw= 2.824 S/m Ir a 1.0 Fig # — Resistivity Index curve L3*2 * Cw= 3.304 S/m— pe + exp. points Fig + ~ Resist ty index curve Mist s/n ws ss —oe + exp. points Fig é ~ Resistivity index curve M1+2 Cw=11.986 S/mistivity Index curve CONS#1 Cw= 3.230 S/mFig ? — Resistivity Index curve AUS12+2 ‘Cw= 0.354 S/m Index curve AUS12#1 230 S/mw/s z9s'y =o HOLM enuno xePUI Ayansisoy — zy Bly w/s 269° =KO — LL enino xepul AyAnsisoy — 77 BLY * pi 1 i He fe t MS im syujod -dxo + syujod -dxe + vie lL 89Z"b=U aIHOUY ——— | | OS6"}=¥ SIHOUY $68"1=U siHowy — r re # 8 For ey poz Loz Loe 7