A Proposed Classification of Petroleum and Natural Gas Fields Based on Structure2

ECONOMIC GEOLOGY

WITH WHICH IS INCORPORATED

THE AMERICAN GEOLOGIST

Vox..V SEPTEMBER, I9•O No. 6

A PROPOSED CLASSIFICATION OF PETROLEUM AND NATURAL GAS FIELDS BASED ON STRUCTURE2

FREDERICK G. CLAPP.

Although not necessaryto take time here to give the history of the "anticlinal theory" in full; it seemsdesirable to point out severalprominent features of that theory and of its history and practical application, before giving some of its limitations and stating the standingof the theory to-day. For over fifty years various geologistsand others have publishedpapers attempting to solve the problem of the distributionof oil and gas fields. Among others T. Sterry Hunt (I859 and I863), E. B. An- drews (I86X) and Hans Hoefer (I876) long ago recognized certain general relations of oil and gas pools to the anticlinal structureof a region. It remainedfor Edward Orton and I. C. 5¾hite, however, to bring the theory before the oil and gas world in sucha way as to force a measureof belief in it. Later investigationshave expandedand limited the theory and have made its applicationmore practicalto the operators,but much of the credit belongsto the early investigators. The anticlinaltheory was, in brief, that oil and gas were originally widely disseminatedthroughout the formationsin which they are found, or in contiguousformations, and their segrega- tion was believedto be due to the different specificgravities of oil, gas and water. If a porous stratum contains these sub- 'Read before the Geological Society of Washington, March 9, •9xo. 504 FREDERICK G. CL.4PP. stances,when it is tilted by geologiccauses they will arrange themselvesaccording to specificgravity; the gas, being lighter, will be driven into the higher parts of the stratum (towards the crestof the anticline), the oil will be floatedon top of the water, •vhile the water occupiesthe lower portions of the stratum (those nearest the syncline). Limitations of the .4nticlinal Theory.--In practice, however, the anticlinal theory frequently did not meet expectations. Operators, after making careful locationsof .wells based upon this theory, were frequentlyrewarded only by dry holes. The successes,overshadowed by the fancied failure of the theory, were lost sight of by practicaloil men, and thus the theory fell largely into disreputefor a time. The reports of the United States GeologicalSurvey and the various state geologicalsur- veysthroughout the countrycontained references to the theory, and eachwriter tried to show the geologicalrelation of fieldsin the particularterritory covered. In many casesgeologists were successfulin finding sucha relation, and in all casesthe major axes of poolswere found to correspondin a generalway with the main anticlinal and synclinalaxes. Many of the gas pools correspondedclosely with the crestsof anticlinesand seemedto provethe theory. Other casesprevailed, however, in whichthe relationwas lessstriking and there were somein which no relation could be determined. To explainthe defectsof the theory,various so-called "limitations" were formulated. The limitations, like the original theory,were imperfectin their application.They have been added to and rehashedby various geologists,until now little seemsto be left of the original "anticlinal theory." However, greatadvances have been made consequent upon detailed map- pingof geologicalstructure by governmentsurveys and private geologists.While we have not yet reached,and may never reach, the state where a productioncan be located correctly everytime, we may truthfully assertthat geologycan now save a considerableproportion of dry holes. Moreover,much of the territory held for yearsby oil and gas companieshas been shown to be not worth paying rentals on. CL,'ISSIFIC,'ITION OF PETROLEUM FIELDS. 505

