ApPENDIX Petrographic Analysis of Sandstones Introduction effectiveness, however, depends on the imagination and skill of the operator. The petrographic analysis of sediments, sand­ stones in particular, became an organized disci­ pline with the development of thin-section tech­ Rock Description and Analysis niques and the polarizing microscope-a devel­ opment attributed to Henry Clifton Sorby. Rock description and analysis are based on Sorby was making thin sections in 1849, pub­ study of outcrops, cores, hand specimens, and lished a paper on the microscopical structure of thin sections. Thin sections can be easily pre­ "calcareous grit" in 1851, and his paper in 1880 pared for unconsolidated sands also (Middleton on the non-calcareous stratified rocks was a ma­ and Kraus, 1980). Emphasis here is placed on jor milestone in the thin-section analysis of hand-specimen and thin-section study of either sandstones. Folk (1965) has ably summarized core or outcrop samples. Sorby's petrographic contributions as has Sum­ The art of rock description and analysis is merson (1978). Until shortly before World War learned by doing and by the study of published II, most students of sedimentary rocks, unlike examples. Good descriptions of rocks of all those of igneous and metamorphic rocks, failed types have been published in Bulletin 150 of the to "follow through" on Sorby's auspicious be­ U.S. Geological Survey. A number of abbrevi­ ginnings. One brilliant exception was Lucien ated descriptions are given in Grout (I932, pp. Cayeux (1929). In more recent years, however, 22-28). See also the approach to rock descrip­ the thin-section analysis of sedimentary rocks tions of Ferm and Weisenftuh (1981), an ap­ has become commonplace and is now a fully proach that uses colored pictures of rock types exploited tool for research. with a code name and number. The prime object of the study of a thin section The use of a well-designed petrographic form is, or should be, the reading of rock history. The develops a regular pattern of description and microscope is the most useful, general method maximizes the efficiency and effectiveness of for close study of the mineral composition, fab­ any microscopic study. The level of descrip­ ric, and general makeup of a rock. Such close tion, however, may vary widely-from a con­ study is a necessary complement to field studies cise paragraph based largely on qualitative ob­ in interpreting the origin of sands and sand­ servation and semiquantitative visual estimates stones. to a three- or four-page typed report based on The study of sand and sandstone in the labo­ counts of 200 to 500 or more grains; or in place ratory has proceeded in other directions also, of a written report, the data may be directly and there has been a proliferation of methods entered on a computer disk for later processing. for study of grain size, grain shape and round­ Choice between semiquantitative and quanti­ ness, porosity and permeability, and the like. tative estimates depends on the investigator's Many of these methods, however, are applica­ objectives, his judgment, and the available ble only to unconsolidated deposits and, useful time. If very large numbers of samples are to be as they may be, the thin-section approach re­ studied and little time is available for the task, mains the single most effective means of inves­ semiquantitative estimates must suffice. Min­ tigation of sandstones in the laboratory. Its eraI percentages can be estimated by compari- 520 Appendix: Petrographic Analysis of Sandstones 1.00 0.50 2.00 DIAMETER RATIO PHI STANDARD VERBAL (MIlliMETERS) DEVIATION SCALE lO -~----~ 000 very well sorted 16 0 35 --~~--- MATURE well sorted f 20 050 -----~-----~ moderately sorted t ,::~-~-=::-~ ---~ J__ '"~ (After Folk, 1965, p. 104-105) son charts with a reticle (Terry and Chilingar, FIGURE A-I. Sorting images from Harrell (1984. 1955). Sorting can also be estimated by compar­ Figs. 3, 4, 5, and 6) and sorting classes from Folk ison charts (Fig. A-I). Roundness of individual (1968, p. 102). Published by permission of the au­ grains is always so estimated (Fig. A-2). Counts thors and the Society of Economic Paleontologists of 50 to 100 grains have usually been made to and Mineralogists. estimate either average grain roundness or per­ centage of angular grains. But the earmark of the modern petrographer estimate based on a given number of counts. is the point count of 200 to 500 grains per slide Van der Plas and Tobi (1965) provide a compa­ for estimates of composition. Quantitative esti­ rable chart (Fig. A-3). Dennison (1962) also mates are needed for many petrographic classifi­ gives useful charts, all of which are based on cations and for most subsequent statistical anal­ the normal and binomial distributions. In prac­ ysis. By using binomial and Poisson confidence tice, it is not uncommon for a petrographer