..J cPTb{,-03-- Ob Quantitative log Evaluation of the Prairie Evaporite Formation In Saskatchewan
By E. R. CRAIN and TV. B. ANDERSON'
(17th A.nnual Tecllnical ~lIeeting, The PetTole1.t'm Society of C.l.lll" Ed'manton, llfaYJ 1966) Downloaded from http://onepetro.org/JCPT/article-pdf/5/03/145/2165444/petsoc-66-03-06.pdf by guest on 27 September 2021
ABSTRACT 'I'he problem of solving for the fraction of s:rlvite7 car~ nallite, halite and insoluble material in the Prairie Eva porite formation can be performed. by a suitable inter pretation program based on Gamma Ray, Sonic. Neutron and Caliper logs. Empirical relations were established between the log '"aloes and the formation parameters, the result being- a set of four simultaneous equations which may be reduced to obtain the desired fractions_ Tedious hand calculation can be eliminated by using computer techniques and automatic log digitizing machines. Correla~ tion between core and log analysis is good, and the speed and efficiency of the method is valuable in initial forma ,_:.: .. tion studies.
INTRODUCTION HE Prairie Evaporite formation has been the object of extensive study in the past several years , T(1) (2). It is the richest known potash-bearing bed Figw"e 1. .-~' I in the world, and, as such, it is important that any .. information gathered concerning- the zone be accurate POTASH CONTENT and immediately useful. NOMOGR,aPH Electrical and radioactiVity well logs have proved ,~, to be of value for formation evaluation in the oil in dustry. A recent paper (3) illustrates their use for
both qualitative and quantitative interpretation in .. ~ evaporite sequences in various parts of North Amer ica_ .i I This paper will outline the theory and technique .~ .' '~ used for a quantitative interpretation procedure in -J I the potash beds of the Prairie Evaporite formation in '-.: . <:,"-,- the Province of Saskatchewan. The data are set. up , "",..,-0<1[<_ -.: '- ...... "'< u.>'T~I"'l :: ..--- so that the}r can be handled by an electronic computer. ~,TIt . llRlUlO:rrUT_' .....r•• ,,'to _'0.u_n"". 11 The equations can also be computed by hand at the .lO'l'II'an ...... 'JT ~.' f well site to supplement the data already available_ The computer program is presented as an Appen- I· dix. An extensive bibliography! covering potash geol ., oglr, development and logging techniques, is included, Figure 2_ ,. THEORY Figltre 2 illustrates a nomogram which facilitates It is well known that potassium has a radioactive conversion of gamma ray activity to apparent K::!O k;otope 'which emits gamma ray energy. This isotope content. It is derived from the graph of Figure 1 (K4U) comprises a constant fraction of the total amount and therefore gives the same results. The result is labelled "apparent" K~O content because the insoluble of potassium, 50 that a Gamma Ra,}r Log, which meas r ures the amount of natural radioactivity in a forma content of the formation generally is slighth radio tion, frequentllc gives a measure of the potassium active and the chart tllus gives an incorrect K::!O value if insolubles are present. A correction can be applied, I content. which will be dealt \\·ith later, but it is small in many Considerable \II,.-ork \vas done in 1964 to establish cases and will not greatly affect the total ~O value of an empirical correlation between gamma ray activity a zone. These results are valid only for the logging and the K,O content of a potash bed (3). The graph tools listed on the chad and bor beds thicker than shown in Figure 1 illustrates the results obtained in 2 feet. oil-base muds. As borehole conditions affect the re sponse of gamma ray logging instruments, hole size and mud weight must be taken into account. :;·Schlum,be1'um' of Canada, CalgaT1.I, Alta.