State•nentof the Structural Theory.--The writer of thispaper prefersto usehere the term "structuraltheory" instead of "anticlinal theory" to explain the relations which accumulations of oil and gas hold to geology within certain limitations, evenwhere n.o definite anticline or synclineexists. The theory, as understood,is as follows: Through some means or other, by organic or inorganicagencies, or by both, the petroleumand gas have got into the sandstoneor limestoneformations in which they are found. The deposits may have originated through the decompositionof plant and animal remains on an ancient sea bottom, as the adherentsof the organic theory claim. Or they may be the product of chemicalaction on carbidesof iron or other substancesexisting under a stateof potential fusion deep in the earth, as the adherentsof the inorganictheory claim. Or certainpetroleum deposits may be of organicand certain deposits of inorganicorigin, which seemsa conservativeopinion. Whichever theory be true, the oil, gas and water in the formations (assumed to have been approximately horizontal at the time the substancesentered them) were at first widely diffused in the oil sandsor contiguousstrata. Such is true at the present time in somesections of the world, notably in severalcounties in western West Virginia fronting on the Ohio River, where only small quantitiesof oil and gas, too slight for profitabledevelop- ment, have been found. The dip of the rocks in that sectionis very slight. Where the beds have been folded, however, as is the case throughoutthe greater part of the Appalachianregion and in most of the other oil fieldsof the world, the slopeand changesin dip have enabledthe oil, gas,and salt water to separateout accordingto their relative specificgravities. This separation and concentrationmay have been assistedby rock pressure,hy- draulicpressure, seepage, capillarity, molecular attraction, internal heat and other causes;but zvhatever causesprevailed, the law of gravitation,being ever operative,must be consideredof most importance,and the separationand subsequentaccumula- tion was in the order of the densities of the substances. Thus on a stated anticlinal, monoclinal or quaquaversalstructure, the 506 FREDERICK G. CL•tPP. gas is nearest the top, the oil lower down, and still lower is the salt water, when that is present. All this is accordingto the original "anticlinal theory." Sometimesthe poolsoccur at the top of the anticlines, sometimes lower down on them, sometimes in synclines,sometimes merely on monoclines. Conver•7enceof Formations.•Oneof the mistakesof the earlier geologistsand oil men was the assumptionthat the surface formations were universally parallel with the oil "sands.TM This is true in many fields,but in othersit is far from being so. For example, the Clinton sand in southern Ohio is not even ap- proximatelyparallel with any surfaceformation; but the interval between the Clinton and the surface strata increases toward the east at a rate ranging from 3¸ to zoo feet per mile, thus bring- ing the Clinton rapidly to a great depth. This convergencewas illustratedby Orton,Volume VIII. ot•the "Geologyof Ohio," but it is now known to be lessuniform and in placesmore marked than Orton supposed. With the assistance of a few •vell records within a few miles of any locality, where developmenthas been commencedor test- ing done, it is possibleto work out the structureof this sand to a gooddegree of accuracy.This is doneby firstmaking a contoured structure map of some surface formation; then the exact intervals from this stratum to the Clinton sand are determined in whatever wells have been drilled to this sand in the vicinity, and a convergencemap is preparedon the same scale. This was first done, so far as I know, by the writer of this paper, for the Amity, geologicalfolio, and for Bulletin 3oo of the United StatesGeological Survey, in •9o7, and the nameisochore lines (meaning lines of equal interval) xvasgiven to the contour-like linesdenoting the convergence. Fig. 52 is an example of sucha map from Bulletin 3oo. In the Amity region,however, the interval representedwas that betweenthe Pittsburgcoal and the Gantz (or Hundred foot) sand,the Clinton sand never havingbeen reached so far east. In Bulletin3•8 of the United StatesGeological Survey the samemethod of representingcon- xIn this paper the word "sands" is used, being the well-drillers' term for the productive formations. CLASSIFICATIONOF PETROLEUMFI•ELDS. 507

vergencewas usedby Griswoldand Munn, who appliedthe sim- pler, and perhapspreferable, name "convergencesheet" to maps of this type. In preparing a map of the structure of the Clinton sand, of

Fro. 52. Convergence map of the Amity quadrangle, in Washington County, Pa. The isochorelines (lines of equal interval) show the changesin interval between the Pitts.burg coal bed and the Gantz sand throughout the quadrangle. (From Bull. 300, U.S. Geol. Survey, •9o7.)

the formations below the Greenbriar limestone in southwestern Pennsylvania,and of certain other formations, a convergence , map must be prepared. The necessityof this is due to uncon- formities. The convergencemap is superposedon the contour map which showsstructure of the surface formation, and the intervals representedby isochorelines are subtractedat every point, the resulting elevationsof the sand being noted on an overlyingtracing. The sand is then contoured. In practisethe finishedmap always showsthat the structurebears some deft- 508 FREDERICK G. CLAPP. nite relation to the occurrenceof gas and oil. The direction of the extension of the field can therefrom be predicted witha good degree of certainty; and, if there has been no deep sand development,the locationsof producingwells can be predicted with fair probability. Fig. 53 is a map of the Clinton sand in a •mallarea in southernOhio, where the structure was determined in this manner. It will be noticed that the positionsof oil de-

