Twelfth Forum on the Geology of Industrial Minerals, April 22-24, 1976, Atlanta, Georgia

TWELFTH FORUM ON THE GEOLOGY OF INDUSTRIAL MINERALS
APRIL 22-24, 1976 ATLANTA, GEORGIA
STATE OF GEORGIA DEPARTMENT OF NATURAL RESOURCES
Joe D. Tanner, Commissioner
ENVIRONMENTAL PROTECTION DIVISION
J. Leonard Ledbetter, Director GEORGIA GEOLOGIC SURVEY
ATLANTA 1978
49 INFORMATION CIRCULAR

For convenience in selecting our reports from your bookshelves, they will be color-keyed across the spine by subject as follows:

Red Dk. Purple
Maroon Lt. Greer1 Lt. Blue Ok. Green Ok. Blue Olive
Yellow
Dk. Orange Brown Black

Valley & Ridge mapping and structural geology Piedmont & Blue Radge mappang and struc
tural geology Coastal Plain mapping and stratigraphy Pa leontology Coastal Zone studies Geochemical and Geophysical studies Hydrolo.gy Economic geology Mining directory Environmental studies Engineering studies Bibliographies and lists of publications Petroleum and natural gas Field trip guidebooks

Colors have been selected at random. and will be aug men ted as new subjects are published

TWELFTH FORUM ON THE GEOLOGY OF INDUSTRIAL MINERALS
APRIL 22-24, 1976 ATLANTA, GEORGIA
STATE OF GEORGIA DEPARTMENT OF NATURAL RESOURCES
Joe D. Tanner, Commissioner ENVIRONMENTAL PROTECTION DIVISION
J. Leonard Ledbett~r. Director GEORGIA GEOLOGIC SURVEY
ATLANTA 1978
49 INFORMATION CIRCULAR

CONTENTS

Keynote Address for the Forum on the Geology of Industrial Minerals, Atlanta, Georgia, April 22, 1976. Philosophical Test-Pitting or Thinking While We Dig by DONALDA. BROBST . . . . . .

Table 1. Mineral materials in the United States economy

Table 2.

Major mines in the United States producing industrial minerals . . . . . . . .

Table 3.

Production, value and rank of 20 major industrial minerals . . . . . . . .

Table 4.

World and United States barite production and United States imports, 1850-1975 . . .

The Crushed Granite Industry of the Atlanta Metropolitan Area by ROBERT L. ATKINS . . . . . . . . . . . . . Figure 1. Physiographic provinces of Georgia . Figure 2. Quarries in the Atlanta Area Figure 3. Example of a portable crushing plant

The Origin of Georgia's Kaolin Deposits by ROGERS. AUSTIN

Figure 1. Kaolin and bauxite areas of the southeastern United States . . . . . . . . .

Figure 2. Pen sketches of selected quartz grains

Figure 3.

Sketch of typical kaolin mine Washington County, Georgia . . .

Figure 4.

Sketch of kaolin mine high wall showing relationship between kaolin and bauxite . . . . . . . . . .

Figure 5.

Sketch of burrow at the Cretaceous-Tertiary unconformity in Twiggs County, Georgia . . . . . . . . . . . . .

Figure 6. Sketch of kaolin mine high wall in Twiggs County Georgia

Figure 7. Generalized stratigraphic section in the middle Georgia kaolin region . . . . . . . . . . . . . . . .

Alumina from Domestic Resources by DON H. BAKER . . . . . .

Figure 1. Generalized acid-leaching-process flowsheet

Table 1. Petrographic grain count . . . . . . . .

Table 2. Chemical analysis of dry clay . . . . . .

Table 3.

Cumulative weight-percent of material finer than indicated size . . . . . .

Table 4. Chemical analysis of calcined clay . . . .

Table 5. Theoretical energy data, Nitric Acid miniplant

Table 6. Theoretical energy data, Hydrochloric Acid mini plant.

Table7.

Theoretical energy data for Evaporation Crystallization option . . . . . . . . . . . . . .

Table 8. Theoretical energy data for Bayer Process . . . . .

Page
2
3
4
6 6 7 9
10 11 12 12 13 13 14
15
16 17 18 18 18 19 20 21 22 23

CONTENTS- Continued

Geologic Classification and Evaluation of Heavy Mineral Deposits by THOMAS E. GARNER, Jr. . . . . . . . . . . . . .

Table 1. Magnetic opaque minerals . . . . . .

Figure 1.

Idealized depositional sequence for titanium mineral deposits . . . . . . . . . . .

Figure 2. Electronmicrograph showing ilmenite grain from continental deposit . . . . . . . .

Figure 3. Photomicrograph of ilmenite grains from marine shoreline-dunal deposit . . . . . . . . . .

Figure 4. Truncated dune bedding in marine shore -dunal deposit

Figure 5.

General outline for laboratory evaluation of heavy mineral prospect samples . . . . . . . . .

Figure 6.

Outline for laboratory determination of mineral separation characteristics. . . . . . . . .

Table 2. Examples of heavy mineral value calculations .

Figure 7.

Photomicrograph showing ilmenite grains from marine shoreline -dunal deposit . . . . . .

Figure 8 . Photomicrograph showing ilmenite grains from
continental deposit . . . . . . . . . . .

Reconnaissance Investigations of Offshore Phosphate Deposits of Georgia and South Carolina by JAMES L. HARDING and JOHN E. NOAKES Figure 1. Gamma -ray detection sled . . . . . . . . Figure 2. Static underwater gamma detection system . . Figure 3. Original traverse lines in Georgia coastal waters Figure 4 . Radiometric readings recorded by shipboard electronics from towable underwater sled . . . . . . . . . . Figure 5. Sub-bottom profiler record . . . . . . . . . . . Figure 6. Final grid pattern run of Georgia - South Carolina coast Figure 7. Diagrammatic cross-section . . . Figure 8. Diagrammatic fence diagram Table 1. Phosphate analysis of grab samples

Development of High Extraction Magnetic Filtration by the Kaolin Industry of Georgia by JOSEPH IANNICELLI . Figure 1. Frantz Ferrofilter Figure 2. Jones Carousel Separator . Figure 3. Hybrid Separator . . Figure 4. Short Coil Separator Figure 5. Long Coil Separator. Figure 6. 84-1 nch Magnetic Separator . Figure 7. HEMF treatment of high quality Georgia # 2 kaolin

ii

Page
25 26
28
29
29 30
31
33 34
35
35
37 38 38
39
39 40 40
41 41 42
44 44 45 45 46 46 47 47

CONTENTS - Continued

Figure 8. HEMF treatment of marginal kaolin Figure 9. HEMF treatment of high Ti0 2 kaolin Figure 10. HEMF treatment of sideritic kaolin. Figure 11. HEMF treatment of discolored kaolin . Figure 12. HEMF treatment of lignitic kaolin Figure 13. Figure 14. Figure 15. Figure 16. Figure 17. Factors for converting English units to international system units

Defining a Commercial Dimension Stone Marble Property by LANCE MEADE

Southeastern Ceramic Raw Materials by J. L . PENTECOST

Economic Geology of the Georgia Marble District by ROBERT POWER Figure 1. Location map of Georgia marble district . . . Figure 2. Geologic map and section of the Tate-Marble Hill Area Figure 3. North-south cross-section through New York Mine . Figure 4 . Geologic map of the Whitestone Belt . . . . . .

Geology of Kyanite by DENNIS RADCLIFFE . . . . . . .

Figure 1. Pressure and temperature conditions for kyanite.

Figure 2. Location of the principle kyanite quartzite deposits

Table 1.

Typical composition of ore and specifications of kyanite concentrates . . . . .

Figure 3. Figure 4.

Phase diagram for the Al 0 -Si0 system . . .

2 3

2

General schematic for kyanite mining and beneficiation

Table 2. Summary of industrial applications of kyanite

Table 3. Illustration of the expansional characteristics of kyanite .

Figure 5. Photomicrograph of kyanite ore . . . . . . . . .

Table 4.

Chemical similarity of typical run-of-mine kyanite ore and a typical sandy kaolin . . . . . . . .

Figure 6. Generalized geologic map of Graves Mountain. . . .

Page
48 48 49 49
50 50 51 51 51 51 53
54
57
59 60 63 65
67
69 69 69
70 70 71 72 72
74
75 76

iii

KEYNOTE ADDRESS FOR THE FORUM ON THE GEOLOGY OF INDUSTRIAL MINERALS, ATLANTA, GEORGIA, APRIL 22, 1976
PHILOSOPHICAL TEST-PITTING, OR THINKING WHILE WE DIG
Donald A. Brobst Deputy Chief, Office of Mineral Resources
U.S. Geological Survey, Reston, Va.

It seemed appropriate to me as we start into the Twelfth Forum on the Geology of Industrial Minerals that we take a few minutes to examine the position of industrial minerals in the economy today and to con sider how geology and geologists can assist in the continued availability of these minerals in the market. An old shaggy dog proverb says, "If you don't know where you are going, you may wind up someplace else." These are the days of highly specialized skills in business, -technology, and science. Each of us looking in one of these fields has a tendency to get bogged down in the details that are strewn about our own particular part of the field. Today I choose to present the perspective of one professional geologist who has had an interest in industrial minerals for many years.
Mineral materials are a highly critical part of the economy of our nation. I am constantly amazed these days how few people really grasp this uncomplicated fact. Table 1 shows the value of the U.S. production and imports of fuels, nonmetals, and metals in wellhead and mine portal condition for the years 1972 and 1975. I chose 1972 for comparison with 1975 because 1972 was the last of the "normal" years prior to the oil

embargo of late 1973. Those figures represent the peak of economic growth in the last upward cycle. The data show several especially important things about the changes of the last 3 years. The value of domestic production had nearly doubled by 1975 to 63 billion dollars, whereas the value of imports had more than doubled. In both 1972 and 1975, the lion's share of the values shown are for fuel materials, but note the greatly increased value of both the domestic and imported fuels
in 1975. Another point needs to be made from this
table, namely that the value of the nonmetals in both
1972 and 1975 was nearly twice that of the metals.
These relative values of materials are virtually unknown to the public and, unfortunately, I suspect, unknown as well to a large segment of those involved in certain aspects of business, economics, and government. The table indicates that in 1972 the value of these materials was 35.5 billion dollars in a 1.1 trillion dollar economy, or 3.1 percent of the nation's GNP; in 1975, these materials had a value of 84 billion dollars in a 1.5 trillion dollar economy, or 5.6 percent of the GNP. In other words, in 3 years, the share of the value of these materi als in the GNP also had nearly doubled, chiefly because

TABLE 1

Mineral materials in the United States economy (In Billions of Dollars)

[Data from U.S. Bureau of Mines]

U.S. Production Fuels Nonmetals Metals
Imports Fuels Nonmetals Metals
GNP Fuels & minerals (percent of GNP)

1972
22.0 6.4 3.6
32.0
2.0 0.5 1.0 35.5 1,150 3.1

1975
49.5 8.6 4.9
63.0
19.0
2.0 84.0
1,500 5.6

1

TABLE 2

Major mines (150,000 tons or more annually) in United States producing industrial minerals1

Number of Mines 31 7 2
Total41

Commodity Phosphate Potash Asbestos Boron

1Does not include cement, sand and gravel, crushed stone

Area of U.S. Southeast and West New Mexico California and Vermont California

of increased cost of fuels. Whether the value is 3 or 6 percent is again not the most important point, which is that these fuel and mineral materials are the very foundation of our economy. Who would care to do without them?
Another way to view this mineral business is the manner of the "London Mining Journal" in presenting its annual summary of the world mining industry. The magazine lists the major mines that produce 150,000 tons or more annually of 22 major commodities involved in world trade: asbestos, bauxite, born, diamonds, titanium, potash, fluorspar, phosphate, iron, copper, lead, zinc, tin, gold, silver, platinum, manganese, nickel, uranium, chromium, molybdenum, and mercury. The Journal claims that 95 percent of all noncoal mining activities can be accounted for by a close examination of only these 22 metals and minerals. The 1975 list contained the names of 1,072 mines. The United States has 172, or 16 percent of the world's largest mines. Of these, 41 mines, or 4 percent of the world's largest mines, produce the four industrial minerals shown in Table 2. Some high-volume commodities are conspicuously absent from the list of 22, in part because they do not make a sizable part of international trade, and also because, among other things, they are more widely distributed around the world, and their value is such that long-distance hauling to market is not feasible.
Thus, it becomes clear that one must do some digging to get into the subtleties of the industrial minerals. Although these materials traditionally have carried a little less glamour than metals, many of them are indispensable in the production of energy materials and metals themselves, all the way from exploration and mining to fabrication.
Table 3 is a summary of the first 20 industrial mineral commodities, ranked by dollar value and production in thousands of short tons. Although it does not show on the table, the dollar value of the first 10 items accounts for virtually all the 8.6 billion dollars for the industrial minerals in last year's GNP. The rank of the next 10 industrial minerals emphasizes the fact that the dollar value and production of these minerals drops off very quickly after the first 10. When it comes to dollar value and production figures, however, the table really

does not tell the whole story of the critical nature of these materials.
Certainly the major share of the industrial minerals business lies in cement, crushed stone, sand, and gravel. Together these accounted for more than 5 billion dollars' worth of sales last year. The cememt came from 57 companies, which operate 175 plants. The crushed stone came from 4,800 quarries. According to the U.S. Bureau of Mines, however, 77 companies operating 740 quarries did 50 percent of the business. Limestone was 70 percent of the output, granite 10 percent, and other types of rocks constituted the remaining 10 percent. These rocks were used in construction and maintenance (70 percent), cement and lime (15 percent). agricultural uses (12 percent), and flux (3 percent). Sand and gravel last year came from 7,000 operations, but 68 percent of all the sand and gravel came from 1 ,200 commercial operations. Forty-eight percent was used in streets and highways, 49 percent in buildings and heavy construction, and 3 percent was consumed in all other uses.
The industrial mineral business has some features that make it quite different from the metals industry. Many of the companies that produce these industrial minerals are scattered over the United States, supplying materials to a great degree on a local or regional basis as well as nationally from centers of commercial ore deposits. Many of the companies have tended to be small. Because deposits of their materials have been relatively easy to find, and because relatively few restrictions have been placed on them until recently, these companies have survived and prospered, supplying the markets with needed materials at relatively low cost. Many companies have accomplished this without fulltime, or even part-time geological service.
Many signs now point to a new era ahead. Energy costs have risen for the production of everything from the basic raw materials to the finished product. Not only is it getting harder to find new, large mineral deposits close to the markets, but in certain types of industrial minerals, it is even getting harder to find a commercially usable deposit at all. The easy-to-find, wellunderstood types of deposits of many minerals close to the surface have already been found. This in turn is requiring longer lead-time to find and get new deposits into production. The reasons for this involve both

2

Cement Crushed stone Sand and gravel Phosphate Sulfur Clay Lime Salt Potash Bromine and compounds Soda ash Boron and compounds Gypsum Diatomite Sodium sulfate Barite Asbestos Vermiculite Feldspar Fluorspar

TABLE 3

Production, value, and rank of 20 major industrial minerals

[Data from U.S. Bureau of Mines]

Rank 1 2 3 4 5 6 7 8 9
10 11 12 13 14 15 16 17 18 19 20

Value ($Millions) 2,000 2,000 1,500 880 486 423 419 376 227 200 167 150 45 42 30 16 15 14 12
12

Production (1,000 Short Tons)
69,600 857,000 861,304
49,000 7,250
50,900 18,850 43.418
2,530 208
4,122 1,100 9,760
567 665 1,066 100 340 775
130