Technolo9V, July-September, 1966, Montreal For thin beds (from 1j~ foot to 3 feel), a con-eetioH can be made using the empirical chart shown in Figure :]. The gamma ray reading in API units i~ rnultiplietl by the conection foetor derived from Figure 3 to ar l-iYe at the corrected valuE'_ Fnr beds less than 1 foot in thickness, the correction be('omes quite large and o ~ is not accurate. Bed boundaties are chosen at the in flection points of the gamma ray curve and the b~d , thit:kness is that distance between any two successive inflection points. -:++-H~4.7V: o As onl.y thfJ,Se zone:.; which are low in carnallite content aloe commercially athactive at pres~nt, a means of delineating these beds must be employed. The N eu i~:~~~~~:~-!:: ~'~B ! '::I,l~ ?:il~~tl-,,~: f1~f:~f~;~:: tron Log is an excellent carnallile logging tool, because " it responds to the hydrogen content of the formation. ,-,- '+Ji :HII:n',i,'~L."'~;fi::,abLfr"',"" "~::,~'-:',:.'~',,, The water of hydration associated with carnallite comprises a large part of its volume, so that a zone Figure .1. rich in carnallite will have II large hydrogen index. 65 The hydrogen index of pure carnallite is per centDownloaded from http://onepetro.org/JCPT/article-pdf/5/03/145/2165444/petsoc-66-03-06.pdf by guest on 27 September 2021 (4). SylVite and halite have an index equal to zero, except for a small (1 to 2 per cent) volume of included water. Again, the insoluble content of the zone affect~ this log, and it should be taken into account if it is found to be very large (greater than 5 per cent 1. An empirical chart similar to Fig/ll"i' .4 can Le made for each well to be interpreted tf) obtain the h~rctrogell index_ A pivot point at GOO API units and a 65 pel' cent hydl·ogen index defines olle end of the straight line lOll tiemi log paper 1 _ The Neutron Log value (API units) in a clean salt :~one and a 1 per cent hydrogen index define!; the other end. The pivot point used for this example applies only for the tool, spacing and source type noted on the chal·t. Different pivot points must be determined for different tools. Precise interpretation using the Neu tron Log is limited to beds thicker than 2 feet. A Sonic Log is employed as an aid to determine the insoluble content. This is a required fador if an accurate interpretation is to he made. Knowledge of the insoluble cuntent is al!io necessary bQcause exces sive amounts of insolubles can make an apparently good zone commercially unattractive, as it is an ex pensive process to refine these imllUritles from the final product. Studies on labol·atory samples and field con-elatiom:\ (3) have given ~;onic travel time \"ulueH in halite, sylvite, carnallite and insoluble material as 67, 74, 78 and 120 microseconds per foot respe.diveb'. Thifl data. combined with the inf(lrm~ltion which can be