EXI=!..AN A TI ONS. • 2400• STRUCTURE CONTOUR L/NE$ SHOW/NO DEPTH OF CLINTON SAND BELOW $E,•-LEYEL. DRYHOLE, 0,•$ WELL, OIL WELL, • • • •

SCALE ONE MILE

Fro. 53. Example of sub-classI(c). An oil pool coincident with a change in monoclinal dip, in the Clinton sand of southern Ohio. Structure map of the sand in a small area determined by the convergence method. velopmentcorrespond well with the strike of the sand, and that the productivewells lie on the outer edge of a slight structural terrace. Fig. 54 showsthe surfacestructure at the samelocality, CLASSIFICATION OF PETROLEUM FIELDS. 509 and illustratesthe entire lack of agreementwith the structureof the-sand. The foregoing summarieshave been made becauseit seems necessaryto take them into accountin making any classification

880

EXPLANATIONS. •885 • •TRUCTURE CONTOUR •INE• •HOWING E•EVATION OF • THIN RED OF 8•ACK FLINT ABOVE TIDE. • DRYHO•E, • • WE•. • 01• WE•L

8CA•E ONE MILE

FIG. 54. Structure map of the surface formations in the same area as Fig. 53. of fields based on structure. The structure of the productive stratumitself must be considered independently of the confi#u- ration of structure of any surface formation. Classificationof Oil and Gas Accumulations.--In order to il- lustrate the differencein geologicconditions, a tentative classification of fieldshas beenmade. Only accumulationsof economic 51o FREDERICK G. CLdIPP. importanceare considered,and even some prominentfields, which the writer has not had opportunityto investigate,'-are not consideredat all. The classificationdoes not pretendto be per- fect, by any'means, and doubtlessit has many deficiencies,but it is publishedhere for whateverit may be worth as a geologic classification.There is no attemptmade to distinguishoil struc- turesfrom gasstructures, nor betweenthe differentkinds or ages of productiveformations. Thesecould be broughtin as subdi- visions in the classification if desired. The object of the tentative classificationis to show that accumulationsof oil and gas can be grouped into classes,each division of which follows a specialrule of structure,and all of which have certain aspectsin common. The classificationpro- posed is as follows:

Cn^SS•F•C^T•ON OF Om ^N•) C^S ^CCU•tUn^TXO•S. I. Where anticlinaland synclinalstructure exists. (a) Strong anticlinesstanding alone. (b) Well definedanticlines and synclinesalternating. (c) Monoclinalslopes with changein dip. (d) Terrace structures. (e) Broad geanticlinalfolds. (f) Overturned folds. II. Domes, or quaquaversalstructures. III. Sealed faults. IV. Oil and gas sealedin by asphalticdeposits. V. Contactof sedimentaryand crystallinerocks. VI. Joint cracks. VII. Surrounding volcanic vents. ClassI. •Vhere Anticlinal and Synclinal Structure Exists.. This is the type of oil and gas accumulationwith which we are most familiar. It includes,a large majority of the known oil fields of the world. The Appalachian,Mid-Continent, Illinois, Indiana, Wyoming, Colorado, northern Louisiana and northern Texas, and someof the California fieldsin this country,and the Russian, Austrian, Burma and Borneo fields in the eastern hemisphere,all belong to this class. It is divided into five sub- CLASSIFICATION OF PETROLEUM FIELDS. 5 1 I. classes,in order to distinguishbetween various structural relations in which oil is found in connectionwith anticlinesand synclines. Subclas• (a). Where Strong •lnticlines E.vist Standin# •llone.--In this division I would include fields that bear a direct relation to very. pronm•nceduplifts, easily recognizable,and which constitutea marked geologicfeature of the region. The only prominentexample in the easternfields of the United States is the famousEureka-Volcano Burning Springsanticline in West Virginia, whichis 25 milesin length,ranging in directionfrom North •o degreesWest to North 20 degreesEast, being from an eighthof a mile to half a mile broadon its flat crest,and having side-dipsof from 20 to 60 degrees. This differssomewhat in directionfrom the main Appalachianfolds, and was probably producedat a differenttime. It is oneof the earliestrecognized anticlinesin the country,and probablyhas had as many wells drilled on it as any other anticline. It has been describedby White,• Andrews and Evans. Some of the California fields havingsharp anticlines probably belong in this class. Possibly the Baku field of Russiamay alsobelong here. Subclass(b). Where Well-defined•tlternatin# •tnticlines andSynclines Exist.-•This is reallya compositeof subclass(a). With minor exceptionsit includesthe entire Appalachianfield in Pennsylvania,West Virginia, and easternKentucky, southern Indiana and Illinois, the Oklahoma fields and the Caddo field in Louisiana, the north Texas fields,and thoseof Colorado, Wyoming and Montana. The Caddofield has nothingin commonwith the Beaumor/t andJennings fields and other fields of theCoastal Plain of Louis- iana and Texas, but it is very similar in structureto the fields of Pennsylvania,West Virginia and Illinois. The properstruc- ture in northern Louisianais affordedby the Sabineuplift. The final distributionof oil and gas in the Caddofield is presumably dueto slightanticlines and synclinesand differencesin porosity of the UpperCretaceous formations which exist there. • I. C. White, Bull. Geol. Soc. Am., Vol. •o, •899, pl. 29. FREDERICK G. CLztPP.