Rank 3 2 1 5 9 4 7 6
12 21 10 13
8 19 17 15 25 20 16
22

geologic availability and various aspects of economic availability.
Now land-use and environmental problems are arising. These problems involve environmental impact from the production of the materials and even the more basic questions of allowable land use for mining and beneficiation operations. The capital for investment in exploration and production facilities is getting harder to find and, even worse, it is becoming more expensive to use.
What we are seeing is a constant struggle between the constraints placed upon the supply system and the exponential growth of varying rates in each industry for supplies of those materials. The population grows, the urban community grows, and as America enters its third century there will be considerable demand for building of all sorts, including new construction as well as re placements of old facilities.
Through all this expansion, we must continue to keep in mind that exponential growth in the use of various geologic materials cannot continue forever. The rate at which we use or consume materials is certainly a very important factor, which must be considered along with the amount of materials geologically and

economically available to us. We have a tendency to look at mineral production
on an annual basis. Was this year's production larger or smaller than last year's? Often we do not look at what the trends are really telling us. For example, the annual consumption of sand and gravel in the United States rose from 300 million tons in 1950 to nearly one billion tons by 1975. That is spectacular growth, even for a commodity that is geologically reasonably abundant.
Let me give you another example of this exponentially increasing production from some data on barite, 1850-1975 (Tab.le 4). The year 1850 is a good place to start because we may consider that time to be about the beginning of modern industrial expansion in the western world. The world has produced 110 million tons of barite; nearly 80 percent of that has been pro duced in the last 30 years. The United States has produced nearly 75 percent of its barite and has received 94 percent of its imports in the last 30 years. In the 1970 edition of "Mineral Facts and Problems," the U.S. Bureau of Mines estimated that the growth rate of barite demand would be about 2 percent per year. On this basis, the U.S. demand by the year 2000 would require the cumulative production of about 37 million

3

tons of barite; a similar projection for the rest of the world would require 105 million tons. Those numbers virtually equal the total U.S. and the total world production from 1850 to the present. The oil embargo of 1973 and the resulting scramble in the search for oil in other parts of the world suggest that these estimates of barite-demand made in 1970 are perhaps low. Work done recently by the U.S. Geological Survey suggests that the United States might consume 25 million tons of barite in oil and gas exploration and production drilling in the next 12 years. That will come very close to asking industry to produce, or at least to have available through production and imports, as much barite as has been produced domestically in the last 30 years.
The chilling finale is that a one billion year supply of anything at the present rates of consumption would be exhausted in less than 600 years if there is a 3 percent rate of expanded use annually-the general rate of U.S. economic growth in the postWorld War II years. A 2 billion year supply would last only about 25 years longer, because it is equal to only one more doubling time of the use of the one billion year supply. Of course, the consumption of a finite supply would not follow such a pattern of increase to its final exhaustion, but the point is that the level of consumption is just as significant an element in the life of a supply as the magnitude of the resource. Exponential growth in consumption can quickly exhaust even a huge supply.
In short, we wi II have to learn to use better what we have geologically available to us. We shall have to go out and search for new types of deposits of many minerals. This will take time, money, and people working together and exercising brain power. We cannot afford to wait until the new deposits are needed tomorrow morning and then begin to look. Everyone involved in this business is going to have to begin taking a longer range view of the problems. It will be to everyone's advantage to do so. The cost of the burden of action can be spread around between industry, academia, and government, although it is the consumer who eventually pays. Let us not forget who that consumer is. We are, along with all the other people to whom we refer as

"they" and "them" in our criticisms. What can geologists do? A great deal, I think, from
all the places where they might be working-within companies, as consultants to companies, in the universities, and within geologic agencies of the State and Federal governments.
Geologists can deal with specific problems. Wedding science and the dollar has been done in the past and can be done with greater intensity in the future. Geologists themselves have not always taken the pains to explain in their reports what some of their rather scientific-sounding prose really means in terms of practical applications. By the same token, management in many corporations perhaps has not utilized people with much or any geology in their background; as a result, management has been unwilling to be talked into expenditures whose value to them cannot be clearly foreseen. There have been countless cases in which some basic geologic information might have saved a company not only its shirt, but its life. A few hours of a geologist's time could have prevented disaster for a company who though it was buying a 50-foot-thick gravel deposit, only to discover when the bulldozers moved in that it was only 2 feet thick and simply looked 50 feet thick because the gravels on the edges of the perched deposit had rolled downhill.
To a great degree, our industry is specificationoriented. Specifications for raw materials and products are fine, but I think that there are times when we continue to live with specifications for materials that are more historical and more dinosaurian than they really need to be today. A rock becomes a resource only when someone is willing to pay for it to do the job that they have in mind. In some jobs, limestone has been the preferred material for crushed rock. It may be hauled a few extra miles to its market area, when a granitic rock or some other type of igneous rock may be closer and just as adequate for the work to be done. In the days of World War II, premium prices were paid for so-called ruby mica, when laboratory testing showed that the other varieties, including the much more abundant green variety, were just as good for the use intended. If a

1850 - 1914 1915 . 1918 1919 1944 1945 - 1975
TOTALS
Percent of total since 1945

TABLE 4

World and United States barite production, and United States imports, 1850- 1975 (Millions of Tons)

World 6.5 1.2
16.7 86.2

u.s.
1.2 .7
7.1 27.6

110.6

36.6

U.S. Imports 0.2 .0 .9
15.2
16.3

78

75

94

4

mmmg company is looking for new deposits of a material within a given radius of its market areas, a geologist can assist the company in assessing the potential for finding a commercial deposit within the area in a short time. This is considerably more advantageous in time and money to the company than to have its marketing experts running around the area looking for something that may not be available geologically in that area.
Geologists undoubtedly will be looking for new types of deposits as well as hidden deposits of types that we know how to work. New methods must constantly be devised to do this, and we still have a long way to go. For the near term, geologists can assist in the search for what we might call undiscovered deposits of "conventional" types not known in some areas. I wi II give an example again from my own work with barite. Deposits of black bedded barite such as those of commercial value found in Arkansas and Nevada are not known in the Appalachian region now. On the basis of my work, I believe that such deposits probably do occur in the Appalachian region. Some successful effort in search of them could go a long way toward assuring a domestic supply of barite for the drilling of oil wells off the Atlantic coast. This pattern of thinking holds promise for more commodities than just barite.
State and federal agencies may produce regional, state, or national assessments of the various resources available in those areas. These generally well-documented reports will help the company and consulting geologists to provide quickly the information the company management needs to make decisions pertaining to its future operations. The results of such studies can be made available quickly through publication of maps and reports that may combine much data, presented in ways that are useful to more people than geologists. Maps

showing potential mtntng districts are really more valuable than those showing where we used to mine.
Geologists, wherever they are, can assist in educating the public about the nature of resource problems. Such problems have two fundamental aspects - geologic availability and economic availability (I would include legal aspects in this category). Geologic availability simply defined says that you cannot mine what you do not have, and that you cannot mine anything until you find it.
In the area of economic availability, we have another sign of that new era that I was talking about- environmental problems associated with mining. This is the old story that nobody wants a big hole nextdoor, but everyone wants the mineral materials they need at the lowest price and as handy to their project as possible. A basic conflict of interest exists here, and unfortunately both the consumer and the mining industry get caught. Zoning laws are frequently set up by boards that really do not understand that the availability of raw materials plays a role in the delicate balances within the economic growth that they so greatly desire to control. Many a suburban development has covered up a sand and gravel or other mineral deposit needed to make its contribution to the life of the metropolitan area. The public has not yet qrasped the face that trade-offs in environment and zoning will be required even to perpetuate a lifestyle that approaches today's, much less one that exceeds it.
I think that geologists, in industry, academia, and all levels of government, can offer constructive suggestions to assist in the formulation of realistic policies that will allow private industry to supply the needs of the nation at realistic and reasonable price levels. It will take the best and most creative thoughts of all of us; we must all keep on thinking and talking while we dig.

5

THE CRUSHED GRANITE INDUSTRY OF THE
ATLANTA METROPOLITAN AREA
Robert L. Atkins Georgia Geological Survey
W. Robert Power Georgia State University
Atlanta, Georgia

Introduction
Atlanta is the only major metropolitan area in the United States that depends on crushed granite as its sole source of local aggregate. Most cities rely on crushed limestone, natural sands and gravels, or trap rock. However, none of these rocks are available in the Atlanta area. Hence, the greatest concentration of crushed granite quarries in the country are in this area.
In 1973, 1.06 billion tons of crushed aggregate and 935 million tons of natural sands and gravels were produced in the United States. Approximately 95 percent of the natural sands and about 67 percent of the crushed aggregate were utilized as aggregate. Of the crushed aggregate in 1973; limestone accounted for 75 percent; granite, 12 percent; trap rock, 8 percent; and other rock types, 5 percent (Minerals Yearbook, 1973).
Georgia produced approximately 28 percent of the crushed granite in the United States in 1973. Georgia, North Carolina, Virginia, and South Carolina produced approximately two-thirds of the crushed granite in this country (Minerals Yearbook, 1973). Though very important to the southeast, it is apparent that this industry is minor to the nation as a whole.
Geology
Georgia includes parts of three major geologic provinces- the Coastal Plain, the Piedmont, and the Appalachian Valley and Ridge (Figure 1 ). As a result several different rock types are used for crushed aggregate within the state. The Valley and Ridge and Coastal Plain Provinces are comprised of sediments, indurated and non-indurated respectively, and limestone is the main rock type quarried for crushed aggregate purposes. Within the Piedmont Province, where granite and granitic gneiss are quarried, the major lithologies are granite, granitic gneiss, biotite gneiss, schist, and amphibolite. These high-grade metamorphic rocks and igneous rocks meet Atlanta's needs for crushed aggregate, and although the most economical aggregate sources are natural sands and gravels, the relative proximity and abundance of the city's aggregate are fortunate characteristics for this growing metropolitan area.

I. Cumberland-Plateau Section II . Southern Valley and Ridge Section Ill . Southern Blue Ridge Section IV. Southern Piedmont Section
IVa. Upland Georgia Subsection IVb. Midland Georgia Subsection V. East Gulf Coastal Plain Section VI . Sea Island Section
FIGURE 1.- Physiographic provinces of Georgia
Economics
The crushed granite industry began as a spin-off from the dimension stone industry. In 1883, crushers were installed at some of the dimension quat ies to crush waste rock yielding stone for railroad ballast and macadam roads. (Watson, 1902).
Initially the location of the quarry depended on two factors: good rock exposures and railroad facilities. The quarries were located on flat rock exposures (pavements) or bosses with steep slopes. Many of the bare rock exposures were very large - up to several tens of acres. The rock was typically massive without internal joints except for exfoliation sheeting. With increasing costs of haulage and with the advent of modern highways, quarries were forced to locate closer to the market.
The distance from quarry to market is the single most important factor in the placement of a crushed stone quarry. Because the price of crushed stone at quarries is relatively standard, it is the additional cost of haulage that determines marketability. This haulage fee is based on the difficulty of loading and hauling the aggregate off of the work site; the total amount of

6

time spent enroute to market; the amount of aggregate that is transported; and the size of the aggregate (Thurman, personal communication, 1976).
With haulage costs always on the increase, the closer a quarry is to its market, the more competitive its prices can be. An example is the small Bellwood Quarry which is located in west Atlanta. Blasting is required every two days to maintain production, and the size of its blasts are restricted. What is more, the quarry probably will not be able to expand laterally but must continue vertically (Pertsch, personal communication, 1976). Despite these restrictions, it is still able to compete for the downtown business because of its close proximity to the market.
At the present, there are 19 quarries serving the Atlanta area. Of the 19, four have been opened within the last three years (Figure 2). All are operating far below their capacity due to the current recession. These 19 quarries, all located within 20 miles of Atlanta, have individual capacities ranging from 300,000 to 3,200,000 tons per year and a combined capacity for producing 25 million tons of aggregate a year (Hitchcock, personal communication, 1976). The typical market area for each is usually within 6 miles of the quarry, and the haulage distances average between 6to 15miles.

Specifications and Testing
Specifications of the State Highway Department are written so that no rock type is automatically excluded. However, the highway business represents the major consumer of aggregate, and very few quarries are successful ventures when the rock being quarried is not acceptable for highway department t,tse.
It is interesting to note some significant differences between limestone and granite, the two most widely used rock types for Georgia highway construction. Generally, limestone makes a more suitable highway base because it grades and compacts better than granite. Granite is less dense, more durable, and not as subject to polish, therefore making it a better surface treatment material. Typically, limestone contains deleterious materials such as chert and shale and is generally stratified (inhomogenous). Quarried granite, on the other hand, is usually massive with only one lithology (homogenous) (Malphurs and Crissler, personal communication, 1976).
Where large masses of uniform rock cannot be found in a marketable area, interlayered gneiss, amphibolite, mica schist, and quartzite have been substituted into the aggregate market. The lithologies must be selectively quarried to meet varying specifications. For

FIGURE 2.- Quarries in the Atlanta area.
7

example: up to 35 percent mica schist can be used in some asphalt mixes, and up to 5 percent schist can be used in concrete mixes. Mica schist weakens concrete mixes if it exceeds the 5 percent limit because of its platy characteristics. Hornblende gneiss or amphibolite is generally not used because of its high density since it costs more to transport and results in less competitive prices.
Before placing a quarry on the approved Iist of the Department of Transportation, the department performs the following: a geologic map of suitable lithologies and deleterious material is compiled; an appraisal of the crushing operation is made; the rock is sampled from the quarry walls and/or from test corings; and the following tests are run on the samples:
1. Soft fragments (mica schist) Soft fragments, consisting primarily of mica shcist, fail the scratch test. This test consists of scratching a mineral with a rod that has the hardness of calcite. Up to 35 percent mica shist may be used in surface treatment and asphalt mixes, and up to 5 percent may be used in the concrete mixes. 2. Minus 200 Material Minus 200 material is utilized in both asphalt and concrete. Problems with minus 200 material include the following: in asphalts, too many fines prevent the asphalt from sticking to the aggregate; and in concrete with too many fines, more water is required in order to get workability. (Too much water is not acceptable because it weakens the concrete.) 3. FIat or Elongated Pieces Flat or elongated pieces are defined as having the longest dimension five or more times the least dimension. Flat or elongated pieces prevent adequate compaction and weaken the concrete. This can be a problem in the gneiss quarries due to breakage along the foliage planes. Force feeding the cone crusher usually eliminates this problem. 4. Sulfur Content Georgia is probably the only state that controls the pyrite and other sulfides in the aggregate to be used for bridge construction. Sulfides do not weaken the structure but may cause staining which is not aesthetically pleasing. 5. Soundness Soundness is not usually a problem with the granites of the state. 6. Abrasion Abrasion is the percent weight lost due to wear, and it varies from quarry to quarry. Generally, quarries west of Atlanta are class A in terms of abrasion, and those east of Atlanta are class B. Class A indicates the rock has no restrictions placed upon it because of its abrasive characteristics, but class B granites are restricted due to a high abrasion loss. Class B granites cannot be used in surface coarse treatment and certain types of plant mixes. What is more, the caprock in quarries does not meet the state specifications; it can be crushed for use in private developments such as subdivisions where

loads are not as heavy and traffic is not as great as along state roads.
The Georgia Department of Transportation groups the aggregate into two groups, siliceous and non-siliceous because the characteristics of limestones and granites are different. The state specifications on aggregate quality take these differences into consideration.
In order to meet the state's specifications, most quarries have initiated quality control programs and maintain rigorous standards. At least once a week, each quarry participating in the aggregate certification program is visited by representatives of the Georgia Department of Transportation Aggregate Control Branch. Samples are taken from the belt, truck, or stockpile, and tests are run to determine if the quarry is providing materials that meet the specifications of the state. In addition, a small number of samples is taken on the job site for statistical correlation with quarry samples. Each quarry is required to maintain an adequate quality control laboratory. The laboratories and personnel are tested and certified by the state to insure accuracy and proficiency. Each quarry has a commitment to maintain high standards of quality control and may be removed from the approved list for not fulfilling this commitment (Malphurs, personal communication, 1976).
Exploration and Development
When exploring for possible quarry sites in the Atlanta area, it should be remembered that the best quarry rocks are free of joints and other openings. This is just as important as the rock type itself because jointed rock is harder to break into smaller sizes and drilling and blasting costs increase. Also, chemical weathering which occurs along joints weakens the rock, sometimes below state specifications.
Once a suitable site is agreed upon, the rock is core drilled, and tests are run prior to the opening of the quarry. Usually a small or portable crushing plant is set up to determine if the rock is suitable to install a more permanent system (Figure 3). In addition, with a portable crusher, less expense is incurred if the market for the stone does not develop.
Most quarries are opened on flat rock exposures and require no initial stripping. If they expand horizontally beyond the exposed pavement, removal of overburden by pans may be necessary. Even where soil overburden is absent, the upper few feet of rock that is caprock may be unacceptable because of incipient chemical alteration. Some quarries must be expanded vertically using conventional open-cast techniques. Although this method may eliminate stripping and require less land, water becomes a problem and haulage costs increase. In some quarries haulage costs are reduced by placing the primary crusher in the quarry and transporting the rock from here to the plant on a conveyor belt.
The size and shape of a quarry varies. Some

8

FIGURE 3.- Example of a portable crushing plant
quarries have attained a depth of 300 feet; others occupy up to 100 acres of land. Typically, a quarry has two or more levels on which to work. Depending on the rock classification and/or the plant's capabilities, all sizes of crushed aggregate are produced for highway construction, railroad ballast, subdivision development, and building construction.
Production Methods
After a new quarry site has been selected and determined economically feasible, production begins. This entails blasting, crushing, and screening of the granite.
Blasting patterns, the number of blasts necessary, and the depth to which the blasts are set all depend upon the type of rock and equipment, the strength of the explosives used, and the environmental factors involved. Next, the rock is typically loaded with an electric shovel of 4 to 7 yards capacity. Oversized blocks are drop-balled, and a rubber tire loader cleans up. Then trucks of 35 to 85 ton capacity haul the rock 300 to 3500 feet to the primary crusher. Usually, this crusher is located on the lip of the quarry, but several operators have located them within the quarry. Though the primary plant operation varies slightly from plant to plant, it generally consists of crushing and screening until the desired size of aggregate is obtained. This crushing of the granite is expensive and very hard on the equipment. In fact, the crusher has to be adjusted almost on a weekly basis (Lambert, personal communication, 1976). When the screening process is completed, the stone is stockpiled according to size. Underneath each stockpile is a vibrating feeder which can transfer the stone to the loading bins or blend different sized stones together. The aggregate is ready for transporting to its respective market at the end of this process.