NEUTRON DEfLEc:nON - API IHIITS determined from the Gamma Ray Log and the Neu tron Log, can be used to set up simultaneouR equa Figure 4- tions to solve for the per cent of halite, sylvite, car nallite and insolubles (w, x, y and 7. respectivelY*l (51. Table I shows the values used for the coefficients of the equations. These coefficients represent values TABLE I of the parameters for 100 per cent pure minerals. COEFFICIENTS FOR POTASH LOG EVALU,~TION A~ it is a:isumed that only halite, .sylVite, carnallite and :lome insoluble:i are present, their total volume mllst comprise the total formation volume. This is IHillCTal Halite Sylvite Carnallite Insolubles represented by equation l. w y z Symbol w + x + }! + z = 100 equation AppaTeIIl /(20 Conlent 0,00 0,63 0.17 The total apparent K~O content of the formation, (fractional) based all the Gamma Ray Log. is made lip to each com re~peetive K~O Hvdrogw Index 0.00 0,00 O,G5 0.30 ponent fraction multiplied by its value_ - (fractional) Thus: 0_63 :.. 0.17 Y 0.05 z = K:!O""" cqll~tiun 2 Sonic Travel Time 67.0 74.0 78.0 120,0 + + (microseconds per [oot) -! 14~ The Journol of Canadian Petroleum --~ --~-_ .._- -' --' ','-- TABLE II EVAPORITE MINERAL PARA]\,IETERS Apparel'lL Sonic K,O Hydrogen Apparent True Travel Rating Index Density Density Time 1\1illerat Grams per Grams per cubic cubic lVIicroseconds Per cent Per cent centimeter centimeter per foot { Anhydrite...... " .... 0.0 0.0 2.98 2.96 50 Carnallite...... -...... , . ... 17.0 65.0 1.57 1.61 78 Gypsum .. _...... - ...... - 0.0 49.0 2.35 2.32 52 Halite ...... -- ...... 0.0 0.0 2.03 2.16 67 Kainite...... _...... ' .. _.- 18.9 45.0 2,12 2.13 - Langbeinite...... 22.6 0.0 2.82 2.83 52 Polyhalite. _...... - . 15.5 15.0 2.79 2.76 57 Downloaded from http://onepetro.org/JCPT/article-pdf/5/03/145/2165444/petsoc-66-03-06.pdf by guest on 27 September 2021 Sylvite. __ ...... 63.0 0.0 1.86 1.98 74 Insoll1bles. _...... - ... .. ' ... .., . 5.0 30.0 2.60 2.60 120 ;-".. ., ~ .. .!.' ;'-'.' "016.0 't. 3~DO. lO.S ~ b.l ~.) oJ />.1 89.3 4o.'Ii .1 ... ~ ~DI5.5 ez. J~5D. TO., . 8.1 4.oIr .s '.9 90.1 3_5 D.D ].1 _,. -';01".00. _ 60•• 3l6D•.__ 'O.~ __.6.1, _J.~ __l.:\ -----l._.__BJ.:4 __ J._D ...__...._~ • __ 'liOt9.5 1.r;0. ~lO(l. 'o.~ 6.l Z_g 3.2. 13.~ BO.~ 9~3 .5 B.1 - ~020.0 u,,~ 3100. '2~O 6.l ~.~ 2.0 1~.1 n.) 1I~7 .1 ll.~ _~o2e.~ 22'. HOC. 12_5 6_1 5.a I.) n.o 7D.6 L~~7 • .i H.5 FiglO'e 5.- Potash Log Analysis. A similar expression for hydrogen index gives: K:!O~ ~ K;!Onpp - 0.05 Z ..equation 9 ·1 ...equation 3 0.65 l' + 0.30 z - '0 K:tO, ~ 0.63 x ..equation 10 ,\Vyllie's time avel·age equation extended to four components gives: K 2 O.. ~ 0.17 y .equation 11 67 w + 74 x + 78 ~r + 120 :t = 100 ~tlo'f ... _ ..equation 4 As an arithmetic check, one should compute the The value for L\tlcg is read directly from the log; sum w + x + y + z, \\'hich should equal 100 per cent. the values for K:!Oapp and ¢n are derived from the Also the sum K,O, + K,O. should equal KoO,. charts (Figures 2, 3 and 4). A simple crosscheck of the data determined from Reducing these equations to obtain the required equations 5 to 8 can be accomplished by running a values - " W", "x", fly" and HZ" - in terms of the Formation Density Log_ The actual density reading derived values from the three logs gives the following: can be compared with a computed density value, cal culated from equation 12. z = 2.07 .u bOG - 0,23 K 20"I'/l - 0.29 Technology, July-September, 1966, Montreal o I. I.: "J ,,, ". "" Downloaded from http://onepetro.org/JCPT/article-pdf/5/03/145/2165444/petsoc-66-03-06.pdf by guest on 27 September 2021 .,, I," , " ~ . ., ,. ."", , I ~ •I ,. a, ." .'- ". , ". , . . , ,- .0 ,., , N ", ". ., jj I '-I,' , " Figure 6. 