Several of the oil fields of California also belong in this class. Examplesare Coalingafield and the Los Angelesfield,

AMITY OI/ADRRNOLE BROWNSVILLE OURDRRNOLE

EXPI..ANA 7'10N•, •STRUCTURE 0031TOUR8 8HOWI31G ELEVA 7'1031 OF PITTSBURG COAL BED...... A31TIOLI31ES. t • mSy31OLI31ES. •-- O/L POOL& •-- GASPOOLS.

Fro. 55- Example of sub-classI(b). Structure of five quadranglesof the U.S. GeologicalSurvey in southwesternPennsylvania, showing relation of the oil and gas pools to surface structure. (After Campbell, Stone and Clapp.) accordingto descriptionby Eldridge,• andby Arnold and Ander- son? In •895 Noettlingestablished the fact that the oil fields •Geo. H. Eldridge, Bull. 213, U.S. Geol. Survey, pp. 3o6-321, 19o2. CL.4SSIF1C.4TION OF PETROLEUM FIELDS. of the Irawaddy, in Burma, correspondwith the structural theory; and I believethey belong in this subclass.The oil and gas in those fields are directly related in their accumulationto anticlines and domes in the Miocene sandstone. The rocks in the fields comprisedin subclass(b) are folded into alternating anticlinesand synclines,bounded by moderate dips, seldom more than 3ø degreesfrom the horizontal. This is the subclassto which the anticlinal theory as originally pro- pounded applies strictly; the gas occurs in the upper part of individual sands or "pay streaks," the oil occurs somewhere below the gas, while salt water, if it occurs at all, fills in the remainingspace in the sand (if the latter is uniformly porous), or occursin the separate"pays" (where the sand as a whole has not a high degreeof porosity), all thesedeposits being con- trolled mainly by the force of gravity. Where oil or salt water occurshigher in the sandthan gas, it is presumablydue to sharp changesin the dip, or to a multiple nature of the "pay streaks." Subclass(c). Where there are Monoclinal Dips; i.e., where the RocksDip in one GeneralDirection Throu#hout (•llthou#h the Dip may Vary in Steepness).--The majority of the oil and gas poolsin southeasternOhio belongin this division. Definite anticlinesare not so commonin Ohio as in Pennsylvaniaand West Virginia, as the prevailing dip of the formations is all in one general direction (towards the southeast). However, the applicationof the structural theory, properly understood,is al- most as certain of profitableresults as it is in Pennsylvania,be- causethe dip is not uniform, but varies from flat to over 5 degreesfrom the horizontal. Roundedor somewhatelongated semi-anticlinesof subclass(b) are occasionallyfound. Oil and gas can be predictedin subclass(c) throughrecog- nitionof the principlethat any changein the rate of dip is to be consideredas a possibleplace of accumulation;and that such a place of accumulationonce discovered,the pool can be easily extended(within structural limits) by followinglines of hor- izontalityin the sand. SubclassI(c) is illustratedin Fig. 53- •Ralph Arnold and Robert Anderson,Bull. 357, U.S. Geol. Survey, pp. 7o-7I, :I9o8. 514 FREDERICK G. CL.4PP.