to charts (Pearson and Hartley, 1954, Tables 40 utilize semiquantitative estimates for some of and 41), one can determine the reliability of an the petrographic variables and reserve system- ~ o (J :>':"" ti (1) (JC/O ::1. -g o· ::3 ~ 0-::3 ~ ::3 po .:z[J; (ji. y y I L-------"II III III II~ II Very Sub- 3 Sub- 4 5 Well- 6 o angular Angular 2 angular rounded Rounded rounded FIGURE A-2. Roundness images and classes. Columns show grains of similar roundness but different sphericity. (Redrawn from Powers, \953, Fig. I). t"'J> Appendix: Petrographic Analysis of Sandstones 5000,--,~~~~~~~~~-,~~~~-' 4000 3000 - 1500 600 500 400 - n 300 200f ! ISO! 60 50 40 ----1-- I 40 50 70 80 90 100 P atic point counting for the one or two most sig­ FIGURE A-3. Ninety-five percent confidence limits nificant ones. An automatic electronic stage and for mineral proportions, where 11 is total number of counter is both a convenience and an efficient grains counted and p is the estimated proportion of a time saver and back scatter election image anal­ particular mineral. Curved contours in percent give confidence limits. Worked example: 11 is 500. p is 28 ysis offers promise of automated point counts percent. and the confidence limit is 4 percent so that (Dilko and Graham, 1985). in repeated sampling the true proportion will lie Sorting, angularity. and clay content define within 24 and 32 percent. (Modified from van der textural maturity, which should be specified. Plas and Tobi. 1965. Fig. 1). Table A-I gives a flow chart for this procedure. Sorting can be estimated by comparison with Fig. A-lor one can estimate it by determining items in these report forms will not be appropri­ the ratio of two representative grain diameters: ate and that there may be special comments that the diameter in millimeters of a grain of the 84th we have not included. The detailed form em­ percentile and the diameter of a grain of the phasizes, however, the comprehensive nature 16th percentile. Conversion to phi units. sub­ of a full description. The analysis of the Trivoli traction, and division by two yields the sorting. Sandstone presented below follows this form. Clay of authigenic origin should be ignored, We have included a few amplifying remarks when clay content is determined. for the most effective use of these forms. After To facilitate petrographic analysis, we have looking at the hand specimen with the naked included two petrographic forms. Table A-2 is a eye. and hand lens. or binocular, it is best to skeletal form and Table A-3 a very detailed one scan the thin section with low power to appraise modified from Folk (1968. pp. 133-38). We rec­ its general characteristics. This should be fol­ ognize, of course, that in many studies all the lowed by a grain size analysis of its terrigenous Rock Description and Analysis 523 TABLE A-I. Textural maturity flow chart. (Modified from Folk. 1968, p. 102). STEP 1 Clay content (micaceous material less than 30 JJ., excluding authigenic material) a) If greater than 5 percent, sand is immature. b) If less than 5 percent, determine sorting. STEP 2 Sorting (See Fig. A-I) a) If sorting is greater than 0.5~ (diameter ratio over 2.0), sand is submature. b) If sorting is less than 0.5~ determine roundness. STEP 3 Roundness (See Fig. A-2) a) If sand-size grains are subangular to angular (3.0 or less on the Powers scale, Table 3-3), sand is mature. b) If roundness exceeds 3.0, sand is supermature. TABLE A-2. Short petrographic report for sandstones. HAND SPECIMEN Color, grain size, sorting, induration, bedding, etc. and field name. THIN-SECTION DESCRIPTION Abstract: Digest and condense all the petrography and summarize in 25 words. Texture: Modal size, sorting, and nature of grain-to-grain contacts. Bedding. Percent sand, silt, and clay. Roundness. Mineralogy: Give percent of terrigenous, orthochemical (precipitated cement) and allochemical (transported grains formed within the basin of deposition) material plus description and amounts of different types of terrigenous debris. A brief paragraph for each constituent. Interpretation: Character of source area plus type of transportation and character of depositional basin, if possi­ ble. Diagenesis. GENERAL COMMENT Always keep description and interpretation separate. At times you may estimate abundances with 100 point counts or by using comparison charts. components under medium or high power. Size tions of one mineral to another are most impor­ analysis is commonly the best way to become tant and should be recorded. acquainted with both the texture and mineral­ Any comprehensive study of thin sections ogy of the section. Counts of as low as 50 grains should be supplemented by X-ray analysis of to 100 grains are satisfactory for the mean and interstitial matrix, (Wilson and Clark, 1978) and sorting for many problems, or images can be microscopic heavy mineral analysis.