148 The Journal of Canadian Petroleum The four logs, Gamma Ray, Neutron, Sonic, and but there is no problem concerning interpretation as .' Density, could be used to set up equations for a five this amount of insoluble matei-ial would condemn the mineral matrix, but is not often necessary, because zone even if it had contained sufficient potash to be halite, sylvite, carnallite and insolubles usually pre considered a prospective commercial orebody_ Bed dominate over other potash minerals. boundaries are clearly defined and correlate exactly As the calculations, although simple, are quite with the core. tedious, a computer program was developed to handle Bed thickness corrections should be applied to the this task. In order to use log values directly, the Gamma Ray Log readings before inserting them in charts of Figu..1·es Z and .& were reduced to tabular the machine computation program. Thin-bed effects form, as shown in Tables III and IV. The data were can be seen on the examples where rich are intervals normalized to 6-inch hole size with a mud ,,,'eight of are less than 3 feet thick 7.2 pounds per U.S. gallon (diesel fuel). Some discrepancies with core analysis, caused by ~~., l 'J, Corrections for different hole sizes and mud weights problems other than thin beds, can be noted in the J were developed in equation form. Any value of hole higher grade sylvite and carnallite intervals, The Gam .~ size from 6 to 12 inches and mud weight from 7.2 to ma Ra}' logging tool does not respond linearly to K:lO 12.0 pounds per U.S. gallon can be accounted for by content, and resolution is poorer in high-K:lO-grade using the following equations: ores. Downloaded from http://onepetro.org/JCPT/article-pdf/5/03/145/2165444/petsoc-66-03-06.pdf by guest on 27 September 2021 . ' '.' 320 (d - 6) '( Due to the logarithmic response of the Neutron GR, ~ GR,", {II + 0.05 (d - 6)1 + GR,", + 100 ( Log to h}'dl-ogen index, the resolution of this logging . [1 + 0.10 (rn - 7.2)[ .equation 13 tool in the zones of high carnallite content is not as good as in the lower grade carnallite. Figure 7 illus :.....-.- N, ~ N,", II + 0.05 (d - 611 .equation 14 trates this effect. Here, a high-grade sylvite bed f''';:; _ These are approximations only, but are sufficiently grades into a high-carnallite-content zone_ The com accurate for the range of values stated above_ The parison between core analysis and log calculations is computer program first corrects the log values, and not as good as in Figlt1-e 6. However, no one could then enters the tables, interpolating between points if mistake the content of the zone. The log-derived bed necessary, to obtain K20::lP~ and cpo. boundaries agree very closely with the core analysis. Unfortunately, this bed is not thick enough to be A sample of the results from an electronic compu considered commerciall}r attractive. ter analysis of digitized log values is shown in FiguTe 5. These same values are plotted against the core The result of the bed thickness and tool resolution assay results in Fig1.o·e 6. Close correlation is shown problems is a pessimistic sylvite and carnallite assay in many cases. in thin rich zones. Limited accuracy is possible in thick, rich ore3, and good accuracy is attained in Those intervals in