Environmental Protection and Restoration
In 1969, Georgia passed its first state law governing the removal of rock, the control of waste products, and reclamation of the land. As a result, every operator must be licensed and bonded, and each year he must submit a mined land-use plan to the Environmental Protection Division. The purpose of this is to make the operator aware of his future needs for expansion, development, location of reserves, water control, dust control, and land reclamation (including the anticipated cost of the reclamation) . This requirement is for all types of mining in Georgia . The following is a summary of points which the land-use plan must include:
1. Total acreage involved in the property. 2. Total acreage to be affected when the mining
is complete. 3. An anticipated date for the completion of the
mining. 4. Number of acres to be affected and licensed
during this licensing period. 5. Map illustrating use of the land-haul roads,
quarry, and ponds. 6. Reclamation date and objective . None of the
quarries in the area is yet reclaimed. Most operators suggest that the quarries would make a lake or a solid waste disposal site. 7. Plans for natural drainage and water disposal. Quarries typically have closed water systems. The water is recycled through the plant. 8. Dust control plans. The operator must give the source and details of control methods. Equipment may include waterwagons to control the dust on haul roads and spray bars in the plant to control dust in the crushers, belt, and stockpiles. 9. Noise . Noise is regulated by the city, county, and Mining Engineering Safety Administration. Quarries located near housing or apartments are restricted as to their working hours; others are not. Most complaints about noise refer to trucks hauling the stone from the quarry. The in-house noise is controlled by the MESA. Ninety decibles per hour is the maximum noise allowed for an eight hour day.
References
Georgia Department of Transportation, 1972, Standard specifications, construction of roads and bridges: 669 p.
U.S. Bureau of Interior, 1973, Stone in Minerals Year-
book: Vol I, p. 1383.
- - - - - - - - - 1973, Sand and gravel in
Minerals Yearbook: Vol I, p. 1383. Watson, T., 1902, Granites and genisses of Georgia :
Georgia Geol. Survey, 367 p.

9

THE ORIGIN OF GEORGIA'S KAOLIN DEPOSITS
Roger S. Austin Freeport Kaolin Company
Gordon, Georgia

Introduction
Commercial deposits of white kaolin clay occur along the northern margin of the Georgia Coastal Plain. The center of kaolin production is in Twiggs, Wilkinson, and Washington Counties, with related operations extending into Alabama and South Carolina. Low grade bauxite is associated with most of these deposits, and an important alumina industry has developed in the same geological area, near Andersonville, Georgia. The kaolin and associated bauxite occur within Cretaceous and Early Tertiary strata which dip gently toward the present coast. These strata unconformably overlie metamorphic and igneous rocks which crop out immediately to the north in the Piedmont Province.
Raw kaolin is processed to form a fine, white, high-brightness, pure kaolinite clay product. This material is shipped in dry or slurry form to the paper, paint, plastics, and rubber industries and is used as both pigment and filler.
Of all the kaolin occurring in the "laolin belt", only a small portion is of commercial value. These unique deposits generally must:
1. Have an overburden to kaolin ratio of less than10:1.
2. Contain less than 10% sand. 2. Have a clay particle size distribution of 55%
to 90% finer than 2 microns. 4. Be extremely white and have high reflectance. 5. Be able to produce a highly fluid aqueous
slurry containing up to 70% kaolin. The commercial kaolins are thus only those portions of the strata which consist of nearly pure kaolinite. The deposits must be at least 1.2 to 3.0 meters (4 to 10 feet) thick, many acres in areal extent, and contain hundreds of thousands of tons of raw clay.
The origin of the kaolin includes three principle factors.
First- The Georgia kaolin deposits occur within a regional system of aluminous laterite and lateritic sediment extending from Arkansas through Mississippi, Alabama, Georgia, and South Carolina in the Gulf and Atlantic Coastal Plains and from Alabama to Virginia in the Appalachian Mountains (Fig. 1 ).
Second- The kaolin and associated bauxite occur among 2 significantly different stratigraphic units. The difference is not simply one of age, but also mode of origin. The kaolin and bauxite in the Cretaceous units were formed by laterization, and are residual products derived from former aluminous sediments. The kaolin and bauxite in the Tertiary units include both lateritic and sedimentary deposits. They were derived largely

from the Cretaceous deposits by erosion and were deposited in normal depositional environments. Some were then subjected to even further laterization.
Third- Though many of the Georgia kaolin deposits occur among Cretaceous sediments, the episode of laterization, erosion and redeposition which formed them is an Early Tertiary phenomenon. The kaolin and bauxite deposits are not the result of some unusual mechanical sorting or sedimentation process.
Description of the Deposits
Below the kaolin-bearing strata are a wide variety of metamorphic and igneous rocks. Where exposed and when encountered by exploratory drilling, the upper portion of these rocks are weathered, in some cases to a kaolinite and quartz residuum almost indistinguishable from the overlying sediment. These rocks are separated from the overlying Cretaceous sediments by a distinct unconformity.
Above are the Cretaceous units which consist of poorly stratified, nearly horizontal beds of quarz sand, kaolinitic sand, sandy kaolin, and kaolin. Locally, portions of the kaolin-rich sediments contain pisolitic kaolin and bauxite. Only the gross features of stratification remain to be seen, ranging from the characteristic parallelism of sediments to extensive crossbedding and associated quartz and kaolin pebble conglomerates within large cut and fill structures. The Cretaceous sediments in Central Georgia contain no marine invertebrate fossils. Only recently have palynological studies supported the oft-proposed Cretaceous age for these units.
The Cretaceous and Tertiary strata are separated everywhere by a major unconformity . The Tertiary strata consist of a variety of sediments including quartz sand, kaolin, conglomerates, lignitic kaolin montmorillonitic clays, and limestone. It is within, Tertiary sediments that marine invertebrate fossils first appear in the Georgia kaolin district. The distinctly marine units of Eocene age form the upper boundary of kaolin occurrence.
Residual bauxite and kaolin deposits are mostly of Cretaceous age in the central and eastern portions of the Georgia kaolin belt. The Tertiary kaolins in this area are largely non-pisolitic, though some exceptions do occur. In the western portion of the belt, the Tertiary kaolins and bauxites are the dominant commercial units, for most of the Cretaceous deposits appear to have been removed by erosion since their.

10

- - - --...:
I
'~- ----------\

INTERIOR

-- - --..1> < ~ ) ,
,/
r'
LI __ ___ /
'
1

'I--N-- --- --- - -~..r -

100

200

MILES

~ EOCENE
~tiCRfliNt ~~>,o.,>.,.Hs
l lh( ... ..u ltJe.~ ll l ~ . , , ..
Will PALEOCENE
[_~-]CRETACEOUS
[, !BAUXITE AREA
Early Wilcox Shoreline

Kaolin and bauxite areas (black) of the southeastern United States (after Gordon, Tracey, and Ellis, 1958)

development. Only partially kaolinitized feldspathic sands and fossiliferous marine shales remain . Throughout the belt, bauxite occurs mostly in the middle portions of larger kaolin and sandy kaolin beds, or at the top of kaolin beds if erosion has removed the upper portion of formerly larger units.
A detailed comparison of Tertiary and Cretaceous kaolins reveals several marked differences. White Cretaceous kaolin commonly has very faint brown spots or mottled patterns compared to the uniformly white or greenish white Tertiary kaolin. The Tertiary kaolins commonly contain tubular forms, ranging in diameter from 1 to 3 mm (.04 to .1 in .) Whether burrows, fecal remains, or bryozoa casts, these tubes apparently are absent in Cretaceous kaolin. Isolated grains of very coarse sand and pebbles are present in some Cretaceous kaolin, suggesting in many cases an unnatural degree of poor sorting, a fa bric produced only by extensive solution and recrystallization. In contrast, large isolated grains of quartz sand are less common in the T e rtiary kaolin. Abund a nt submicroscopic quartz occurs only in the Tertiary kaolins, suggesting a truer sorting. Tertiary kaolins generally are more dense, have a lower porosity, and appear more compact than the Cretaceous kaolin s. These have often been referred to as "hard" kaolin s in contrast to the "soft" Cretaceous kaolins which tend to be mealy or even powdery when dry.

Examination with the petrographic microscope of kaolin thin sections showed even more distinction . Quartz grains in Cretaceous kaolins (Fig. 2) exhibit severe chemical corrosion, sometimes to the extent of deep embayment or even division into several fragments, all in optical continuity. Severe corrosion of the Tertiary quartz grains is less common. Large vermicular crystals of authigenic kaolinite, up to several mm in length are abund a nt in Cretaceous kaolin, yet uncommon or rare in Tertiary kaolins. The Tertiary kaolins are generally finer-grained than the Cretaceous, the large majority of their clay fraction being finer than 2 microns.
Heavy min e ral s of microscopic size are present in both Cretaceous and Tertiary kaolins, and there seems to be no significant difference between the mineral suites of the two . Th e suite, however, is re markable in its simplicity. Zircon, rutile, some tourmaline and amorphous iron compounds were commonly the only phases pre se nt. These are all extremely stable pha se s under conditions of even th e most severe chemical weathering.
Kaolin mine s in central Georgia provide excellent exposures of the Cretaceous units. One in Washington County (Fig . 3) conta ins a 6 meter (20 foot) thick layer of nearly pure kaolin. Above is 6 to 9 meters (20 to 30 feet) of the more characteristic sandy kaolin and sand with vi sible large-scale sedi-

11

1mm
Pen sketches of selected quartz grains observed in thin sections of kaolin.

mentary structures. These are separated from over lying Tertiary sediments by an unconformity. Immediately below the unconformity, as is most common, is bauxite in discontinuous pods and lenses. Another mine nearby (Fig. 4) is similar in many respects. Again, sands and sandy kaolin are dominant. A bed of commercial kaolin has been exposed at the base. Bauxite occurs at the uncon formity at the top. Above the unconformity is a typical Tertiary sequence of sands and montmorillonitic clays. The development of the bauxite appears ot have gone to completion prior to the final development of the Cretaceous-Tertiary unconformity. The bauxite had been partially removed by erosion prior to Tertiary sedimentation.
At a mine in Twiggs County, the relative age of the bauxite is established by the burrow of a small animal that lived on the erosion surface. Figure 5 shows a sketch of a portion of the network of chambers and tunnels these organisms formed. At the bottom of the sketch is an enlarged view of the burrow showing where the animal cut indiscriminately through individual pisolites. The bauxite certainly predates this activity.
Another mine in Twiggs County provided an excellent exposure of Tertiary strata, and a sketch of the mine wall is shown in Figure 6. The Cretaceous-Tertiary unconformity is exposed throughout the mine. The Cretaceous materials include massive commercial kaolins

TOP OF WHITE COMMERCIAL KAOLIN

LOWEST

Sketch of typical kaolin mine in Washington County, Georgia.
12

.. .... - - .. .
* ..... . .

- - ... .. . . >-

0::

:-----=-- .. - .-..

_.._-_....

<r

f-

SAND AND FULLER'S EARTH ..

0w::

- ..

f-

o .

., . _ .. _ ....... . ., . . . UNCONFORMITY

- ~::- 1' ~ ~ CRO SS-BEDDED . SA ND;;:<,:'.:-~ -
?~:~; t/!A+-~~?fltf~f!:;-)~ ~-
~/:&.-.;!...:-;--
.-~"!~~?::';\~-'--
. : . : : .... .. . i

KAOLIN MATERIAL

(f)
:::> 0
w
(.)
<r wf-
0::
u
30 feet

Sketch of kaolin mine high wall showing relation between kaolin and bauxite.

and poorly bedded sand and sandy kaolin. The Tertiary strata are the most striking. A large channel is present and is filled with quartz sand, kaolin pebble and boulder conglomerates, bauxite boulders, and cross-bedded bauxite pisolites. Though the bauxite detritus is present, no residual bauxite was exposed. These units grade upward into sandy kaolin beds containing terrigenous plant fossils and finally, small invertebrate marine fossils. A succession of marine sands, limestones, and montmorillonitic clays follow.
Geologic History
The geologic history of the Georgia deposits is essentially outlined in the generalized stratigraphic section shown in Figure 7. By Late Cretaceous time, a thick sequence of feldspathic sands, silt, and clay (perhaps already kaolinitic) had been deposited in the Georgia region, part of a more extensive series of apron-like deposits generally near and paralleling the present inner margin of the Atlantic and Gulf Coastal Plains. At the end of Cretaceous or in very early Tertiary time, the Cretaceous sea retreated and exposed the sedimentary deposits to subaerial weathering. The sediments were subjected to a climatic and subsequent weathering episide of intensive or prolonged rainfall which brought about extensive leach ing of the newly-exposed sediments by a continual

~

~ ."
.........

o~- --~ .* -~~
:.... . + , , " .

;;.,t~.rt;! IAR'f:~~;:.~."<4'
C JI ET"A,( EOUS

............ . . . (a) BURROW COMPLEX
............. .. .............. . . l .

ONE FOOT
.

@@
@
@
o
g

( b) DET A IL OF BURROW PA SS AGE (ac1ual size)

Sketch of burrow (crustacean) at the Cretaceous-Tertiary unconformity in Twiggs County, Georgia.

13

Sketch of kaolin mine high wall in Twiggs County, Georgia . Cretaceous kaolin is below unconformity and Tertiary deposits, above.

SAND FULLER'S EARTH
SAND

I 10 FEET

downward flux of fresh ground water. The products or residues of this episode were an aluminous laterite which consisted of massive beds of bauxite and kaolin interspersed among quartz sands. The bauxite formed in the upper portions of the Cretaceous strata where leaching was relatively more effective (Type lA); and kaolin below (Type IB), where leaching was less effective, the ground waters having become too saturated with silica for bauxite to be stable.
During the development of the lateritic deposits of kaolin and bauxite, streams were cutting down into the altering sediments, eventually causing the removal of large portions. Some of the eroded material accumulated in the stream channels. These new kaolin beds, if deposited early enough, were then subjected'to further leaching, further kaolinitizing and bauxitizing portions (Types II A and B).
By Late Paleocene or Early Eocene time the climate became more temperate, the amount of dissolved silica in the ground water increased, and bauxitrization ceased. Portions of the bauxite bodies then began to be resilicated especially around their margins, forming a secondary residual kaolin (Types IC and II C). This is the only adequate explanation for the common occurrence of bauxite bodies being completely enveloped by kaolin. The kaolin is simply an alteration rim around an unaltered bauxite core. Erosion and redeposition of kaolin and bauxite continued during this period, producing deposits

which were subjected to little or no further alteration (Type Ill).
During Middle to Late Eocene time, the sea began to advance across the altered, eroded, and redeposited Cretaceous, Paleocene and early Eocene deposits. Stream valleys were flooded by marine waters. Where the saline waters invaded the bauxites, montmorillonite, rather than kaolinite, became the alteration product.
Finally, the sea covered the entire region and the formation episode was brought to a close . A blanket of sand, montmorillonitic clay, and limestone was deposited, preserving the deposits for our use today.
Economic Aspects
The kaolin industry now benefits from several specific episodes.
1. Solution, recombination and recrystallization have produced great quantities of kaolinite. At the same time, it reduced the abundance of or eliminated most other minerals. Quartz and degraded white mica constitute most of the nonclay residue.
2. Solution was effective enough to remove most carbon and iron, producing a bright, white product.
3. Resilication altered large masses of bauxite to kaolin, much to the benefit of the pigment industry.

