HHHra Hi KB am nfiHBmSl m 111111$ HIP BMnnn ■ ■nRmg ■•■•-' StSff Km Bfl HHHHI3 : - " HHHaHanO EsxnNSB HBMmiiElll IBhIII BHBB11 • ■ ■'■,.■' ■ URBANA STATE OF ILLINOIS DEPARTMENT OF REGISTRATION AND EDUCATION A. M. SHELTON. Director DIVISION OF THE STATE GEOLOGICAL SURVEY M. M. LEIGHTON. Chief REPORT OF INVESTIGATIONS-NO. 12 LIMESTONE FOR SEWAGE FILTER BEDS CAUSES OF DISINTEGRATION, DESIRABLE PROPERTIES, AND METHODS OF TESTING BY J. E. LAMAR PRINTED BY AUTHORITY OF THE STATE OF ILLINOIS URBANA, ILLINOIS 1927 STATE OF ILLINOIS DEPARTMENT OF REGISTRATION AND EDUCATION A. M. SHELTON. Director DIVISION OF THE STATE GEOLOGICAL SURVEY M. M. LEIGHTON. Chwf Committee of the Board of Natural Resources and Conservation A. M. Sheltox, Chairman Director of Registration and Education Charles M. Thompson Representing the President of the Uni versity of Illinois Edson S. Bastin Geologist Schnepp & Barnes, Piunteiss Springfield, III. I !i 2 7 65102 ■:: LIMESTONE FOR SEWAGE FILTER BEDS CAUSES OF DISINTEGRATION, DESIRABLE PROPERTIES, AND METHODS OF TESTING By J. E. Lamar OUTLINE PACiE Introduction 5 General statement 5 Acknowledgments 5 Causes of disintegration of filter stone 6 Mechanical disintegration 6 Freezing and thawing 6 Heating and cooling 7 Physical factors influencing the rate of mechanical disintegration.... 7 Porosity 7 Texture 7 Bedded or laminated structure 8 Chemical disintegration 8 Solution : 8 Oxidation 9 Hydration 10 Cooperative effect of mechanical and chemical disintegration 10 Properties desirable in filter stone 10 Testing of limestone and dolomite filter stone 11 Tests for hardness, toughness, and wear 11 Accelerated soundness test 13 Freezing test 15 Water absorption test 15 Microscopic examination 15 Solution tests 15 Etching 15 Rate of solution 16 Studies of residue 16 Chemical analysis 16 ILLUSTRATIONS FIGURE PAGE 1-5. The effect on different types of limestones of etching for the same time period and with the same acid concentration. 1. Niagaran dolomite IS 2. Ste. Genevieve oolitic limestone 19 3. Chester limestone 20 4. Chester limestone 21 5. Pennsylvanian limestone 21 TABLES PAGE 1. Variations in physical tests of Niagaran dolomite 12 2. Hypothetical accelerated soundness tests 14 Digitized by the Internet Archive in 2012 with funding from University of Illinois Urbana-Champaign http://archive.org/details/limestoneforsewa55712lama INTRODUCTION General Statement The matter of predetermining the suitability of a given limestone for use in sewage filter beds is one which is yearly coming to be of more interest to quarrymen and engineers of sanitation. This interest is linked with the growth of towns and cities, the consequent increase in the amount of sewage, and the endangering of public health by improper sewage disposal. As a result, the use of sprinkling and trickling filters has come to be of increas- ing interest. In a filter recently constructed for a town of 25, 000 people, 6S0 carloads or about 13 trains of limestone were used. It is apparent, therefore, that the initial cost of such a quantity of stone is an important item to a municipality, and that an enduring stone is desirable not only because of initial cost but because of the cost of removing poor stone and replacing it with new material. The problem of sewage disposal during the time required for replacement of the filter stone is still another factor making this operation one to be assiduously avoided. The study of limestones to determine their suitability as filter stone is new and still in the experimental sta.^e. As yet, very few data are avail- able which correlate laboratory tests and the actual life of the stone in the filter bed. The collection of such data should be the ultimate aim of research and investigation in this line. However, the processes operating to destroy the limestone and their action on the stone may be evaluated, and such tests as are at hand for determining the resistance of the stone to destructive agencies may be applied. Although the only real test of a filter stone is its behavior in the filter bed, the results of the laboratory tests give, at the present time, the most intelligent basis for judging the comparative merits of proposed filter stones. The function of filter stone in a sewage filter is essentially that of a lodging place for bacteria which gather and grow upon the surface of the stone, and by their life processes effect a purification of the sewage. Acknowledgments The author wishes to acknowledge with thanks the helpful criticisms and suggestions bearing on this paper, made by Dr. A. M. Ruswell, Chief, Illinois State Water Survey, and Dr. Ah M. Leighton, Chief, Illinois State Geological Survey. b LIMESTONE FOB SEWAGE FILTEB BEDS CAUSES OF DISINTEGRATION OF FILTER STONE The principal processes effecting the destruction of filter stone in sprinkling and trickling filters are mechanical disintegration and chemical disintegration. These two processes though