which large discrepancies occur low- to mediurn-grade beds. are probab1;}T due to the different volumes of investi gation of the core and the logs_ The logging instru The value of the Caliper Log is evident from Fig ments are capable of analJrsing as much as 100 times nre 7. The sylvite fraction could have been in error the volume of a conventional 2%-in. core_ Particular by as much as 14 pel' cent in Example 4 had the hole examples of such discrepancies occur at 4,014 to 4,022 diameter been assumed equal to the bit size of 9 feet and at 4,031 to 4,044 feet on the insoluble ca[cu inches_ lation shown in Figu..Te 6_ A more complex computer program has been de The high insoluble content interval between 4,053 veloped to translate directly the recorded logs to a and 4,056 feet on Fig1lre 6 is shown quite clearly_ The computed log_ A sample of the output from such a absolute values do not agree ,\'ith the core analysis, program is illustrated in Figure 8. In this application, the logs aloe digitized, calculations are performed, and the results are plotted by an incremental digital plotter in a continuous sequence, eliminating the TABLE III source of en·or in manual digitizing and plotting_ "~So GAl'v1lV1A RAY 1(,00 ," Gamma Ray DeHediotl API units TABLE IV ------1------, ; 0 .. o NEUTRON vs. 4>" , 45 .. __ 2.5 J 90 .. 5.0 135 .. __ 7.5 Neulron Dejler/ioll 175 .. 10.0 API units ! " 220 .. 12.5 265 .. 15.0 6000. o 310... 17.5 4300. I 355.. 20.0 3600. 2 400 "". 22.5 3200. 3 435 .. 25.0 2600 ... 5 470 . 27.5 2400. 7 505 .. 30.0 2200. 9 530 .. 32.5 2000 . 12 550 . 35.0 1700 . 16 565 .. __ 37.5 1500 . 20 580 . 400 1300 .. 26 590 . 42.5 1100 .. 35 600 .. 45.0 800. 50 605 .. 47.5 600. 65 999. 99.9 o 99 Technology, July-September, 1966, Montreol 149 ," 'I "~'~~~T~'I;T':,"I,I I' " ,,' IlL t:J ' J:.....I Ir- " 'III II· -:- ' ;- I~I r' Downloaded from http://onepetro.org/JCPT/article-pdf/5/03/145/2165444/petsoc-66-03-06.pdf by guest on 27 September 2021 , " ~_ ~_...J __ LJ,'j:~___.- +- '.' _.. I--'t-+I-+-Y~' ,~J, L ;••...: "I :- I " , '- ' 11,1 , " 1 ' I' I: 150 The Journal of Canadian Petroleum SYLVITE· .. CARNALLITE i:: IIIISOLUBLES ::!: 7 ...', " '.'.. " '~~ . Downloaded from http://onepetro.org/JCPT/article-pdf/5/03/145/2165444/petsoc-66-03-06.pdf by guest on 27 September 2021 Figun; 8.-Potas'~ Log Analysis. The computer program can be adapted to any com puter. At present, logs must be digitized b)r hand_ The output is in the form of punched cards which are .~PPENDIX used for off-line listing. Plotting of results from this LIST OF SnlBOLS list is accomplished by hand, but the data can be re corded by the incremental digital plotter if thc Computer punched cards are first converted to digital magnetic Program Definition Uuits or paper tape. Symbol Equil'aleul The direct digitizing, computing and plotting tech~ D Deptb Feet GRlc &, G Gamma Ray Log reading nique is available as a commercial service at those from log at depth D API Units centers which have the suitable hardware. GR Gamma Ray Log reading corrected for hole size API Units CONCLUSIONS GR. GRC Gamma Ray Log reading corrected for hole size It has been shown that with three standard well and mud weight API Units logs it is possible to evaluate quantitatively the Prai Nlo&' XNL Neutron