To this subclassmay also belong (in part at least) the Flor- enceoil-field in Colorado,where the oil existsin bedshaving a dip of less than 5 degrees,lying betweendips as great as 20 degrees (according to Fenneman,1 and occupyinga portion of a structuralslope where the sandslocally are rather flat for a fexv miles; being a semi-terracein the formation. In this field the bedscarry little water. Subclass'(d). Terrace Structuresare an Exa9t7eratedForm of the Flattenin9sof Dip, Included in Subclass(c).--As a rule where gas occurs, it is found on the outside of the terrace and oil on the inside, though this is not an infallible rule. The changein the rate of dip, forming a local interruption, seemsto be the essentialfactor. An example of a structural terrace in southwesternOhio is shown in Fig. 56. The effect of terrace structure was first explained and illustrated by Edward Orton in •886.2 The terraces describedby him were in the Trenton limestone fields of northwestern Ohio. In the Findlay field the oil and gas were found in two terraces, separatedby a monoclinaldip. The upper terrace yielded dry gas, the lower terrace yielded oil and water. Orton gave the name "arrested anticlines" to structural terraces,and cites the Macksburgfield of southernOhio as an example.a The terrace structureof the Macksburgfield was first recognizedand describedby Newhall in the same volume. During the past decadehundreds of similar terrace structures have beendiscovered throughout southeastern Ohio and to some extent in adjacent states, and most of them are available for oil and gas development. Generally, though not always, the structure can be determined from the geology of the surface without the need of borings until one is ready to make his test. Other good examplesof terracestructures and the relationsof oil to them were shownby Griswold and Munn in JeffersonCo., Ohio. 4

N.M. Fenneman, Bull. 26o, U.S. Geol. Survey, I9o4, p. 437. Science, Vol. 7, P. 563. ,, Geology of Ohio," Vol. 6, p. 94, I888. Bull. 3•8, U.S. Geol. Survey, •9o7. CLASSIFICATION OF PETROLEUM FIELDS. 51.5

'Subclass(e). Broad GeanticlinalFolds.--This is an extreme type of I(a). By a geanticlineis meant an anticline which is extremelylong and broad, and constitutesmore than a local fea- ture, extendingover thousandsor tens of thousandsof square

•XPI-ANATION•o SCALE ONE • 110 • STRUCTURECONTOUR L/NE8 8HOW/Na ELEVAT/ON8 OF TOP OF BEREA 8AND ABOVE T/DE. •IA8WELL WITH LARGE PRODUCT/ON •IA$WELL WITH $f4ALL PRODUCTION DRYHOLE (NO OIL OR•IA$).

FIG. 56. Example of subclassI(d). A gas pool coincident with a struc- -tural terrace, in the Berea sand of southeastern Ohio.