14

On the other hand 1. Not all deposits had enough quartz or mica re moved by solution and are too sandy for current production standards. 2. Resilication under saline conditions permitted the development of montmorillonite, the major cause of non-fluid clay slurries. 3. More recent infiltration by ground water has carried in iron which has been subsequently oxidized or reduced, producing red, yellow and purple colors on the one hand and grays on the other. 4. Extensive beds of kaolin still remain buried too deeply to permit economic mining operations because of excessive overburden.
Selected References
Buie, B. F., Fountain, Richard C., 1967, Tertiary and Cretaceous age of kaolin deposits in Georgia and South Carolina (abs): Geol. Soc. America- Program with abstracts for the Southeastern section meeting, p. 19.

Gordon, Mackenzie, Jr., Tracey, Joshua 1., Jr., and Ellis, Miller W., 1958, Geology of the Arkansas Bauxite Region: U.S. Geol. Survey Prof. Paper 299, p. 1-261.
Kesler, T. L., 1963, Environment and origin of the Cretaceous kaolin deposits of Georgia and South Carolina: Georgia Mineral Newsletter, Georgia Geol. Survey, v. 16, p. 3-11.
Overstreet, E. F., 1964, Geology of the Southeastern bauxite deposits: U.S. Geol. Survey Bull., 1199-A, p. 1-19.
Smith, R. T., 1929, Sedimentary kaolins of the Coastal Plain of Georgia: Georgia Geol. Survey Bull. 44, p. 1-482.
Tschudy, Robert H., and Patterson, Sam H., 1975, Palynological evidence for Late Cretaceous, Paleocene, and Early and Middle Eocene ages for strata in the kaolin belt, central Georgia: Jour. Research U.S. Geol. Survey, v. 3, p. 437-445.
Veatch, Otto, 1908, The kaolins of the Dry Branch Region, Georgia: Georgia Geol. Survey Bu II. 18, p. 1-453.

o

o

o

o

I

B

..UJ
1-

- -- -

...J

FIGURE 7 Generalized stratigraphic section in the middle Georgia kaolin region .

CRETACEOUS

TYPE I

A

B

C

Bauxite Kaolin Kaolin

lateritic product derived Cretaceous sediments Alteration rim derived from bauxite.

TERTIARY TYPE II A Bauxite B Kaolin C Kaolin

Lateritic product derived from trans ported altered Cretaceous materials. Alteration rim derived from bauxite .

TYPE Ill

A Bauxite (rare) B Kaolin

True sedimentary deposits from I and II.

15

ALUMINA FROM DOMESTIC RESOURCES
Don H. Baker, Jr., Research Supervisor U.S. Department of the Interior Bureau of Mines
Boulder City Metallurgy Engineering Laboratory Boulder City, Nevada

Introduction
At present, nearly all of the primary aluminum produced in the United States is obtained by the electroreduction of alumina extracted from bauxite by the Bayer process. Although world reserves of bauxite are adequate for many decades at the current rate of aluminum production, domestic reserves are limited, and as a result, the United States now rei ies on foreign suppliers for about 90 percent of its raw material. Global demand for bauxite is increasing rapidly, and this combined with production levies and the nationalization of foreign bauxite mining operations, has placed the United States in a position in which it faces additional increases in the cost of imported bauxite, as well as possible constraints on its availability.
There are, however, vast quantities of non bauxitic aluminous minerals in the United States, such as clay, anorthosite, alunite, and dawsonite, which represent potential alternatives to bauxite. In the interest of establishing a domestic productive capability for alumina from substitute materials, the Bureau of Mines is pursuing a miniplant program at the Boulder City Metallurgy Engineering Laboratory. This program is designed to provide a comparative evaluation of the most promising technologies applicable to these materials.
J. E. Husted (1974) 1 estimated that there are about 9 billion tons of clay. His tabulation outlined only the significant clay deposits, most of which contain more than 50 million tons. Clay reserves that would be amenable to acid leaching (those containing 30 percent or more Al 2 0 3 and less than 6 percent iron) amount to 7 billion tons, representing the equivalent of 1.24 billion tons of aluminum. This is sufficient aluminum to meet the domestic annual primary demand (industrial demand less secondary supply) for almost 200 years, based on the annual consumption in 1973 of 5.242 million tons (Bureau of Mines, 1975) . Husted also indicated that there are massive (440 billion) tonnages of anorthosite with an average alumina content of about 26 percent, which is sufficient to supply the primary demand at the 1973 rate for more than 19,000 years.
In December 1970, the National Materials Advisory Board made a study entitled "Processes for Extracting Alumina From Nonbauxite Ores". The immediate recommendation of the report centered around nitric and hydrochloric acid processes to treat clay. Other resources were considered to have potential, but each had technical problems associated with the
1Numbers refer to items in the list of references .
16

extraction of alumina which needed additional study. The plan was not implemented at that time. However, in July 1973, the Bureau of Mines instituted a six-point miniplant program to provide comparative engineering data for clay/HN03 , clay/HCI, anorthosite, clay/H 2S03 , alunite, and dawsonite processes outlined in the literature. The first-year funding was $400,000, and the program was scheduled to continue through September 1979.
The program was accelerated in July 1974, when eight aluminum companies joined in a cost-sharing effort, with each company contributing $50,000. The original cooperators were Aluminum Company of America, Aluminum Company of Canada Ltd., Alumax Inc., Anaconda Aluminum, CONALCO Inc., Kaiser Aluminum and Chemical Corp., Martin Marietta Aluminum Inc., and Reynolds Metals Co. Combustion Engineering Inc., and Vereinigte Aluminum Werke AG of West Germany joined the group in the spring of 1975. Representatives from each of the 10 companies and the Bureau meet quarterly as a steering committee to coordinate and assist in directing and evaluating the miniplant program.
Accomplishments to date include nitric acid leaching of kaolinitic clay which has been tested in a miniplant, and about half of the unit processes involved in leaching clay with hydrochloric acid have been tested. Some of the observations and data developed in these efforts are reviewed in this report.
General AcidLeaching Process
The clay raw material must be heated to a temperature sufficiently high to cause decomposition of the hydrated aluminum silicate and thereby make the alumina soluble in acid. This occurs between 700 and 800 C, at which temperature amorphous oxides of aluminum and silicon are formed. The calcined clay is then treated with acid, which dissolves the alumina as the aluminum salt of the acid but does not dissolve the silica. This salt is crystallized and separated from the mother liquor and is subsequently thermally decomposed to produce a pure alumina and the acid for reuse in the process. A generalized flowsheet for this process is presented in Figure 1.
Installation of the nitric acid miniplant was initiated in August 1973, and the plant was first operated as an integrated series of unit processes through the crystallization step in May 1974. Construction of the hydrochloric acid miniplant is still in

Makeup acid

Row cloy Calcinoti on

Leaching

Recycle acid

Organic solvent

Liquid/solid separation
Solution purification

Solid to waste
-
Impurities

Recycle

Acid

liquid

regeneration

LAcid

vapors

Crysto llizotion
Liquid/crystal separation
Crystal decomposition

Calcination

Alumino

FIGURE 1- Generalized acid-leaching-process flowsheet
17

progress, but major portions of the plant have been operated.
Raw Materials
Clay was obtained in 50-ton quantities of runof-mi ne materia l through the courtesy of the Thiele Kaolin Company 2 of Sandersville, Ga. Petrographic analysis showed m inor quartz fragments and feldspar (mostly potassium) and scattered accessory mine rals in the clay matrix. Most opaques we re extremely fine grained and probably kao linite. A petrograph ic grain count is presented in Table 1.
Table 2 gives a typical chemical analysis of this clay after drying at 115 C. The analysis indicates the clay to be of high quality and reasonably low iron content . Data in Table 3 indicate the fineness of the dispersed clay as determined by the ASTM D-422-63 soil hydrometer method .

TABLE 1. - Petrographic grain count

Constituent

Estimated percent

Clay ....... ... . . . . .. .. . . ... . .. .... . 96-97 Quartz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Feldspar . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Epidote. . . . . . . . . . . . . . . . . . . . . . . Not detected Fluorite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Zircon. . . . . . . . . . . . . . . . . . . . . . . . Not detected Organic matter . . . . . . . . . . . . . . . . . . . . . . . . 1 Muscovite ... .. . .. ........ . . . ...... . . Opaque ..... ................... .... .

TABLE 2.- Chemical analysis of dry clay Component

Wt-pct

Al 2 03 . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . 37.41 FE 2 P3 . . . . . . . . . . . . .85 Na 2 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .09 K2 0.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .09 CaO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .09
MgO.. . . . . . . . . . . . . . . . . . . . . . . ... . . .. .06
P2 0 5 . , . . . . . . . . . . . . . . . . . . . . . . . . .05 Si02 .. . .. . ... ... . ..... . .. .. .... .43.31 lnerts 1/ .. . . . ... . . .. ..... .... ..... . . 4.85
LOI .. . ... . . . . . . . . . . . . . . . . . . . . . ... . 13.20

.!_/Obtained by difference, mostly TiO 2

TABLE 3.- Cumulative weight-percent of material
finer than indicated size

Size fraction, micrometers

Cumulative weight, percent finer

105 .. . . ... ....................... .99.6 74 . . ......... ... . .......... . ..... .98.9 74 ....... ...... .. ... . . ... . . .... ...97.8 37 . ................ .. ...... .. .....97.6 16.1 .... .. . . . . ... .... .. ...... . .... .91.0
16.4 1I . ........ . ... . . ... ...........90.9
9.6 1/ ........ .. ...... . . .. . .. ..... .88.9 6.8 1/ ... . ...... ... . . . .............87.9 4.8 1/ .. . .. .. .. . . . . .. . . . . . . . .. .. . . .86.9 3.5 1/ . . . . . . . . . . . . . . . .. .. . . . ... ... .85.9 2.0 1/ ....... . .. .. .. ..... ... . . . ....84.8

.!_/Equivalent spherical diameter of particles determined using soil hydrometer method. Calculations were made using 2.62 as the specific gravity of the clay.

In the nitric acid miniplant, commercial-grade acid (42 Be) diluted to 50 percent concentration was used. Di (2-ethyl-hexyl) phosphoric acid (DEHPA) combined with kerosine comprised the organic phase in the solvent extraction (SX) circuit.
For the clay/hydrochloric acid miniplant, commercial-grade acid (20 BE) was used. Reagents used in the solvent extraction circuit consisted of kerozine, a tertiary amine (Aiamine 336), and n-decyl alcohol.
Discussion of Process
Clay Preparation
The clay crushing and calcination sections were common for both the nitric and hydrochloric acid miniplants. The bulk clay was crushed to pass a 2-mesh screen and the minus 2- plus 20-mesh fraction was fed to a countercurrent gas-fired kiln. The minus 20-mesh fraction was pelletized on a disc pelletizer, and the 3/8-inch wet pellets also were fed to the kiln.
The kiln was 15 inches inside diameter by 21 feet long. Calcination temperature was maintained by an automatic gas burner controlled by the temperature of the clay as it entered the enlarged section of the combustion zone. Calcined clay, the analysis of which is shown in Table 4, was crushed to a minus 10 mesh for feeding to the leach circuit of the mini plant.

2 Reference to specific companies or brands is for

identification only and does not imply endorsement

18

by the Bureau of Mines.

TABLE 4.- Chemical Analysis of Calcined Clay

A l2 0 3 , total .. .. .. ... . . . . . . Acid-extractab le A l2 0 3 . .

. . . . .43.10 . 1/40.81

FE2 0 3 ... . . .

.98

Na2 0 . . .. . .... .. ..... .

.10

CaO .. . ..... .. . ..... .... .10

MgO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .07

.06

.. . . . .. . .. . .. .. .. .. .. .. . .49.90

Inerts

. . . . . . . . . . . . . . . . . . . . . . . . . . 5.59

Total . . . . . . . . . . . . . . . .... .. . . . . . . . . 99.90

1/Extractable under the condition employed in the miniplant.

Nitric Acid Process

The leach section of the nitric acid miniplant

consisted of a stirred cold-mix pot where cold calcined

clay was slurried with nitric acid before entering a series

of three stirred reactors. The slurry produced in the

cold-mix pot overflowed into the first reactor, a jacketed

vessel constructed of Carpenter 20 stainless steel and

heated with a circulating heat transfer fluid to maintain

a minimum temperature of 100 C. The vent of the

reactor was fitted with a water-cooled condenser to

return any acid vapor to the reaction chamber. The

reaction,

A 12 0

3

+

6HN0a+2AI(N03 )3

+

3H 0,
2

repre-

sents the leaching step. The slurry leaving the first

reactor overflowed and cascaded through two additional

insulated, stirred 316 stainless steel reactors to complete

the leach section.

The discharge from the third reactor flowed into

a drag classifier that separated the coarse sand residue

from the slurry. The sand discharge after one washing

was stored for further testing. The pool overflow from

the No. 1 classifier went to a 3- by 3-ft-diameter thicken-

er. The thickener underflow was stored for future testing,

and the clear overflow raw, pregnant solution was

filtered in a 12-sq-ft plate and frame filter to remove

suspended solids. This yielded a polished pregnant

solution from which the iron was removed by solvent

extraction .

To remove the iron, the polished pregnant solution

was heated to 40 C and treated with organic extrac-

tant in three cells of an 11-cell, countercurrent, mixer-

settler system. The organic phase consisted of 59 percent

kerosine, and 41 percent DEHPA. The iron-laden organic

phase was washed with water in the fourth and fifth

cells prior to being contacted with 17 percent HCI in the

sixth through ninth cells to strip the iron. The organic

solvent was washed free of chloride in the 10th and

11th cells before being recycled.

The purified pregnant solution was pumped to an evaporator operated at 120 C, in which the concentration of the aluminum nitrate solution was increased by evaporating some of the water. The concentrated solution was then cooled to about 55 C, which resulted in the crystallization of aluminum nitrate nonahydrate. A drag classifier was used to separate the crystals from the solution, and the mother liquor was recycled to the evaporator and crystallizer. The drained crystals were stored for subsequent use in studies involving thermal decomposition to produce alumina.
Thermal decomposition of the aluminum nitrate nonahydrate was accomplished by several methods includ ing a direct-fired, fluidized-bed roaster and a heated-tube roaster. Decomposition produced nitric acid, oxides of n itrogen, and hydrated alumina . The hydrated alum ina was then fired at 1,050 C to produce alumina that was about 50 percent in the alpha form, the crystalline form desired by part of the aluminum smelter industry. Most of the decomposition of the AI(N0 3 )3 9H 2 0 took place at about 400 C according to the following equation :
AI(N0 3 ) 3 9H 2 0(sl-+1/2 Al 2 0 3 (sl
+ 7 1/2 H2 0(gl + 3HN0 3 (gl
The miniplant evaluation of the nitric acid extraction of alumina from clay shows it to be a technically viable process. Data from these operations, supplemented by laboratory tests, have shown that 97 .8 percent of the extractable alumina can be recovered. This represents 93.0 percent of the total alumina in the clay tested . The acid loss, which is about 7 percent, results primarily from nitrates left in the trailings and from oxides of nitrogen formed during crystal decomposition, which were not recovered by scrubbing.
Theoretical energy data summarized in Table 5 show that, for the conditions listed, there is a th eoretical heat requirement of 651,000 British th ermal units (Btu) to produce 40.4 pounds of Al 2 0 3 . This is equivalent to 32.2 X 106 Btu/ton of Al 2 0 3 produced . Consideration of the theoretical energy requirements serves two functions: (1) It provides a basis on which various processing techniques can be compared, and (2) it points out the energy intensive process steps where efforts to minimize energy requirements should be made by either modification of the processing steps or by recove ring or saving process energy.
Hydrochloric Acid Process
The flowsheet for the hydrochloric acid leach process for the treatment of calcined clay is very similar to that presented in Figure 1. The process differs from the nitric acid technique in that the crystallization of the aluminum acid salt in the nitric acid process is temperature dependent because the crystals melt at 73 C in their water of crystallization, whereas in th e HCI process, crystallization is concentration de-

19

TABLE 5- Theoretical energy data*, nitric acid miniplant

Process step Calcination
Leaching

Flowsheet materials, lb
143.7, raw clay 100, calcined clay 43.7, H2 0
324.3, 50 percent HN03 100, calcined clay 424.3, slurry

Theoretical energy requirements, Btu

Needed

Available reclaim

118,000

35,000 74,000

45,000

Liquid-solid separation

424.3, slurry 131.3, wash water 336.2, pregnant solution 115.6, sands

2,000

Solvent extraction
Evaporation/ crystallization

440.2, pregnant solution 437.0, purified solution
2.0, FeCI 3
437.0, purified solution 134.9, vapors 311. 8, crystals
5.0, bleed stream

195,000

148,000 66,000

Decomposition Acid recovery misc.