Log reading at rie Evaporite potash beds. The arithmetic is simple depth D API Units No XNC Neutron Log reading cor- and the presentation, as illustrated in the examples, rected for hole size API Units is informative and self-explanatory. Auxiliary surveys w W Per cent of halite in such as the Caliper Log or the Formation Densit)r formation Per cent Log, can be used to supplement and cross-check data x X Per cent of syh'ite in formation Percent derived from the three basic logs. I' Y Per cent of carnallite in Computer techniques are admirably suited to this formation Per cent z Z Per cent of insolubles in interpretation approach, and eliminate much tedious formation Per cent hand calculation. The graphical display of the calcu IC,O, XICOT Total K:!O content of lations allows correlation, mining studies and map fOlmation Per cent ping to be accomplished' with ease. 1(2 0 5 XKOS K 20 contentof formation in s~'lvite Per cent ACKNOWLEDGMENTS K,O. XKOC K:!O content of [onnation in carnallite Percent The authors wish to express their appreciation to K 2OIlPI' XKOA Apparent K 20 content of formation [rom Gamma the companies whose logs were released for presenta Ra}' Log Per cen t tion in this paper. Access to the confidential informa P Hydrogen Index from tion required to complete this study is gratefully Neutron log Per cent acknowledged. d HS Borehole diameter from Caliper Logar from bit Inches REFERENCES size m WM 1vlud weight Pounds per U.s. gallon (1) Goudie, ill., ".i\'liddle Devonian Potash Beds of Cen tral Saskatchewan/' Sask. Dept. of 1l1inel·al Re T Sonic travel time from Microseconds sow·ces (1957) - unpublished report.. Log perfoot (2) Saskatchewan Depu1·tment of l1iineral Resources, P. Computed formation Grams per cubic "Potash in Saskatchewan," (1965)_ density centimeter (3) Alger, R. P. and Crain) E. R., "Defining Evaporite Deposits with Electrical Well Logs," T1·ansactiolls B (1) Gamma Ray Log values Northen~ O/~io Geological Society Second SY1nposi1l1n in Table III API Units on Salt, (1965). C(J) K 20 npfl values in (4) Alger, R. P., Raymcr, L. L., Hoyle, W. R., and Table III Per cent Tixier, M. P., "Formation Density Log Applications corresponding to B (1) in Liquid Filled Holes," Journal of Petrolcum Tech R (11.) Neutron Log values in nology, (March, 1963). Table IV API Units (5) Sa'IH'c, W. C.; UDetermination of a .More Accurate V (K) Hydrogen Index in Table Porosity and Mineral Composition in Complex Lith IV corresponding to ologies with the Use of the Sonic, Neutron and Dens R (IC) Per cent ity Surveys:' JOlu"1!.al of PdnJlelun Technology, (September, 1963). Technology, July-Sepf'ember. 1966, Montreal )51 APPENDIX II (33) Dt"wall, ./. T., "Neutron Log Correction Charts for B~rehole Conditions and Bed Thickness," Jounwl BIBLIOGRAP HY OJ PctJo[eulII Tcchnologu (l!l55). (34) Fc'arolJ, R. E., "Radioactivity Well Logging," W('f{ (I) Su..::kafchewUH Dellal'(lIlellt 0; .1Iil1cl'ul R('.~u/{rces. Stu lley,,; hiC. (HI-!6). "Subsurface Mineral Distribution Map," (1965) (35) A{g('/'~ R. p.., and era.ill, E. R., "Definihf.