they are doubtless greatly cooperative in their action, will be discussed separately, and their combined activity will be taken up in a subsequent statement. Although their destruc- tive effect is slow, — not to be measured in single years — yet ultimately and in the aggregate these processes do produce a greater or lesser destruction of limestone filter media. Mechanical Disintegration freezing and thawing The most important process producing mechanical disintegration i- freezing and thawing. The destructive effect of this process depends on the number of times it is repeated which in turn is related to the climate and to the operation of the filter bed. During the winter months filters may be operated continuously, or only during the daytime, to treat the more con- centrated sewage, or, if the dilution is sufficient, the sewage may be In- passed entirely. In continuous operation in a moderate climate a certain amount of ice forms usually in wide circles around the nozzles of the sprinkler and at the surface of the filter bed. The temperature of the sewage and the biological action are usually sufficient to prevent frost from entering to any considerable depth. Under severe conditions, such as a sudden cold wave accompanied by wind, the entire bed may become covered with ice. Under these condition^ a given portion of the bed may freeze and thaw four or five times during the winter. In filters in which only the strong day sewage is treated and the nighl sewage is by-passed, the exposed portion of the bed may freeze at night and thaw again when treated with the warm sewage so that freezing and thaw- ing under certain weather conditions might occur daily, thus affecting tin filter stone more severely. Filters which are not operated during the winter months, should ft shut down in time to avoid the earliest probable freezing temperature, so that they may drain completely and dry out. Freezing and thawing disrupts a stone principally by reason of the fad that when water changes to ice it expands about 1/10 its volume. There- fore, if a stone is saturated with water, and the water freezes, the stone will be disrupted unless it has sufficient strength to overcome the force of the expansion accompanying the change of water to ice in the sub-surface pores. In the surface pores the expansion accompanying freezing is prob- LIMESTONE FOR SEWAGE FILTER BEDS 7 iblv partly accommodated in the open direction of the pore, but not always sufficiently to eliminate disruptive stresses completely. Though a stone may withstand the forces occasioned during the freezing of water in its pores nid in pits or pockets on its surface for a number of repetitions of the pro- :ess, it eventually gives way and fractures or chips. Inasmuch as there is inherently less resistance to breakage in the surface portions of a stone, other things being equal, the common effect of freezing and thawing is a chipping or scaling off of the surface. HEATING AND COOLING Although not nearly so important as freezing and thawing, the expan- sion and contraction of the stone in a filter bed, as the result of heating by the sun's rays in the summer and subsequent cooling by dousing with sew- age, is doubtless a contributory factor in the destruction of filter stone. The : effect on the stone would probably be to cause it to chip or spall. PHYSICAL FACTORS INFLUENCING THE RATE OF MECHANICAL DISINTEGRATION POROSITY The pores of a stone which can be penetrated by a liquid must have a connection with the surface of the rock and therefore the amount of water a stone can absorb is related to the volume of the connected pore space it possesses. This porosity is an important factor governing the effectiveness of freezing and thawing in destroying the stone. The distribution and size of the connected pores is also important in this regard. For example two stones may have the same porosity, the first with evenly distributed small pores, and the second with but a few large pores concentrated in a given zone ; the second would fracture more quickly than the first, other things being equal, because of the concentration and localization of the stresses set up in the few larger pores of the stone by repeated freezing and thawing. TEXTURE Limestones in general may be grouped into three main classes accord- ing to their texture as apparent to the naked eye, namely, crystalline, non- crystalline or dense, and granular. There are some limestone formations which fall definitely into one of these three classes, but in general limestones have such a variety of texture that any two or even all of the above classes may be embodied in one formation. The terms are of value, however, in expressing the relative importance of the three textural divisions for pur- poses of discussion. There is a definite relation between the physical struc- ture of filter stone and its resistance to disruption, for the strength of a crystalline limestone depends on the degree