311.8, crystals 271.4, acid vapors
40.4, Al 2 0 3

338,000

252,000 22,000
3,000

Total

40.8 lbs Al 2 0 3 in 40.4 lbs Al 2 0 3 out

651,000

Energy needed= 651,000 Btu per 40.41b Al 2 0 3 produced or 32.2 x 106 Btu/ton alumina produced .

*For ease of comparison these data and the data presented later for the clay/HCI process are based on a feed rate to the leach section of 100 pounds of calcined clay per hour.

pendent. There are two ways of effecting crystallization of AICI 3 6H 2 . One is through evaporation to increase the concentration to saturation thereby causing crystallization. The second, referred to as gas sparging, uti Iizes the addition of hydrogen chloride gas to raise the acid concentration in order to depress the solubility of aluminum chloride, thus resulting in crystallization.
The hydrochloric acid miniplant equipment was fabricated from fiberglass and plastics or metal coated with glass, rubber, and/or Plastisol.
The leach section equipment consisted of a cold-

mix chamber in which a clay/acid slurry was formed, four jacketed, glass-lined Pfaudler reactors, and a heat exchanger. The coarse waste in the slurry discharge from this section was separated and washed in two spiral classifiers while the slimes (or fines) were separated from the pregnant solution in two 5- by 5-feet thickeners. Final polishing of suspended solid from the solution was accomplished by a plate and frame filter prior to treating the solution by solvent extraction for the removal of iron. The polished pregnant solution is contacted with an organic phase, consisting of 16percent Alamine 336 stabilized with 11 percent

20

TABLE 6.- Theoretical energy data, hydrochloric acid miniplant sparging process

Process step

Flowsheet materials, lb

Calcination ............... 143.7, raw clay 100, calcined clay
43.7, H2 o
Leaching . . ..... . ........ 361.0, 26% HCI 100, calcined clay 461.0, slurry

Liquid-solid separation . . . . .

461.0, slurry 57.8, wash water
403.2, pregnant solution 115.6, sands

Solvent extraction

403.2, pregnant solution 401.2, purified solution
2.0, FeCI 3

Sparging crystallization ... ...

401.2, purified solution 106.3, HCI gas 191.1, crystals
20.5, bleed stream 295.9, recoverable acid]!

Decomposition . . . . . . . . . . .

191.1, crystals 150. 7, acid vapors 40.4, Al 2 0 3

Acid recovery . . . . . . . . . . . . Miscellaneous

Total . . . . . . . . . . . . . . . . . . 40.8 Al 2 0 3 in 40.4 Al 2 0 3 out

Theoretical energy requirements, Btu

Needed

Available for reclaim

118,000

35,000 74,000

13,000

8,000

221,000 12,000 351,000

68,000
7,000 2,000 30,000
153,000 22,000
26,000

Energy needed= 351,000 Btu per 40.4 pounds Al 2 0 3 or 17.4 x 106 Btu/ton of Al 2 0 3 produced by sparging. _!!Estimated magnitude for cooling acid to base temperature prior to recovery system.

n-decyl alcohol in 73-percent kerosine, in two mixersettler cells to remove the iron. The iron in the loaded organic phase is stripped with 3-percent HCI in three additional cells before the organic phase is recycled.
The leaching, liquid-solid separation, and solvent extraction sections of the hydrochloric acid miniplant have been operated as an integrated continuous process. The purified aluminum chloride solution produced has
21

been stored for crystallization studies by both the evaporation and sparging techniques.
A review of the basic thermochemical data for the HCI process, incorporating the sparging option, is presented in Table 6. Table 7 presents thermochemical data specific to the evaporation/crystallization option. These data show that the production of 40.4 pounds of Al 2 0 3 by the sparging crystallization tech-

TABLE 7.- Theoretical Energy Data for Evaporation/ Crystallization Option

Process step

Flowsheet materials, lb.

Calcination . . . . . . . . . . . . . . . 143. 7, raw clay 100, calcined clay 43.7, H2 0

Leaching ................ 361.0, 26% HCI 100, calcined clay 461.0, slurry

Theoretical energy requirements, Btu

Needed

Available for reclaim

118,000

35,000 74,000

13,000

Liquid-solid separation .. __ .

461.0, slurry 57.8, wash water
403.2, pregnant solution 115.6, sands

8,000

Solvent extraction . . . . . . . . . 403.2, pregnant solution 401.2, purified solution 2.0, FeCI 3
Evaporation crystallization (double effect) . . . ..... .... 401.2, purified solution 105.1, crystals 258.2, vapors 86.0, crystals 87.0, vapors 8.9, bleed stream 7.5, bleed stream

195,000

4,000 120,000
2,000 96,000
700 350

Decomposition . . . . . . . . . . .

191.1, crystals 150. 7, acid vapors 40.4, AJ 2 0 3

Acid recovery . . . . . . . . . . . . Miscellaneous

221,000

153,000 22,000
25,000

Total . . . . ... ...... .....

40.8 Al 2 0 3 in 40.4 Al 2 0 3 out

Energy needed = 534,000 Btu per 40.4 pounds Al 2 0 3 or 26.5 x 106 Btu/ton of Al 2 0 3 produced.

nique would have a theoretical energy requirement of 351,000 Btu. This is equivalent to 17.4 x 106 Btu/ton of Al 2 0 3 produced . By comparison, to produce 40.4 pounds of Al 2 0 3 by the evaporation/crystallization option would require 534,000 Btu or 26.5 x 106 Btu/ ton of Al 2 0 3 produced. In addition, the sparging technique offers the greatest potential for energy recovery, followed by the evaporation/crystal Iization technique, and then the nitric acid process.
The alumina industry today is principally based
upon the treatment of bauxite by the Bayer process.
Therefore, a similar theoretical energy tabulation

has been developed for the Bayer process (Table 8), which shows an energy requirement of 28.4 x 106
Btu/ton of Al 2 0 3 . However, the processing generates considerable steam, which theoretically can contribute up to 1,087,000 Btu per 100 pounds of Al 2 0 3 produced or 21.6 x 106 Btu / ton of Al 2 0 3 . This would reduce the energy needed to 7.87 x 106 Btu/ton of Al 2 0 3 if all the steam were used. Since the steam cannot be
recovered completely, it is felt that a more realistic comparative value would be about 10 x 106 Btu/ton
of Al 2 0 3 .

22

TABLE 8.- Theoretical Energy Data for Bayer Plant

Theoretical energy requirements, Btu

Process step

Flowsheet materials, lb .

Needed

Available for reclaim

Digestion

. . .. . .. . ... . .. ..

2095.3 caustic 258.2 bauxite
3.9 impure lime
218.6 H2 0 2134.9 digest liquor
351.1 steam

386,000

-397,000

Washing and separation . . . . . . 835.4 wash water 1234.9 digest Iiquor 2772.0 green liquor 198.3 red mud

113,000

-17,000

Precipitation of gibbsite . . . . .

2772.0 green liquor 101.0 wash water
2697.8 mother liquor 175.2 gibbsite

-106,000 -12,000

Mother liquor recovery and evaporation . . . . . . . . . .

2697.8 mother liquor 7.5 soda ash
2095.3 caustic to digestion 61 0.0 water vapor

748,000

-682,000

Gibbsite decomposition . . . . . .

175.2 gibbsite 79.2 stack loss
101.0 impure Al 2 0 3

Limestone calcination . . . . . . .

1.0 limestone 3.5 stack loss 3.91ime

215,000 10,000

-202,000 -4,000

Al 2 0 3 produced . . . . . . . . . . . 100.0

1.472,000

Energy needed@ 1,472,000 Btu per 100 pounds Al 2 0 3 or 29.4 x 106 Btu/ton of Al 2 0 3 produced.

Conclusions
The production of alumina from clay by either of the acid-leaching processes outlined appears to be technically feasible. Two options are available for effecting crystallization within the hydrochloric acid process. Crystallization by evaporation is endothermic, whereas the gas-sparging method is exothermic, and therefore is less energy intensive. A review of the theoretical energy

requirements of each, without regard for engineering energy efficiencies of process steps or the quantity of the energy that is practicable to recover, indicates that the hydrochloric acid process using the gas-sparging option is the most promising. From a standpoint of theoretical energy requirements, the HN03 process is less promising than either of the HCI process options, primarily because 3 more moles of water must be removed during the decomposition step.

23

References
1. Husted, .J. E., 1974, "Potential reserves of domestic non-bauxite sources of aluminum": Presented at the AIME meeting, Dallas, Texas; February, 1974. TMS paper A74-65, 20 pp.
2. National Materials Advisory Board, 1970, "Processes for extracting alumina from nonbauxite ores": Publication NMAB 278, 88 pp.; available from National Technical Information Service, Springfield, Va., PB 198 507.
3. Bureau of Mines, 1975, Minerals in the U. S. economy: ten year supply-demand; Profiles for Minerals and Fuel Commodities. Available from U. S. Government Printing Office, Washington, DC, 1975-603-752/25, p. 5.
24

GEOLOGICAL CLASSIFICATION AND EVALUATION OF HEAVY MINERAL
DEPOSITS
Thomas E. Garner. Jr. E. I. duPont de Nemours and Company
Lawtey, Florida

Introduction
Heavy minerals have been important to the world over the past 50 years, and increasing demand for them makes it essential that new deposits be discovered and developed. Traditionally, heavy minerals have been de fined as discrete, liberated, sand-sized mineral particles with specific gravities over 3.2. They have generally been classed as beach sand deposits because of their associ ation with shore line activity. The heavy mineral suite of most economic deposits consists of the more stable minerals such as ilmenite, rutile, zircon, staurolite, gar net, epidote, aluminum silicates, etc.
The first beach sand mining operation in this country was located at "Mineral City" on the Atlantic coast just south of Jacksonville Beach, Florida. During World War I, rutile was mined for use in producting titanium tetrachloride for tracer bullets. Following the war, this operation became inactive and the site is now the Ponte Vedra Country Club. There was very little activity in heavy minerals during the intervening years until World War II. The Humphreys Mining Company operated a heavy minerals mine just east of Arlington near Jacksonville from 1944 through 1964. Hobart Brothers operated a mine near Vero Beach to produce rutile for their welding rods during the 1946-1963 period. Both of these are inactive at the present time.
When the Arlington deposit was mined out, Humphreys moved their equipment to Folkston, Geor gia, to mine a deposit owned by E. I. du Pont de Nemours & Company. In 1974, the Folkston deposit was mined out and Humphreys moved their wet mill across the St. Marys River to mine a deposit just west of Boulogne, Florida. The wet mill concentrates are trucked to their dry mill at Folkston where they are processed into titanium minerals, zircon, and monazite.
A deposit at the south end of Amelia Island owned by Union Carbide was scheduled for mining in the late 1950's. Plans to mine it were abandoned, and the property was sold in 1970 for development as a recreational area.
Three other heavy mineral mines operate in Florida: two DuPont mines at Starke and Lawtey; and one just south of Green Cove Springs near Penny Farms which is operated by Titanium Enterprise, a joint venture between Union Camp Corporation and American Cyanamid. In addition, heavy mineral mines are operated for ilmenite in New Jersey by the Glidden Company and the American Smelting and Refining Company (ASARCO). The output from ASARCO's plant is pur-

chased by E. I. du Pont de Nemours & Company under long term contract. Other known heavy mineral occur rences include several small deposits in southeast Geor gia, northeast Florida and two large deposits in Tennes see. The Tennessee deposits are owned by Ethyl Corpo ration and Kerr-McGee. None of these are in operation at the present time.
The purpose of this paper is to review geologic history of heavy mineral deposits and discuss economic evaluation of heavy minerals contained therein.
Titanium Minerals
About 95 percent of the United States ilmenite, leucoxene, and rutile production is used to manufacture titanium dioxide white pigments. The remaining 5 percent is used to manufacture titanium metal. The two processes by which white pigments are manufactured from ilmenite and rutile are described by Lynd and Lefond (1975). In general, the higher the Ti02 content in titanium minerals, the more value they have to pigments manufacturers.
In the case of sulfate pigments manufacture, the degree of solubility in sulfuric acid is an important consideration. Ilmenite and altered ilmenite are soluble and can be used successfully in sulfate pigment manufacture. Rutile and high Ti02 leucoxene cannot be used in sulfate pigment manufacturing because they are insoluble in sulfuric acid.
E. I. du Pont de Nemours & Company pioneered the chloride process which can be used successfully with all titanium minerals. Because of the superior pigment quality produced by chlorination of titanium minerals and because there are less pollutants produced by their process, Du Pont has phased out all of their sulfate plants and produces pigments exclusively by chlorination.
Classification of Deposits
Traditionally, titanium mineral deposits have been regarded as either rock or sand deposits (Lynd and Leford, 1975). Hard rock deposits occL; throughout the world producing ilmenite for use in manufacturing white Ti02 pigments. Ilmenite from the Adirondack deposit is used in manufacturing Ti02 pigment by the sulfate process. Ilmenite produced from the Allard Lake deposit is shipped to a smelter at Sorrel, Quebec, where a titanium-rich slag is produced for use in sulfate pigment manufacture.

25

Mineral Magnetite Chromite
Ilmenite Leucoxene

TABLE I - Magnetic Opaque Minerals

Composition

Sp. G.

Color

Distinguishing Optical and Physical Properties

Fe3 0 4

5.2

Black

FeCr20 4

4.6

Black

Ferromagnetic
Chemical analysis may be required to distinguish from ilmenite

FeO Ti02

4.7

Fe20 3 3Ti02 4.3

Fe20 3 4Ti0 2 3.9

Black to dark Chemical analysis may be reddish brown required to identify

Dark red, tan, yellow, white

Light color, waxy luster, some grains translucent

Use Ore of iron Ore of chromium; four.dry sand
Ti0 2 pigment manu facture TiCI 4 Same as ilmenite

Monazite Xenotime

(Ce, La, Y, Th) 5.2 P04

Y P04

5.1

Staurolite

FeAI 4 Si 20 10

3.8

(OH) 2

Garnet Spinel

Complex FeMg 3.8 AI, silicate

ZnAI 20 4

3.6

Epidote Tourmaline

Ca2 (AI, Fe, Mn) 3 3.4

(OH)

Si 0
3

12

complex bora- 3.1 silicate

Mica (Biotite)

Complex alumi- 3.0 num silicate of K, Mg and Fe

Magnetic Non-Opaque Minerals

Yellow to colorless

High relief and index of refraction

Rare earth and thorium ore

Yellow to colorless

High relief and index of refraction ; high birefringence; rei ief changes noticeably on rotation of 90

Rare earth ore

Red

Color and rei ief

Portland cement

manufacture;

sandblast abrasive

Red

Isotropic- red color

Abrasive

and relief

Green to colorless

Isotropic-green color and relief

No commercial value

Green

Green color, particle shape No commercial value and birefringence

Pink to black

Low rei ief, index just below No commercial value oil , pleochoism

Black to golden Uniaxial figure on basal plates

No commercial value

Gold Pyrite

Au FeS 2

Non-Magnetic Opaque Minerals

17

Gold (R)

Soft, malleable, high sp. g. Precious metal ore

5

Brass yellow

Cubic crystal habit

Ore of sulfur

can't.