miles. One of the best examplesin this country is the Cincin- nati anticline,in which immensereservoirs of oil and gas have been developedand exhausted,the oil and gas being contained mainly in the Trenton limestone. Owing to the broad areas FREDERICK G. CL.4PP. under which oil and gas are found in the Cincinnati anticlinethe chancesof successin drilling were originally much better than in other fields. The decline of the Trenton limestone fields was largely due to wastefulness'of the gas before peoplein general understoodthat the fields were subject to exhaustion. Most of the poolsin the Clinton sand in Ohio are situatedalong the eastern flank of the Cincinnati anticline, but these pools belong under subclasses(c) and (d), of the classification. Subclass(f). Overturned folds.mExamples of oil and gas occurring in connectionwith overthrust folds are not common but some such casesare conspicuousin California, as shown by Arnold and Johnson2 ClassH. Domes,or QuaquaversalStructures.--Certain types of anticlinesdeveloped as well-markeddomes might be classed here; but sincethey are includedin subclasses(a) and (b) of Class I., the domes here consideredwill be limited to those which are not part of any well-developedanticline, and are thus susceptibleof a different classification. The conspicuousexamples of the occurrencein this countryare in the fieldsof the Coastal Plain of Texas and Louisiana, describedby Hayes,2 Kennedy,2 Fenneman,a Harris 4 and others. Some of the best known in- stancesare Spindletop,Sour Lake and Batson. The moundsin which gas and oil exist may not show as suchat all on the surface. They do, however,appear in some casesas circular ele- vations coveringseveral hundred to severalthousand acres and rising5 ¸ to IOOfeet abovethe surroundingplain, in whichcase they can, of course,be easily recognized. They are frequently known as "salines," on accountof containingdeposits of rock salt. Underneath the surface the whole body of strata have a mound-likeshape, containing, in additionto the ordinary formations, limestone,salt, gypsum,and other minerals,all of which have been pressedupwards in approximatelycircular outlines. • Ralph Arnold and Harry R. Johnson,Bull. 406, U.S. Geol. Survey, •9Ol, p. 97. 2C. W. Hayes and Wm. Kennedy, Bull. 212, U.S. Geol. Survey, •9o3. * N.M. Fenneman, Bull, 282, U.S. Geol. Survey, •9o6. 4G. D. Harris, Geol. Survey of La., Bull. No. 7, •9o8. CLASSIFICATION OF PETROLEUM FIELDS. 517

Fig. 57 showsa crosssection of a typicalsaline according to the ideas of Harris. Class III. Along Sealed Faults. The known examplesof this.classconsist of someof the pools in the Los Angeles field and some of those in the Lompoc field in California described by Arnold.x In these casesthe highly inclined oil sands are cut off abruptlybelow ground by a fault, thus sealingin the oil and gasand preventingtheir escapeto the surface. An example I

FIG. 57. Example of class II. Typical cross section of mound in Texas and Louisiana. (After Harris,, Geol. Survey of La., Bull. No. 7, •9o7.) of thisclass is givenin Fig. 58. Someof theother oil fieldsin California probablybelong in this class. To show the probability that such casesare more frequent than known, it may be worth while to mention that oil-springsfrequently, and perhaps generally,occur along fault lines. Someof theseinstances exist in British Columbia and others in Gaspe,Quebec. ClassI?. Oil and GasSealed in by ./lsphalticDeposits.raCer - • Bull. 309 and 322, U.S. Geol. Survey, 19o7. $I8 FREDERICK G. CL.dPP.

tain examplesof this class,like the last, are few, and not well known,but they may be exemplifiedby the Pitch Lake of Trini- dad, where small quantities of oil and gas are reported. Some of the oil found near the vein of grahamite at Ritchie Mines, West Virginia describedby White,x may belongin this class,although these deposits are alsodependent in their original accumulationupon anticlinal and synclinalstructure, as in Class I. Class V. •It the Contact of Sedimentaryand Crystalline Rocks.--The principal known examplesof this class are in the

Fro. 58. Example of class III. An oil pool in California sealed in by a fault line. (After Arnold, Bull. 309, U.S. Geol. Survey, Plate 20, •9o7.)

Provinces of Quebec and northern Ontario, and have not been much studied. It is probablethat gas occurs at some places in northern New York state in this way. No records are at hand of oil occurringin sucha position,but gas occursin com- mercial quantities. It is held in the zone of the lower Potsdam sandstonewhich is of arkosenature, and restsdirectly upon the underlying granite. The depositsseem, so far as the writer has been able to learn from men who know the fields, to occur on top of prominentknobs of the granite. The occurrencemay be somewhatas in Fig. 59- Class VI. In Joint Cracks.--According to descriptionsof Mr. H. S. Gale, some of the oil in the Florence field in Colorado occursin joint cracks in shale. Other examplesof this nature are believed t• exist elsewhere.

x I. C. White, Bull. Geol. Soc. Am., Vol. •o, •899, pp. 277-284. CLASSIFICATION OF PETROLEUM FIELDS. $19

ClassVII. Surrounding Volcanic Vents.--This classseems to be illustrated by certain Mexican fields. It may be considered an exaggeratedform of ClassII. Although most fields are susceptibleof classificationaccord- ing to the foregoing plan, some of them do not fit in easily. It is believed that if all the causes and effects, and the internal