of interlocking of the crystals or, if the stone is granular, on the completeness of the bonding of the granular 8 LIMESTONE FOB SEWAGE I'll.TEH BEDS material. A stone which appears crystalline to the eye is likely to have well- interlocked grains. A stone which appears non-crystalline may or may not have well-interlocked crystals. All limestones when examined microscopic- ally under sufficiently high magnification are seen to he composed, to greater or lesser extent, of crystals of calcite ; if the stone is a dolomite, the crystals are calcite and dolomite. Therefore, some limestones, non-crystal- line to the eye, are simply so line grained that the individual crystals are nol megascopically recognizable. In many limestones appearing to the either crystalline or non-crystalline, though more frequently the latter. cla\ is present in varying amounts. This clay, in general, lessens the strength of the stone because it interrupts the interlocking of grains or coats them at the planes of contact. In some rocks the clay is segregated in zones or thin hands through the stone which often constitute planes of weakness. In the case of siliceous limestones, crystals of silica, either localized or evenly dis- seminated through the stone, will be found interlocking with or replacing calcite crystals. In granular limestones composed of fossils, fossil debris, oolite grains, or calcareous debris of any sort, the strength of the stone depends to a largi degree on the completeness of the bond effected between the granular ma- terial and the matrix and on the strength of the bonding substance. In main limestones the bond is very good; in others a thin film of organic material surrounds many of the granules or fossils and consequently interrupts the continuity of the bond. BEDDED OK LAMINATED STBUCTUBE Some limestones possess a distinct bedded or laminated structure which exerts an important influence on the ability of the stone to resist disintegra (ion. The bedded structure is not the gross feature of bedding hut rather minor, abrupt variations in the texture of the granular or detrital materials composing the stone within a larger bed. The laminated structure is not generally visible to the eye in fresh specimens, hut is usually revealed during the accelerated soundness test by a flaking or splitting of the stone into thin sheets or laminae. Inasmuch as the contacts of texturally unlike sedimentarj materials are often planes of weak, partly or wholly interrupted hond, a stone having a bedded or laminated structure should he thoroughly tested before it is accepted as filter stone. Chemical 1 )isintegration solution As long as a sprinkling filter is in operation, all of the stone in the filter bed is being repeatedly doused with the effluent from the settling tanks. In ordinary domestic sewage this effluent is largely water, hut it contains car- LIMESTONE FOR SEWAGE KILTKK BEDS 9 bonic acid and possibly a number of fatty acids. Although both are in very dilute solution they doubtless dissolve a relatively small but still appreciable portion of the filter limestone in a period of years. Furthermore, the bac- teria growing on the stone produce carbon dioxide which forms carbonic acid. There is some question, however, as to whether the net effect of the coating of bacterial jelly is not more protective than injurious as regards solution of the filter stone. Considering the effect of solution on limestones and dolomites, it is noteworthy that dolomites, though more soluble than limestones, are less rapidly soluble. The rate of solution is of importance because the move- ment of the water through the filter bed defeats any chance of saturation of the dissolving medium, thus nullifying the effect of differences in solubility, and facilitates the solution of the more rapidly soluble substance. The rate nf solution of a stone, however, is influenced by an additional factor, namely, the area of surface exposed to the solvent. The larger the area the greater the potential rate of solution. In general, dolomite, as a result of the manner of its formation, is more porous than limestone, and consequently exposes a greater surface to the solvent than limestone. This tends to and may prac- tically equalize in the net result the difference between the two rocks in rates of solution. If a piece of limestone or dolomite with a plane surface is immersed in hydrochloric acid for a brief time, it will be found that certain parts of the stone have dissolved more rapidly than others (figs. 1-5, pp. 18-21 ). In some stones, the fossils, fossil debris, vein fillings, and the like dissolve the more rapidly and in others, the matrix. In still others the effect of the solution is about equal over the area exposed. If the stone is very pure, solution will leave a surface nearly similar to the original one ; if , however, the stone is an argillaceous or siliceous limestone or dolomite, a protective residual coat will