26

TABLE 1 can't.

Mineral Zircon Rutile
Corundum Anatase Kyanite Diamond S ill im a n i t e Apatite

Non-Magnetic Translucent Minerals

Composition

Sp.G .

Color

Distinguishing Optical and Physical Characteristics Use

ZrSi04

4 .3

Colorless to

High index of refraction

Foundry sand, re-

pink

and relief- shape

fractory, zirconium

metal ore

Ti02

4.2

Red

High relief and index of refraction, birefringent with no interference colors

Ti02 pigment manufacture, welding rod flux

Al 2 0 3

4.0

Blue to

High index of refraction, Refractory

colorless

relief and irregular shape

distinguishes from kyanite

Ti02

3 .9

Colorless to

Rectangular grain shape in Potential Ti0

2

blue

sand deposits

pigment raw material

AI 2 Si0 5
c

3.6

Colorless to

Grain shape and

blue

cleavage

Refractory

3.5

Colorless

Octahedral shape - high

Abrasive

index of refraction

A I2 S i0 5

3.2

Colorless

Matches in bromonapthylene Refractory

cleavage

Ca

5

(F,CI)(PO
4

)3

3.2

Colorless, brown, Index of refraction below

black

liquid. Collophane is

amorphous

Fertilizer mineral

Ilmenite from hard rock source is less attractive than sand ilmenite because Ti02 content is normally too low for economic use in chloride pigment manufacture. An idealized depositional sequence for titanium mineral deposits is shown in Figure 1.
Altered Hard Rock Deposits
Titanium bearing hard rock deposits may undergo extensive alteration, producing new species of titanium minerals substantially different in character from those in the original rock. For example, Brazilian jacupirangites have undergone extensive alteration producing anatase and ilmenite at the expense of the original perovskite. Large octahedral crystals of secondary anatase may be found in the tropically weathered carbonatite near Tapira in Minas Geraes, Brazil. Anatase concentrates assaying up to 86 percent Ti02 can be produced from these deposits, using physical and chemical methods (Lynd and Leford, 1975).

Residual Deposits
Hard rock deposits may be broken down physically by weathering processes without alteration of the ilmenite. In many cases, weathering produces ores which can easily be mined without blasting and do not require extensive grinding, since most of the ore minerals have been liberated. These deposits can be classed as "residual" deposits if the minerals have not been transported. Unless there has been secondary alteration, the Ti02 content in the ilmenite is still relatively low (45 - 48 percent), thus making it unattractive to many pigments manufacturers at the present time.
Sedimentary Deposits
If minerals derived from weathered hard rock sources are transported, they may then be classed as sedimentary deposits. These may be sub-divided into continental alluvial, deltaic, and marine deposits.

27

HARD ROCK

DEPOSITS

r-------------------+------------------1

I

I

I
+ I

f I
WEATHERED

IN SITU

RESIDUAL

ALTERATION

DEPOSITS

I I I
'----------------~~-----------------'

CONT INENTAL
ALLUVIAL
DEPOSITS
T DELTAIC
DEPOSITS
T MARINE
DEPOSITS
r--------------------- ~--------------------- :

STRAND LINE
BEACH SAND
f- ------------ _{ ___ __________ _+

DUNAL DEPOSITS
----------1

OFFSHORE SAND DEPOSITS

I

I

I

1-----------I

FIGURE 1 -Idealized depositional sequence for titanium mineral deposits 28

a) Continental Alluvial Deposits: The liberated and weathered mineral particles
may be transported and redeposited by means of rain water run-off and streams_ During this process the ilmenite grains may undergo alteration through oxidation and leaching to upgrade the Ti02 to 60 percent or higher (Garner, 1971 and Temple, 1966)_ Ilmenite from these deposits may be dull, irregular shaped fragments with a "shaly" structure. Such a grain is shown in Figure 2. The shaly structure may be due to exsolution or to basal parting often noted in ilmenite. Because of this structure, ilmenite from some continental alluvial deposits tends to break down physically producing excessive fines during processing. Alluvial deposits have the advantage that most of the mineral particles are completely liberated sand-sized grains and can be mined using inexpensive methods without crushing and grinding. Continental alluvial deposits are first associated with flood plans. Further reworking and transport downstream may lead to deposition in deltaic sediments. These deposits may be classed as "fresh water deltaic".
b) Marine Deposits:
Heavy minerals from continental alluvial or. deltaic mineral deposits may be carried to the

FIGURE 2 -

Electron micrograph showing ilmenite grain from continental deposit. Note "shaly" character. Grain is about 115 microns across shorter dimension. (ASARCO, Lakehurst, New Jersey).

ocean by rivers and streams where they are deposited in deltas and along coastlines. Reworking of deltaic deposits by waves and tidal currents may redistribute the sand along coastlines.
In some cases, grains may be washed seaward and deposited on the sea floor contiguous to shore lines. The potential importance of these offshore deposits has been recognized and work is underway to investigate the concentration of heavy minerals off the southeast Atlantic coast and in the norther Gulf coast areas. Heavy mineral grains may reside for eons in any of the environments discussed, then be reworked and redeposited many times before being concentrated into commercial deposits. During this process the ilmenite and other grains frequently become rounded with highly polished grain surfaces. Ilmenite particles from the Trail Ridge deposit shown in Figure 3 are typical. For example, heavy mineral bearing Citronelle sands of Miocene age are believed to have been reworked during the Pleistocene (Pirkle and Yoho, 1970). Heavy mineral content of the Citronelle is too low to be considered economic by present day standards. However, during the formation of Trail Ridge, the narrow feature formed at the north end of the Lake Wales erosion remnant, these heavy minerals were redeposited and concentrated

FIGURE 3-

Photomicrograph of ilmenite grains from marine shoreline-dunal deposit. Note grain roundness and surface polish. (E. I. duPont de Nemours & Co., Inc. Trail Ridge Plant, Starke, Florida).

29

to 3 -4 percent heavy mineral in the deposit. Early scientists believed the cross-bedding features noted in the Trail Ridge heavy mineral deposit were foreset bedding of a huge delta. More recent studies indicate the angles of the cross-bedding are those of coastal dunes. The cross-bedding is shown in Figure 4.
Marine heavy mineral beach sands have the advantage that grains are completely liberated, and in many cases, extensive alteration through oxidation and leaching has upgraded ilmenite particles to as high as 90 percent Ti02 . The unconsolidated mineral particles in beach sands are easily mined by low cost methods (dredging, draglining, pan scraper, bucketwheel excavators, etc.). with essentially no crushing or grinding required to liberate mineral particles.
Heavy Mineral Silicates
The heavy mineral assemblage in a deposit depends entirely on the nature of the source rocks. The New Jersey heavy minerals are made up of 80 percent ilmenite with very little zircon (MacKewicz, 1969). Other deposits contain a variety of silicate heavy minerals which were deposited with the titanium minerals. Most of the unstable minerals weather out leaving the more resitant minerals. Younger continental deposits often contain hornblende, augite, and other pyriboles. Older marine beach sand deposits normally shown only traces of pyriboles. The more stable beach sand minerals include epidote, garnet, staurolite, tourmaline, zircon, kyanite, sillimanite, andalusite, corundum, spinel, monazite, and xenotime. Some of these have commercial value, whereas others do not. Heavy mineral uses are shown in Table 1.
Evaluation of Heavy Mineral Samples
Exploration techniques for finding heavy mineral deposits are outside the scope of this paper. However, a general outline for determining heavy mineral composition and economic potential of prospect samples is discussed below. No single method can always be used since samples may be different depending on the origin of the deposit.
Preliminary Examination
After drying at 100C, samples received for evaluation should be given a preliminary examination to determine best laboratory procedure. If the sample contains aggregates it should be crushed to liberation size. If the sample contains plus-28 mesh material, this should be removed by dry screening, the percentage determined and recorded for later use. If the sample is granular, free flowing, without significant amounts of clay, the evaluation continues with a heavy mineral determination using heavy liquids discussed below. If the sample contains significant amounts of caly, it should be slurried and wet screened on a 325 or 400 mesh sieve .

Because of the very fine wire used in 400 mesh sieves, it is extremely important that care be taken not to break the wires while screening. Small amounts of slurry (no more than 300 grams of sand at a time) are introduced onto the 400 mesh screen. Water is used to wash the fines through the screen. Gentle rubbing in circular motion on the bottom of the screen may facilitate screening. The plus400 mesh is transferred to a pan, dewatered, dried, then weighed, and the percent of sample determined.
Heavy Mineral Separation

The minus-28 plus-400 mesh fraction should be riffled into a 100 gram sample using a Jones type splitter. The sample is poured into a pear-shaped, 500 ml separatory funnel filled with acetylene tetrabromide which has specific gravity 2.96. If the minerals are extremely fine, bromoform (specific gravity 2.87) may be necessary for complete heavy mineral recovery. The sample and heavy liquid are stirred together using a stirring rod and the heavy minerals allowed to settle. The separatory funnel is agitated and stirred to make sure that all heavy minerals which may have been entrapped in the float are released and allowed to sink. This is repeated several times. The sink fraction is drawn

FIGURE 4-

Truncated dune bedding in marine shore-dunal deposit. Section is about 20 feet. Heavy minerals are more abundant in the cross-bedded portion. (E. I. duPont de Nemours & Co., Inc., Trail Ridge Plant, Starke, Florida).

30

SAMPLE AS RECEIVED

OVEN

CRUSH TO
LIBER'ATION

Preliminary Examination (Hand Lens-Binocular microscope)
l----- ------, - -- -A---- -- ---- -- --------- ~

1----------------------~

I

I

plus-28 Mesh

B

I

SCREEN

I

I

4--------~

minus-400 mesh

Kaolin Montmorillonite

LIGHTS Quartz 2.6

-

Feldspar 2.6

Mica

Glauconite 2.7

DES LIME
HEAVY LIQUID SEPARATION

c
I l l
I
~-- _!

FERROMAGNETICS

PERMANENT MAGNET
~ Non-Mags

+
MAGNETICS
Petrographic and Chemical Analyses

HIGH INTENSITY INDUCED
ROLL MAGNET
1
NON-MAGNETICS Petrographic and Chemical Analyses
A - Use if rocks or aggregates present. B - Use if no coarse particles noted. C - Use if no clays or coarse particles in sample .

FIGURE 5- General outline for laboratory evaluation of heavy mineral prospect samples.

31

off into a funnel lined with Whatman #4 filter paper. The separatory funnel is again stirred and the heavy minerals allowed to settle to the bottom. The second crop is drawn off after the solution between the heavy minerals has cleared. (If the sample contains collophane or other minerals with specific gravity near that of the liquid, they may suspend.) This procedure is continued until no more heavy minerals are detected. The float is collected in a separate funnel lined with filter paper. After the heavy liquid is recovered, sinks and floats are washed using either alcohol or Solox. (Solox is a trade name of U.S. Industrial Chemicals, Inc.) The sinks and floats are rewashed several times to completely remove all of the heavy liquid. Both products are then dried in an oven at a temperature of 100C, weighed, and their percentages calculated. It should be noted that both acetylene tetrabromide and bromoform are toxic. Care should be taken to avoid skin contact and the separation should be carried out under a hood.
Identification and Evaluation of Heavy Minerals
Heavy mineral content of a sample may range from 0 to 100 percent. The value of the heavy minerals depends on their mineral composition. Mineral composition may be determined by petrographic analysis using bromonapthalene or other oil with index of refraction of 1.66 as an immersion fluid. This analysis may be supplemented with chemical analysis, x-ray diffraction, scanning electron microscope, electron probe, etc. If the sample contains minerals of potential economic value such as ilmenite, leucoxene, rutile, staurolite, zircon, monazite, xenotime, garnet, further evaluation is warranted. If the heavy minerals contained only noneconomic heavy minerals with no value such as hornblende, augite, epidote, further examination is probably unwarranted.
Detailed Laboratory Analysis of Heavy Minerals
If the optical examination shows significant amounts of opaques (potential titanium minerals), zircon, staurolite, monazite, etc.), further laboratory analysis should be undertaken using the method outlined in Figure 5.
Ferromagnetic Minerals
Magnetic separation using a permanent horseshoe magnet removes magnetite and titaniferous magnetite. Silicates with magnetic inclusions may also be removed. Magnetite and titaniferous magnetite both have potential markets. The magnetics are weighed and recorded as percent of heavy minerals.
Paramagnetic Minera Is
The non-ferromagnetic fraction may be further separated into paramagnetic and non-magnetic using high

intensity laboratory magnet. Products from this separation should be examined petrographically to determine composition. The paramagnetic fraction may be made up of the opaque or non-opaque minerals shown in Table 1. The paramagnetic opaques may be high grade Ti02 ilmenite type minerals. To determine this, analyze the fraction for Ti02 , FeO, and Fe2 0 3 and estimate Ti02 in ilmenite as shown below:
Estimated Ti0 2 in ilmenite
If the estimated Ti02 is less than 55 percent, it has only marginal value in the present day market. If it is 60 percent or higher, it has good potential value.
Staurolite by itself has potential economic value as a sand blast abrasive or constituent in Portland cement manufacture. In the presence of low AI 20 3 garnet and epidote, staurolite has not value to the cement industry which specifies 45 percent Al 2 0 3 or more. On the other hand, garnet by itself has potential value as an abrasive. In the presence of epidote, this value is minimized. At the present time epidote alone has no market value. Monazite is used to manufacture rare earth oxides and thorium.
Non-Magnetic Minerals
In general, the non-magnetic fraction will contain the more valuable minerals rutile, zircon, leucoxene, etc. The composition of the non-magnetic minerals may be determined by mineral count or chemical analysis. Rutile and leucoxene have value to Ti02 manufacturers. Zircon is used as a foundry sand, as an opacifier, and to manufacture zirconium metal.
Preliminary Economic Evaluation
The value of a heavy mineral suite can be estimated by multiplying the percent of each heavy mineral constitutents by its present market value. These values can be found in "Engineering and Mining Journal" and "Industrial Minerals" magazines. If there is sufficient value indicated for the heavy mineral suite and there are enough heavy minerals in the original sample to warrant, then more detailed testing should be undertaken to investigate the separation characteristics of the heavy minerals. Examples are shown in Table 2.
Determination of Separation Characteristics
A large sample representing the prospective ore is obtained using composite drill samples, pitting, or trenching. The sample size should be large enough to furnish at least five pounds of heavy mineral concentrate. Using the laboratory procedure outlined in Figure

32

DRY SAMPLE

FIGURE 6

+28 Mesh
SLIMES CLAYS
TAILINGS Quartz, Feldspar and micas

A ------------ ------ .-,.1.,

o I B~
CRUSH TO LIBERATION 1 1

I I

j_

I 1

"~-------- 1

I

SCREEN

c

@ 28 MESH

0

t .--------~

DES LIME

+ -------...,

I

LABORATORY

+ _ ____, SHAKING TABLE
.L.----.--

Middling

FERROMAGN ETICS 4- ------
Magnetite, titaniferous magnetite
Magnetics

A . HEAVY MINERAL

CONCENTRATE

B .

t

DRY

c

- -- ------- ,

PERMANENT

D D -

M AG NET

I

+ -------~

Use if rocks or aggregates are present. Use if no aggregate particles are present. Use if no coarse particles are present. Use if no ferromagnetic particles are present.