Fro. 59. Theoretical example of class V. A gas pool in the arkose zone between granite and Potsdam sandstone, in Province of Quebec. character of the sands, were known in every case, the fields couldbe classified. Another factor of importancein the position and distributionof all typesof oil and gas accumulationsis the degree of saturation of the sand; although, as has been explained,x the positionof the water line is not so importantas some personshave supposed. In the Berea grit, for, instances. many poolshave been found, and are still being tapped,where the oil liesbelow the water line in neighboringpools in the same sand. Some of the newer oil developmentshave taught lessons which operatorsand geologiststook yearsto learn, but now that they have been learned, they are being productiveof good results. For instance,in poolswhere the oil occursin thin "pay- streaks," there are sometimesdry areas several square miles in extent, which lie directly "down-dip" from important pools of oil or of water or of both. Near .the corner of Brighton, Chippewaand South Beaver townships,in Beaver County, Pa., • Ecolv. Gv.ox.., Vol. IV.,' I9O9, pp. 565-570. 52o FREDERICK G. CL•IPP.

is a smallpool in whichthe productionof the wells is very light, which is bordered on the northeast, southeastand south by a numberof strongsalt water wells. The "pay" here,as in the majority of poolsin the Bereasand, is near the top of the sand. The salt water in closeproximity to the pool occupiesa definitestructural position with referenceto it, occurringon all sidesof the poolexcept on the up-dipside. However,this water was not found to extend'far down-dipfrom the oil-pool,several wells between there and the Ohio River, four miles southeast, having been dry of both water and oil. Similar occurrences are found repeatedlyin the Berea sandin BeaverCounty, Pa., and in neighboringcounties and the Pan-Handleof 'WestVir- ginia and in Ohio, where dry areas of the sand occur on the down-dipsides of severalpools of oil and salt water. The point to be noted is that the lowest oil-pool in a stated group of poolsin the Berea sand is successivelyhigher above sea-levelgoing from southwestto northeast parallel with the Ohio River. Near Steubenvilleand Mingo, in JeffersonCounty, O., the lowermostelevation at which oil has as yet been dis- coveredin this sand is 60o feet below sea-level; at Holidays Cove, W. Va., it is about 4oo feet below sea-level; on the West Virginia-Pennsylvania line near the southwestern corner of Beaver County it is 25o feet below; at Hookstown and Smiths Ferry, Beaver County, it is at sea-level;while near the corner of Brighton, Chippewa and South Beaver townships, in the pool first mentioned,it is 2oo feet above sea-level. The pro- gressiverise toward the northeastis explainedin part by the greater general height of the anticlinal folds and the lesserdepth of the adjacent synclinal depressionsin the sand in that direction. The fact that a so-called"pay-streak" lying down-dip from a known pool may be dry of both water and oil doesnot disprove the "structural theory" of the accumulationin its proper application; but the position of the dry area is due to the fact that whateverwater and oil may oncehave occupiedthe "pay-streak" have been drained out, segregatedelsewhere naturally or arti- ficially, or that the "pay-streak" is in reality not so porous as CLASSIFICATION OF PETROLEUM FIELDS. 5:2I would seem from the size of the grains in the sand-pumpings. Moreover, many apparentlydry areas are due to a changein the characterof thesand, by v•hichit becomespractically non-oil- bearing.or non-water-bearing. The fields in the deepersands frequently do not correspond structurally with those in the shallow sands, even where the sandsare parallel, and this is one reasonwhy the deeperfields are not discoveredmore rapidly. As a rule, the shallow sands are productiveof gas near the crestsof antictines,where anticlines occur, as in ClassesI(a) and I(b), and are frequently confinedto domes,in thesecrests, while the deepersands may be, and frequentlyare, barrenon the crestsand productiveof gas on the flanksof the anticlinewhere it is dippingmoderately steeply towards the adjacent syncline. This is generally due to the non-parallelismof the deepersands. And sincethe size of pools is approximatelyinversely proportional to the steepnessof the dips,these deep-sand pools commonly take the longestto be dis- coveredby the drill.