remain. If the residual coating is firm and hard, it is an asset in hindering further solution. The effect of selective solution on any given portion of the stone is to increase the surface porosity. As a result additional surface is made available to the action of the solvent, and the opportunities for me- chanical disintegration are increased. The matter of selective solution is thought to be the crux of the potential destructive effects of solution on filter stone. A stone of even texture and uniform solubility would apparentlv be desirable to offset the effects of solution. OXIDATION Because of the frequent wettings and complete or partial dryings to which filter stone is subjected, the process of oxidation is greatly facilitated. Substances which oxidize are, therefore, to be consistently avoided. Prob- ably the worst offenders in this class are pyrite and marcasite, both sul- 10 LIMESTONE FOB SEWAGE FILTEB BEDS phides of iron. These oxidize to linn mite, the hydrated oxide of iron. which occupies more volume than either of the sulphides. As a consequence oxidation of either of the sulphides to limonite is accompanied by forces tending to disrupt or weaken a stone if they cannot he accommodated by free outward expansion as in surface pores. HYDRATION It is possible that some limestones contain clay minerals along bedding planes, or less probably in a disseminated state, which absorb water as a result of the constant wetting to which they are subjected, with an accom- panying increase in volume. This increase would have the same general effect as that resulting from the change of pyrite into limonite. Cooperative Effect of Mechanical and Chemical Disixtegratiox As previously stated, it is very difficult to divorce and identify singly the effects of mechanical and chemical disintegration, respectively, on a filter stone. Mechanical disintegration increases the surface area available for the agents of chemical disintegration. Chemical disintegration, in turn, if it is selective, results in an increase of porosity and favors further mechan- ical disintegration. Without doubt the upper foot of the filter bed is the zone of maximum effect of combined mechanical and chemical disintegra- tion, ft seems likely, also, that the basal two or three feet of a filter bed are the most favorable for chemical disintegration not excepting the process of oxidation. PROPERTIES DESIRABLE IN FILTER STOXE From the foregoing discussion of the agencies causing the destruction of filter stone, the following properties seem desirable in filter stone : (1) The stone should have a minimum volume of pore space connected with the surf-ace. (2) The pores of the stone should be small and evenly distributed. (3) The stone should consist of well-interlocked crystals or if it is granular, the grains should be firmly bonded by a strong cement. (4) The stone should be of uniform solubility. (5) The stone should be free from minerals which oxidize or hydrate. Pyrite and marcasite especially are to be avoided. (6) The stone should have a sufficiently rough surface to furnish anchorage for the bacteria which are to grow upon it. (7) The stone should be comparatively pure chemically. Stones with high clay contents are generally to be avoided. A high siliceous content is probably not harmful if the silica occurs in fine crystals evenly disseminated. (8) The stone as delivered to the filtering plant should be free from dirt or fine rock particles which might collect in and clog the basal portion of the filter bed. LIMESTONE FOE SEWAGE FILTER liEDS 11 TESTING OF LIMESTONE AND DOLOMITE FILTER STONE Certain tests may be made on limestones and dolomites which individu- ally indicate the relative value of specific properties of a stone. It is impos- sible, however, exactly to duplicate conditions in nature and at the same time accelerate in a laboratory the combined effects of mechanical and chem- ical disintegration. Therefore, the results of any set of tests are indicative rather than absolute. Nevertheless, as such they serve a purpose. The following series of tests, though doubtless not in their ultimate form, appear to give valuable data concerning filter stones and will serve, as well, as a basis for devising improved tests. Tests for Hardness, Toughness and Wear It is natural in developing a new set of tests for a certain material, to turn to known tests and attempt to adapt them to the work at hand. In an endeavor to find adequate tests for filter stone, engineers have turned to the tests made on limestone highway aggregates, and accordingly have em- ployed the toughness test, the Dorry hardness test, the Deval abrasion test and the accelerated soundness test. Considering the first three tests as a whole, they indicate the resistance of rock to repeated impact and to wear and abrasion. Inasmuch as it does not seem probable that impact, wear, or abrasion plays any important part in the destruction of filter stone, it