LABORATORY HIGH INTENSITY
MAGNET

Non -Magnetics

LABORATORY HIGH TENSION
PARAMAGNETIC CONDUCTORS
Ilmenite Leucoxene Chromite

Magnetic Non-Conductors
l

LABORATORY HIGH TENSION

~ Non-Magnetic Non -Conductors
I

VANNING PLAQUE OR TABLE

NON-MAGNETIC CONDUCTORS

VANNING PLAQUE OR TABLE

HEAVIES

LIGHTS

Leucoxen~-Rutile
Pyrite, Gold

+
HEAVIES

f
LIGHTS

Monazite Xenotime

Garnet, Epidote, Staurolite, Spinel

Zircon, Corundum

Kyanite, Sillimanite,

Glauconite

Tourmaline,

Apatite

FIGURE 6 Outline for laboratory determination of mineral separation characteristics.
33

TABLE 2- Examples of Heavy Mineral Value Calculations (Hypothetical Cases)

Potentially Economic Ore:
Ilmenite @60% Ti02 Leucoxene @ 87% Ti02 Rutile Zircon Staurolite Monazite Garnet-Epidote Kyanite-Sillimanite

%of H. M.

X

45 5 2 5
20 1
15 7

Current Market Value
$ 29* 213* 313* 175* 5** 219*

100%

4% H. M. in ground x $41.90 = $1.68 value per ton of ore in ground .

Marginal or Uneconomic Ore:

Magnetite Ilmenite @54% Ti02 Leucoxene Rutile Zircon Pyriboles Kyanite-Sillimanite

2%

$ 18**

20

19*

213*

2

175*

60

15

100%

10% H. M. in ground x $9.79 = $0.98 value per ton of ore in ground.

Heavy Mineral Value
$13.05 10.65 6 .26 8.75 1.00 2.19
$41.90
$ 0.36 3.80 2.13 3.50
$ 9.79

* "Industrial Minerals" magazine No. 100, January, 1976. **Estimated

6, the heavy minerals are concentrated on a laboratorysize shaking table. The table concentrate is driend, then split magnetically at 7,000 gauss to separate the ilmenite. The magnetic portion is passed over a laboratory high tension unit and the conductor fraction recleaned. The final conductor fraction is examined under the microscope, and if it contains 98 percent or more opaques, that is a good indication that mineral particles are free from surface coatings. If there are large amour'lts of silicates in the conductors, this may indicate presence of surface coatings which could interfere with dry milling. Tests should then be undertaken to determine what

can be done to remove these surface coatings. In some cases high energy attritioning alone may do it. Some ores will require chemical treatment. Most acids have been used with some degree of success to remove coatings. Hydrofluoric acid has been very successful in removing surface coatings; however, it is extremely hazardout and should be considered only if all others fail. NaOH may be useful in removing clays and organic surface coatings.
The non-magnetic fraction is run through the same series of tests. The presence of surface coatings, particularly on zircon, can be detected by calcining the

34

non-magnetic non-conductors in a laboratory muffle furnace at 1000C for 30 minutes. This treatment produces a bright orange color on grains with extensive surface coatings. It is important to have clean grain surfaces because surface coatings on zircon grains may not only effect their dry milling characteristics, but may also make the product unacceptable to the foundry industry where wettability and coatability are essential. Zircon is separated from lighter aluminum silicates using a shaking table if there is enough sample. If there is not enough sample, zircon can be upgraded with a vanning plaque.
Particle Size Analysis
Industries using the heavy mineral products usually specify the particle size distribution desired. Therefore, each product from the laboratory testing should be screened. In the event that none of the laboratory products are pure, then screen the entire sample, do petrographic analysis on each size fraction, and then calculate the size distribution for each mineral in the composite sample.
About 85 percent of the zircon produced goes into the foundry industry. Foundrymen prefer rounded grains with size range of 80 to 150 mesh. Du Pont zircon

FIGURE 7-

Photomicrograph showing zircon grains from marine shoreline-dunal deposits. Note grain roundness. Grain surfaces show frosting due to wind action be possibly chemical solution. (E. I. duPont de Nemours & Co., Inc., Trail Ridge Plant, Starke, Florida).

FIGURE 8-

Photomicrograph showing zircon grains from continental deposits. Note angularity and lack of surface rounding on most grains. (ASARCO, lakehurst, New Jersey).

produced at Starke, Florida, is shown in Figure 7. A sample of zircon from a continental deposit is shown in Figure 8 for comparison.
Finer zircon (say - 200 mesh) cannot be used in most foundry sand applications. If zircon quality is good enough, it may be used in manufacturing zirconium metal.

Final Evaluation
Decision to do further work'leading to feasibility studies and economic evaluations will depend on the following:
laboratory work must show sufficient valuable minerals to be economic and no milling problems detected.
Deposit must be located in favorable location with respect to potential markets.
There must be roads, railroads, and labor supply.
Environmental impact. land use after mining.

35

Summary and Conclusions
This paper shows: Heavy mineral deposits have extremely com-
plex geologic histories. Mineral samples vary in character and composi-
tion and reflect the type of deposit from which they were derived. The value of a heavy mineral suite may depend more on composition than on concentration in the ground. Extensive laboratory work is required before potential value of an ore can be accurately determined. Perhaps the most important points are: No single laboratory method will fit all samples and no single set of rules can be set forth to accurately evaluate all heavy mineral ore deposits. Each deposit is different and each must be evaluated on its own merits. There is no doubt that demand for heavy mineral products will continue through the foreseeable future, in fact demand will probably increase over the next two decades. To meet these needs the industry must prepare to meet increasing labor and energy costs with new, innovative methods to more cheaply mine and separate heavy minerals, and be prepared to restore land to 100 percent usefulness and leave no hazardous conditions for future generations.
References
Bailey, S. W., Cameron, E. N ., Spedden , H. R., and Weege, R. J., 1956, The alteration of ilmenite in beach sands : Econ. Geology, Vol. 51, p. 263-279.
Bishop, E. W., 1956, Geology and ground-water resources of Highlands County , Florida: Fla. Geol. Survey, Rept. of Invest. 15.
Brooks, H. K., 1966, Geological history of the Suwannee River in geology of the Miocene and Pliocene series in the north Florida-south Georgia area, N. K. Olson, Ed.: Atlantic Coastal Plain Geol. Soc., 12th Field Trip Guidebook, p. 37-45.
Cannon, H. B., 1950, Economic minerals in the beach sands of the southeastern United States : Symposium on mineral resources of the southeastern U . S., Univ. of Tenn. Press, p. 202-210.
Creitz, E. E., and McVay, T . N., 1948, A study of opaque minerals in Trail Ridge, Florida dune sands: AIME Tech. Pub. No. 2426, p. 1-7.
Garnar, T. E., Jr., Geological review of some North Florida resources, Southeastern Geological Society, 15th Field Conference Guidebook, Tallahassee, Florida, p. 26-33.

Garnar, T. E., Jr., 1971, Economic geology of Florida heavy minerals : Proceedings of 7th Annual Forum on the Geology of Industrial Minerals, Box 631 , Tallahassee, Florida.
Garnar, T. E., Jr., 1973, Heavy Mineral Separation: Light Metals, AIM E.
Gillson, J. L., 1960, Industrial Minerals and rocks : AI ME, New York.
Gillson, J. L., 1959, Sand deposits of titanium minerals: Mining Engineering, Vol 11, p. 421-429.
Grogan, R. M., Few, W. G., Garnar; T . E., and Hager, C. R., 1964, Milling at DuPont's heavy mineral mines in Florida : Milling Methods in the Americas, N. Arbiter, Ed.: VII International Mineral Processing Congress, N.Y., Gordon and Breach Science Publishers, p. 205-229
Hammoud, N. S., 1975, A process for recovering low-chromium high-grade. ilmenite from North Egyptian beach deposits: Proceedings 11th International Mineral Processing Congress, special volume, p. 315-336.
Lynd, L . E., Sigurdson, H ., North, C. H., and Anderson, W. W., 1954, Characteristics of titaniferous concentrates: Mining Engineering, p. 817-824 .
Lynd, L. E., 1960, Alteration of ilmenite : Econ . Geology, Vol. 55, p. 1064-1068.
Lynd, L. E., 1960 Study of the mechanism and rate of ilmenite weathering, AIME Trans., Vol. 217, p. 311-318.
Lynd, L. E., and Lefond, S. J., 1975, Titanium Minerals: Industrial Minerals and Rocks, 4th ed ., AIM E. New York.
Mackewicz, F. J., 1969, Ilmenite deposits of the New Jersey coastal plain : Geology of Selected Areas of New Jersey and Eastern Pennsylvania and Guid ebook of Excursions, S. Subitzky, ed ., Rutgers Univ. Press, p. 363-382.
Mackey, Thomas S., 1972, Alteration and recovery of ilmenite and rutile: Australian Mining, ~Jovember, 1972, p. 18-44.
Mackey, Thomas S., 1974, Selective leaching of ilmenite to produce a synthetic rutile: Light Metals, AIM E.
Pirkle, E. C., Yoho, W. H., 1970, The heavy mineral ore body of Trail Ridge, Florida : Econ. Geology, Vol. 65, No. 1, p. 17-30.
Pirkle, E. C., Pirkle, W. A., Yoho, W. H ., The Green Cove Springs and Boulogne heavy mineral sand deposits of Florida: Econ. Geology, Vol. 69, No.7, p. 1129-1137.
Swanson, V. E. , Palacas, J. G., 1965, Humate in coastal sands of northwest Florida: Geol. Survey Bulletin 1214-B.
Temple, A . K., 1966, Alteration of ilmenite: Econ. Geology, Vol. 61, p. 695-714 .

36

RECONNAISANCE INVESTIGATIONS OF OFFSHORE PHOSPHATE DEPOSITS OF
GEORGIA AND SOUTH CAROLINA
James L. Harding Marine Resources Center
John E. Noakes Geochronology Laboratory
University of Georgia Athens, Georgia

Introduction
An important mineral deposit consisting of bedded marine phosphorite is known to exist in the subsurface strata underlying the marshlands, estuaries and barrier islands of coastal Georgia. As it is centered in the eastern portion of Chatham County (which is the northernmost of Georgia's coastal counties). this deposit is usually re ferred to as the Chatham County Phosphate Deposit. The stratigraphy and economic geology of the deposit were presented by Furlow (1969) in a comprehensive study conducted by the Georgia Department of Mines, Mining and Geology. The objectives of his study were: 1) to determine if the mining of phosphate in the eastern portion of the county would jeopardize the fresh water aquifer which underlies the phosphate-bearing zone and 2) determinations of the depth, grade, volume, value, etc. of the ore.
Although Furlow concluded that mining done under strict supervision of regulatory agencies would not imperil the integrity of the aquifer, the proposed exploitation of these deposits led to an uproar on behalf of the environmentalists so the plans for mining were shelved. While such plans were still viable, interested parties drilled well over one hundred exploratory holes in the general area of interest. Data from these holes indicated that the phosphate-bearing strata (matrix) might extend several kilometers seaward of the Georgia coast.
In 1962, two test holes were drilled 16 kilometers (10 miles) offshore of Savannah Beach for engineering purposes coincident to the emplacement of a light tower designed to replace the Savannah lightship. The holes penetrated strata varying in age from Recent to late Eocene (McCollum and Herrick, 1964). These au thors found a continuation of the subsurface stratigra phic sequence that exists onshore due west of the tower site.
According to Pickering (1976) re-examination of the Miocene section of cuttings and cores from the off shore test holes, now retained by the State of Georgia, showed the presence of the Duplin Marl, the same for mation which contains the phosphate matrix onshore beneath eastern Chatham County. A columnar section drill log prepared by Pickering showed the seafloor at
the tower test hole #1 site to be about 17.5 meters (58 feet) below mean sea level, with the Duplin For

mation being overlain by 0.9 to 1.2 meters (3 to 4 feet) of undifferentiated Recent to Pleistocene sands. The sands contained accessory quantities of shell hash and brown or black phosphate pebbles. The seafloor here is well within wave-base range, so that these few meters of unconsolidated coarse sands probably suffered con stant re-working.
Beneath the sands, Pickering showed about 7.3 meters (24 feet) of the phosphate-bearing Duplin Marl Formation. Whereas the Chatham County deposit as described by Furlow (op. cit.) was overlain by up to 15.2 meters (50 feet) of nonorebearing Duplin Marl, the barren portion of this upper Miocene formation was non-existent in the area of the light tower, apparent ly having been removed by erosion.
Investigative Methods
The reconnaissance investigation described in this paper was based in large part on extrapolations from onshore data and the results of the analyses made on the drill hole samples from the tower site.
In 1970, the investigators began to develop appli cations of nuclear technology for adaptation to prob lems involving marine mineral exploration. The initial efforts (197073) were aimed primarily at evaluating the use of neutron activation analysis, first on the seafloor using a manned submersible and then on shipboard through measurements made on cores and water sam pies. The results of these research efforts were described by Noakes and Harding (1971) and Noakes, et al (1972).
During the work with the neutron activation anal ysis systems, new tools were developed for radiation detection in the marine environment. Although the work had involved the utilization of a man-made isotope, Californium 252 (252 Cf), it became apparent that the same tools could be modified to detect any naturally occurring radioactivity associated with seafloor sediments.
Previous studies, summarized by Emery and Uchupi (1972), indicated that the most common occurences of natural radiation in marine sediments were associated with heavy mineral and phosphorite deposits. Heavy minerals such as monazite, zircon, sphene and epidote may contain thorium and its decay daughters, whereas phosphate deposits may contain uranium and its decay products.

37

The potential of a large deposit of phosphate situated in dredgeable depths within continental waters was of far greater interest and import than was the proposed mining beneath the marshlands of the mid- andlate 1960's. It was therefore decided to focus research efforts toward the development of a system to detect and differentiate between the gamma activity associated with any heavy minerals and/or phosphate in the surficial sediments of Georgia's continental shelf.
Instrumentation
The radiation survey system consisted of 1) a towable underwater sled and 2) a static radiation analysis device. The sled (Figure 1) was constructed of heavy stainless steel and contained four gamma-ray detectors consisting of 7.6-centimeter (three by three-inch) thai Iium-activated sodium iodide Na I(T 1) crystals. It was similar to the one developed by Summerhayes, et. al. (1970) for marine phosphorite prospecting. Also, Bowie and Clayton (1972) and Miller and Symons (1973) developed and tested parallel systems. Preamplifiers and high voltage electronics encapsulated within the sled transmitted electronic pulses via coaxial cable to the surface vessel. There a dual channel recorder traced the gamma activity in terms of background and variations from background (anomalies) while the vessel was underway.
The static detector was designed to rest directly on the seafloor and consisted of a tripod arrangement housing a cylinder which contained a single large lithium-drifted germanium Ge(Li) detection crystal (Figure 2). The Ge(Li) detector was maintained at an operating temperature of minus 180 degrees Centigrade by a cryogenic coolant.
The static detector was a specific isotope detection system, which by obtaining spectrum analyses on a multi-channel analyzer and plotter, could differentiate between uranium and thorium and their respective decay products.
Operational Procedures
The radiation detection sled was towed astern the survey vessel with 2.2 centimeter (7 /8 inch) stainless steel cable as tow cable, to which the coaxial signal cable was taped for the first 15 to 30 meters (50 to 100 feet) underwater. At the onset of the program, while still in the testing phase, tows were restricted to the intracoastal waterway and the various sounds along the Georgia coast. It was then taken into the open waters of the continental shelf where parallel traverses approximating the length of the Georgia coast were then made with the sled in tow (Figure 3). Although much of the work was weather-dependent as the research vessel Kit Jones was a 17 meter (57-footer), the two speeds averaged between three and six knots, depending upon sea state and bottom topography.
Horizontal positioning was accomplished by a combination of Loran-A fixes and radar bearings. Time.

FIGURE 1

Gamma-ray detection sled, housing four Na I (TL) detectors, preamplifiers and associated electronics.

FIGURE 2-

Static underwater gamma detection system. Base of unit must be in direct contact with seafloor to register readings for spectrum analysis.