{ EVUpol·ite i\fu}J #8-33_ D.eposlts Wlt~ Electl'lcnl Well Logs," Tnmsadiolls (2) Sas.katchclt.:an Dcpaj"rmc/(t of Mim'ral Rc,'ww"c.es. Northern OhiO Geological S()cir:tJj SeccllId S!lmposiltlll "Subsurface Ex!)loration GeiJlo~'Y and Deyonian Oll So,!l (1965). Stratig-raphy of j\'learlow Lake, Prince Albert and (3fj) l'i.J.:ia, M. P., A fye,-, R- P., and Doh, C. A., "Sonic Yorkton Arl!:ls." - Report #15. Logging," Journal of Petrolfllll1 TechllvlllYII (Mav. Lane, D. M., "Dawson Bay Formation in the Quill 1059) . Lakes - Qu'Appelle Area," Sl.1.sk. Dept. 0/ Millnal (;37 ) Kokes!•. F. P., Schwartz, R. J_, Wall, 1(". B., [lnd Resollrc'_'ii (1959) - Report #=:18. .l/orris, R_ L. '"A New Appmach to SOllic Log-gill'" (4) Pt'UI'S01t. TV. J., "Developmenls in Potash in Sas and Othcr Aeoustic 3-'1easurement:s." Trflll,~. A/ME katchewan," C.I-iV!. Bill/dill, (Odobel", 1960). (HHi4). (5) l'Ctll'IWI1. W. J.. "\Ve.stern Canada Potash and its (38) ..llgc,', Raymcr, Hoyll! and Ti:xicr, "Fol'mation DCll Future Prospects." EIC JoltrtJul (August, l:JGOL .:'iit~· LOI! Applications in LiqUId Filled Holes," AIMS I (;j SCh"1Cl"ettJln, "Genesis of Potash Rock::; in the -;yrid 0%2), Paper #SPE.,l,35. die Devonian." A.4PG Bullet,iil (July, HHi.,l,). (3!J) ~I·ahf, J. S., TiUman, J" JO!ItIStVlIl!, C, W., and :-l/ (7) TOlllhiH8, R. l'., ·'Potash in Saskettche'...·an," C.I.i\1. gn, R. P.; "The Dual Spaclhg Formation Densit\' Dufh'till (November, 1954). LIJ,!!;"," Trll1l'i. AIME (1964), . (8) "·(Lrrlluw, N. C.. ·'Potash Re~eal·ch," Saskatc.hewan (40) [iuymcr, L. L., and Biggs, 11'. P., "Matrix Chamc Enf!,:ineer 1)963). tetistics Defined 11 ..... Porosity Computations," TraJls.Downloaded from http://onepetro.org/JCPT/article-pdf/5/03/145/2165444/petsoc-66-03-06.pdf by guest on 27 September 2021 (!J) TOtHkiwi, R. 1'., "Canadian Potash - The Wodd's 01' Soc. oj Pro"- Well [.09 .·lr.. rtly,~t,'1 J1[cdiltg (1963). FutLll'e Supply of a Vital i\Iineral," CJ.:\I. Tl'all (41) S'UI1J"l:', 11'_ C., and fJllrke, .J. A., "Detul'lllinatioll of sue/tOllS (1962). a True Porosity and Minenll COlllpl)~ition in Com lID) Guudie, M., "Middle Devonian Pota::.h HeeL, of Cen plex Litholog-ies with the US!! of the Sonic, Neutt·on tral Saskatchewan;' Sask. Dept. of MilH'ru[ R(' and Density Sun'eys," JVU1'IW{ or Pl'll'olt'/w! Gll- soltn.:cs (HH,7) - unpnblished repol:t. gi1/CcrillrJ (SeptembN', 1~Ga). . \ ll) DL1rtE('y, C. M" ··Pota5h," Dept. of !HiJJt~S alld Tech Lfl.1!d('~, Ie [{" "Origin of Salt Depo~it-;," Trallstlc nical SHn>rfls (19G3). tlOllS oj Isf Sult SympnsiulJI (1962). (12) Milllc, Shouldicf' alit! .'\1('/.'1011. "Collapse Strurtul·es Tit/lIlull, .J, ~nd ""'uM, J_ S., "The Physical Founda Related to Evaporites of Prairie Formation, Saskat tions of Formation Density Log~inl?: (gammn-gulll- {'he-wan;' GSA Bnl[ctlll (April. 1964). Intl)" G~'f)P{II/.'1ic's (l965)_ -- (13) Stll'jil., D. £., and Ha.bsoll, G., "Prairie Evapol'itr IHI Baillie, .-1.. D., "Devonian System of the Williston :i\Iappin,g in Minton A.reas of Saskatchewan Em Basin Area," J/il/es IJrallt"h. r/'()lli'I('(' n; Illall.tollU ploying- Se.ismle .l\'Iethods," Gcophysic.'! (December. ( 195:]), Pub. 52-5. 10fj4) . Bi.'1{wjJ, R, .-t., "Saskatchewan Exploratory Pro ( 14) 8uskatche WIt II fJ()jJa/'tJJI~'JI t o{ Jll iH('l'al Rt;j'OII I t,:(',<;; g-.'ess and Problem;;," R/(thf'J"(ol'd ,1l1'IIWl'ifll l'ol,lllll' "Potash in Saskatchewan," (Janllary, Hl(5). (1954) . ( 151 "Potas~ium b~r 1S2 The JouUlol of Canadian Petroleum