would appear that these tests might be dispensed with. However, since the results of the hardness, toughness and wear tests and the percentage of water absorbed are in most cases the only data available from the numerous tests of various stones for aggregate, these data may give clues as to range and character of the regional variations in a given limestone formation. For example, Table 1 gives data on the variation in the physical properties of the Xiagaran dolomite in the three principal areas in which it is quarried in Illinois. From the table it appears that the stone from district No. 1 is likely to be more variable than the stone from districts Nos. 2 and 3 and consequently stone from that district should be more carefully and fre- quently sampled than stone from the other two districts. Regarding the physical tests shown in Table 1, only the water absorp- tion, discussed later, is individually significant. Although it may be said that rocks satisfactorily passing the above mentioned tests from the stand- point of highway material will probably pass the accelerated soundness test. there are known to be exceptions which make it highly desirable that the soundness test itself be applied. This test, if applied, practically eliminates the need of hardness, toughness and wear tests for filter stone. 12 LIMESTONE FOB SEWAGE FILTER BEDS co to 2, O Varia- tion CO CO c- a S in *v -r X cci a M o rH iH r-t i-H to 0) s •- 8 Varia- tion cm co c- in CO *r C5 © CO N C6 ri H i-l cci a HMO oo in co* H H ri CD 'u G CD o u o •~ Varia- tion 8.6 4.2 4.9 d a OS CO C} CO in" in X efl a in oq oo cm' o o i-H H i-l -a . O 3 to w Ctf CD s- a CD *-> co CIS ,C .2 e i- o > n en |{ SEWAGE FILTER UEDS Figures 1-5 show the efifect on different types of limestones of etching tor the same time period and with the same acid concentration. Fig. 1. Niagaran dolomite. Maximum surface relief about 1 mm. The etching has enlarged the surface pores and has left the specimen coated with fine dolo- mite crystals which may be easily brushed off. The effect of surface etching or solution of the calcium carbonate from dolomitic limestone is augmented by the consequent freeing of dolomite grains; hence the actual weight of dolomite dis- integrated may in some cases be equal to or greater than the weight of lime- stone dissolved by simple solution under identical conditions. This specimen chipped slightly after 8 repetitions of the soundness test, but thereafter remained intact through the remainder of 20 repetitions. ( Magnification six times.) LIMESTONE l'OU SEWAGE FII.TKK I1EDS 19 Fig. 2. Ste. Genevieve oolitic limestone. Surface relief about 0.2 mm. The etching has affected the stone very evenly but was slightly more pronounced on the white outer portions of the oolite grains. Near the center of the right margin of the illustration, a hexagonal quartz grain which is the center of an oolite grain projects above the general level of the surface and casts a pronounced shadow. That this specimen showed no disintegration after 20 repetitions of the soundness test indicates a good bonding of the granular material of the limestone. (Magnifi- cation six times.) 20 LIMESTONE FOK SEWAGE FILTER UEDS Fig. 3. Chester limestone. Maximum surface relief about 1 mm. This is a very pure limestone composed of fossil debris bonded by calcium carbonate. Etching has dissolved the calcium carbonate and Left the fossil debris projecting from the general level of the surface. At the right of the picture is a brachiopod which has been filled with calcite. After 16 repetitions of the soundness test this speci- men broke into three pieses along a large fossil. There was no breakage during the four subsequent repetitions of the test. The resistance of this specimen to the soundness test is evidence that in general it is well bonded. (Magnification six times.) LIMESTONE FOIS SEWAGE FILTKH BEDS 21 Fm. 4. Chester limestone. Maximum surface relief about 1 mm. Etching has removed the calcium carbonate cement and left the clayey, siliceous impurities of the stone as projections above the general surface. Disintegration of this speci- men began after 4 repetitions of the soundness test, and complete failure resulted ' after 8 repetitions. The disintegration of the stone into a coarse sand showed that the bonding material was weak. (Magnification six times.) %m i Pig. 5. Pennsylvanian limestone. Maximum surface relief about l 1 ^ mm. This is a siliceous limestone containing fossils composed of calcium carbonate. These were dissolved during the etching and gave rise to the linear depressions. After 6 repetitions of the soundness test this specimen began to chip and split along the surfaces of fossils. After 11 repetitions, disintegration of the entire specimen was complete. This sample shows the type of results which may be expected in limestone with a localized porosity. (Magnification six times.) CHliR'S" LIBRARY BINDERS 507 S. Goodwin Urbano,IlL