38

0 0 0
FIGURE 3- Original traverse lines in Georgia coastal waters.
Fl GU RE 4- Radiometric readings recorded by shipboard electronics from towable underwater sled.

compass bearing and Loran fix positions were marked simultaneously on the dual-channel ship recorder (Figure 4) and on the ship's depth-recorder charts. This was done as a matter of course every ten minutes. When anomalies were apparent on the sled-recorder, additional fixes, etc., were logged.
The original intent was to complete the series of tow traverses, keeping track of all anomalies and then return to such sites to utilize the static, specific isotope detector. This latter unit required that the vessel anchor at each site, because the electronics were such that between 30 and 45 minutes of direct bottom contact were required in order to record ample spectra data. The R!V Kit Jones carried only one anchor and after several frustrating attempts, it became clear that at least a two-point moor was necessary to achieve stabilization of the unit with regards to contact with the seafloor. Its use was therefore abandoned in favor of obtaining samples and doing the spectrum analyses in the laboratory.
Toward the end of the field phase of the investigation, a high-resolution acoustic seismic profiler unit was towed in conjunction with the sled. Both were towed by the surface vessel moving at a speed of about five knots. A sample of a seismic record thus obtained is shown in Figure 5.
Results
The original offshore traverses paralleled the coast and extended from abeam Mayport, Florida (just north of Jacksonville) to Port Royal Sound, north of Hilton Head Island, South Carolina. These lines were conducted in water depths ranging from 3 to 30 meters (10 to 100 feet).
As the work progressed, a definite geographic zonation of gross anomalies appeared, wherein the southern portion of the survey area (south of Sapelo Island) had but one large anomalous area, located directly offshore of the southern half of Cumberland Island. Subsequent sampling indicated that this one anomalous area was due to localized concentrations of heavy minerals in the surface sands. The middle portion of the survey area (between Sapelo and Wassaw Islands) was essentially barren with respect to anomalies, whereas the readings in the northern portion (between Tybee and Hilton Head Islands) were generally the highest encountered. The survey line direction, therefore, was changed to one which ran essentially east-west, in an attempt to grid the area, using the position of the light tower as a control point. The results seemed to indicate a slight increase in above-background readings towards the northeast. The final grid pattern is that illustrated in Figure 6, which also utilized the naviga:mal tower as a reference point.
During the running of the grid pattern shown in Figure 6, the high-resolution sub-bottom profiling unit and the gamma detector sled were being towed from the surface vessel at a speed of about five knots. Radioactive anomalies located by the sled detector could be readily correlated with undulations in the bottom and near

39

Line No, E-3 Bearing 151

vertical, however, and this can only be done by a coring program.
The work done to date was conducted under the auspices of the Georgia Sea Grant Program, and the investigators were limited to using vessels within the Georgia Marine Program, none of which are capacle of the type of coring required. Any further work should utilize a combination of tools consisting of (1) high resolution seismic profiling which should be either a mini-sparker or a Uniboom, so that at least 30 meters (100 feet) of data can be obtained on the sub-bottom strata; and (2) precise horizontal positioning, accurate and reproducible to within plus or minus one meter; (3) side-scan sonar for mapping the micro-bathymetry; (4) the gamma detector sled system and (5) a survey vessel capable of at least a two-point anchoring system in the case of vibracores or a drill barge with support tug for deployment of a three or four point anchor spread in the case of a rotary rig. The detailed sub-bottom profiling should be done first, in order to reduce the number of drill hole sites necessary to define the prospect area.
Summary

FIGURE 5- Sub-Bottom profiler (Eiac-Echograph Laz-17) record of portion of line E-3.

It should be reiterated that the gamma detection system is a reconnaissance tool and is only capable of detecting radioactive material in the surface sediments. The sub-bottom profiler, where used, gave the only clues as to the composition and altitude of the subsurface

bottom strata as shown on the profile recorder. In places, where a sub-bottom reflector would rise to intersect the seafloor, an anomaly would also appear on the shipboard instrument obtaining signals from the sled. Figure 5 shows the sub-bottom profile record taken on line E-3. There are between 1.5 and 6 meters (5 and 20 feet) of fairly well-stratified unconsolidated sediments lying atop a horizon which is acoustically opaque at the frequencies employed. Undulations in the sub-bottom strata are also apparent. It is possible that this opaque surface represents the top of the Miocene which lies a few feet below the seafloor at the tower site according to Pickering (1976).
Grab samples taken in the grid area were analyzed for P2 0 5 and percent BPL, the results of which are given in Table I. As these samples represent only the top several inches of the seafloor sediments, any concentrations of phosphatic material could merely reflect wavebase concentrations.
Suggestions for Future Work
The gamma detection sled and the static spectrum analysis system described herein are applicable for gross reconnaissance work only, because their detection limits are confined to the upper few inches of seafloor material. By themselves, they are excellent guides; when combined with high resolution seismic profiling, the system improves. Additional substantiation is needed in the

-~"' FdJ SOUND . 3ori

~~~t l

W; lIJWO.
--L.

SAVANNAH

'

TOWER

Nculiccl Miles

0

10

60.ft

FIGURE 6-

Final grid pattern run off Georgia-South Carolina coast, using tower location as a control point.

40

Tybee Island

Mean Sea Level

Area of Interest

CH-10

Diagrammatic Sketch Not To Scale

Savannah Tower

FIGURE 7 - Diagrammatic cross-section showing gross relationships between matrix zone and shelf strata.
FT.

FIGURE 8- Diagrammatic fence diagram showing gross relations of near sub-surface strata.
SAVANNAH

Based on Drill Holes and Reflection Seismic Data
after Woolsey + Henry (1974)
41

~ Holocene ~ Pliocene k>.'./~ Miocene

Sample No.

Location

TABLE I
Phosphate Analysis of Grab Samples %P205

%BPL

* 3 * 5
8 11 17 22

Lat. 31 51' 45" N Long. 80 52' 00" W
Lat. 32 05' 00" N Long. 80 34' 10" W
Lat. 32 04' 33" N Long. 80 25' 30" W
Lat. 31 40' 20" N Long. 80 54' 00" W
Lat. 31 12' 30" N Long. 81 04' 45" W
Lat. 31 00' 00" N Long. 81 12' 30" W
Lat. 31 43' 10" N Long. 80 59' 00" W

0 . 15

0.327

0 .57

1.24

1.65

3.60

1.19

2.60

0.12

0.26

0.70

1.53

1.64

3.58

Vanomolybdate colorimetric method used.
*Sample stations are in Grid area depicted in Figure 6.

strata. Nonetheless, the authors feel that because the in vestigation showed a definite geographic pattern to at least surficial concentrations of phosphatic material, and that pattern coincided with the data obtained from the tower borings, the inner Continental Shelf off Chatham County, Georgia, and the adjacent counties in South Carolina may contain a potentially economic deposit of phosphate within dredgeable depths. According to Pickering (1976), the matrix (ore-bearing) section of the Duplin Marl is 7.3 meters (24 feet) thick at the tower site where the BP L content ranges from 19 to 40 percent.
Vertical data are insufficient to identify theregional dip the strata involved, but it is reasonable to assume that somewhere to the east of the tower site, the seaward slope of the shelf causes the Miocene section to thin (see Figures 7 and 8). It is also possible that a significant portion of the ore-bearing section has been removed by erosion.
In order to fully evaluate the ore potential of the offshore deposits, information on the thickness of the matrix, overburden characteristics, and the relationship

to the underlying strata must be obtained. The former two are imperative for design of the recovery system whereas the latter must be firmly established in order to guarantee the integrity of the aquaclude and the principal aquifer beneath.
Coincident to any additional prospecting, information should also be gathered on the current regime, water mass movement and water chemistry in the general prospect area, as such data will be necessary for the preparation of the environmental impact statement certain to be required by the regulatory agencies prior to the issuiance of any leasing, etc. Portions of the area are currently under study by scientists of the Skidaway Institute of Oceanography in Savannah under the auspices of ERDA, and some information will therefore be available for baseline purposes. Although the offshore area is far less critical to the ecological balance of the species which are known to spawn and spend a portion of their juvenile life-cycles in the estuaries and marshdominated tidal creeks, the effects of suspended sediments and detritus stirred up by dredging operations must be established in advance.

42

Acknowledgements
This work was supported with funds from the Georgia Sea Grant Program (U.S. Department of Commerce, NOAA). Shiptime was furnished by the National Science Foundation, Grant No. NSF-GO 31558 (University of Georgia).
References Cited
Bowie, S. H. Y. and Clayton, C. G. 1972, Gamma spectrometer for sea or lake-bottom surveying: Trans. lnst. Min. Metal!. Sect. B, Appl. Earth Sci., vol. 81, p, 251-256.
Emery, K. 0. and Uchupi, E., 1972, Western Atlantic Ocean: topography, rocks, structure, water, life and sediments: Am. Asso. Petroleum Geologists, Memoir No. 17, 379 p.
Furlow, J. W., 1969, Stratigraphy and economic geology of the eastern Chatham County Phosphate Deposit: Bull. No. 82, Ga. Dept. Mines, Mining and Geology, Atlanta.
McCollum, M. J. and Herrick, S.M., 1964, Offshore extension of the Upper Eocene to Recent stratigraphic sequence in southeastern Georgia: U.S. Geol. Survey Prof. Paper 501-C, pp. C61-C63.
Miller, J. M. and Symons, G. D., 1973, Radiometric travers of the seabed off the Yorkshire coast: Nature, vol. 242, pp. 184-186.
Noakes, J. E., Harding, J. L. and Spaulding, J.D., 1972, Californium 252 as a new oceanographic tool: preprints, 8th Ann. Cont. Mar. Tech. Soc., Sept. 11-12, Washington, D. C. pp. 415-425.
Pickering, S. M., Jr., 1976, Occurrence of Miocene Phosphorite in samples from offshore test holes east of Chatham County, Georgia: unpublished manuscript.
Summerhayes, C. P., Hazelhoff-Roelfzama, B. H., Tooms, J. S. and Smith, D. B., 1970, Phosphorite prospecting using a submersible scintillation counter: Econ. Geology, vol. 65, pp. 718-723.
Woolsey, J. R. and Henry, V. J., 1974, Shallow, high resolution seismic investigations of the Georgia coast and inner continental shelf: in Symposium on the Petroleum Geology of the Georgia Coastal Plain, Bull. 87, Georgia Dept. Nat. Res., Earth and Water Div., Atlanta. pp. 167-187.
43

DEVELOPMENT OF HIGH EXTRACTION MAGNETIC FILTRATION BY THE KAOLIN INDUSTRY OF GEORGIA
Joseph lannicelli Aquafine Corporation
Brunswick, Georgia

Status Of Development
Between 1973 and 1975, each of the 5 major kaolin producers in Georgia put hugh 20 kilogauss filters into service. These filters are unique as to size, power, capacity, and novelty of design.
Each magnetic filter weighs 250 tons and can process its own weight of kaolin slurry each hour when operated at the maximum throughput (20 seconds retention time). Under these conditions, flow rates are about 5425 liters (1,400 gallons) per minute or about 100 tons per hour on a dry basis.
These five installations represent a total invest ment of over 8 million dollars (1975 dollars) and can process nearly all of the waterwashed kaolin in the United States if operated at maximum throughput. This commercialization constitutes the most rapid acceptance of novel magnetic separation technology by a mining industry in recent times.
Full scale commercial use of magnetic separation is the culmination of a 10 year effort almost entirely supported by the kaolin industry. 1! is the only commercial use of this technology at this time. However, the experience gained in the kaolin industry is now a good springboard for practical applications of this process in other industries.
Evolution of the Process
High extraction magnetic filtration was carried from initial research to pilot production in 1968 after a 5 year research and development effort on kaolin. During this period the importance of 3 key concepts were identified and successfully combined into a practical prototype separator.
An early breakthrough in HEMF was the finding that separations on micron and submicron kaolin suspensions required very slow transit velocities and maximum exposure to collecting elements to allow capture of paramagnetic particles.
To achieve particle capture, the magnetic attraction of a collecting element must overcome viscous drag of the transporting fluid. Therefore, it was necessary to operate at much lower velocities (Jess than 2 em/sec.) than had ever been employed in practical separations.
Disciminating control of slurry exposure in the collecting zone of a magnetic filter is best expressed by the concept of retention time. It is the subject of a basic process patent on kaolin.

Evolution of Small Scale HEMF Equipment
To exploit the concept of retention time, it was necessary to devise a new magnetic separator which for the first time combined high gradients of the Frantz Separator (Figure 1) with high intensity fields of the Jones Separator. (Figure 2). This critical fusion of separator technology was achieved in 1967 when J. lannicelli constructed a hybrid separator (Figure 3) utilizing a canister filled with high gradient Frantz screens installed in the high field gap of a Jones machine. This hybrid device generated an average field of nearly 10 kilogauss throughout a relatively large volume of 381iters (10 gallons). The new hybrid separator furnished a startling increase in efficiency over the previous Frantz and Jones Separators. For the first time, the equipment and process had reached the stage of commercial practicality.
At this point, further refinements were added as a result of contacts with consultants retained by the J. M Huber Corporation. This effort led to the design of prototype equipment generally resembling a Frantz Ferrofilter, but modified to generate fields of up to
FRANTZ FERROFILTER"

U.S. Patent 2,074,085

FIGURE 1-

Frantz FerrofilterR. Section elevation view (top) and plan view showing Frantz screen matrix and solenoid coil construction_

44

JONES CAROUSEL SEPARATOR

ROTOR

PLATE BOXES

FIGURE 2

Jones Carousel Separator. Schematic plan view (top) showing compartmented rotor, magnetic coils and pole pieces. Elevation view (bottom) shows steel yoke return circuit for flux.

Modern HEMF Production Units
In 1971, the Pacific Electric Motor (PEM) Company (Oakland, California) agreed with Aquafine to undertake design of full scale HEMF equipment using experience gained through 20 years of construction of large high field magnets in their own shops.
Because of extensive experience in the design and fabrication of large high field magnets, PEM was able to proceed directly with design of very large magnetic separators for full-scale use by the kaolin industry . Efforts were devoted to design of the largest machine whose components would be shippable by ordinary carriers.
Design and manufacture of the first 84-inch (213-centimeter) separator required solutions to numerous problems that had never been faced before in construction of magnetic separators.
A key feature in the design of these new separators was the long coil devised by J. Allen of PEM . In the long coil separators, the coil axis is 50 to 100 percent longer than the magnet gap or canister height. (Figure 5).
Since the long coil had a smaller mean diameter (compared to the prototype units). the long coil re quired less conductor than a short coil having the same number of turns. This saving in copper signigicantly reduced the capital cost of separators. Comparison of the two types of coils is shown in Figures 4 and 5.

20 kilgauss. Stainless steel woood had been shown at
J. M. Huber to be more effective on kaolin than coarser
ribbon Frantz screens, so the prototype incorporated this refinement.
The first production prototype high extraction magnetic filter (HEMF) was placed on stream in 1969 at J. M. Hube r. It consisted of a hollow conductor solenoid surrounding a canister 50 centimeters (20 inches) in diameter by 30 centimeters (12 inches) high. The solenoid was surrounded by a heavy box -like enclosure which served as a return circuit for magnetic flux generated by a short coil equal in height to the canister (Figure 4) . This small filter generated a field of 20 kilogauss and required 300 kilowatts of power. Throughput was about 2 tons per hour (dry basis) . Estimated processing cost was about $2 per ton which made the process of interest only for speciality kaolin applications.
A second small filter rated at 10 kilogauss and 4.8 tons per hour was delivered and placed on stream at J. M. Huber in 1970. In 1971, a third small unit of this type was acquired by the Industrial Minerals Department of Cyprus Mines.
Development of this equipment is the subject of numerous apparatus patents issued to J. M. Huber as well as to Magnetic Engineering Associates and M. I.T.

Cii:
I.U
1~
z
u <

0~[ ....-

~

FIGURE 3

Hybrid Separator. Schematic elevation views showing coils, steel yoke magnet with canister in magnet gap.

45

SHORT COIL SEPARATOR COILS
POLE PARALLEL TO COIL

LONG COIL SEPARATOR POLE RECESSED IN COIL

German Patent 2,111,986 AXIAL FLOW

RADIAL FLOW

FIGURE 4-

Short Coil Separator. Schematic cross section in elevation (top) and plan (bottom) views showing matrix, canister, coils and steel return circuit.

The first commercial HEMF unit was constructed by PEM in 1972 (Figure 6) and represented a 13 fold scale-up over the largest filter then in use at J. M. Huber. This PEM machine went on stream at Freeport Kaolin (Gordon, Georgia in the first quarter of 1973. A second PEM 84-inch machine went into production at American Industrial Clay Company in Sandersville, Georgia, during the fourth quarter of 1973, and a third PEM 84-inch machine was delivered to Thiele Kaolin in Sandersville, Georgia in early 1974. In addition, a 76inch (193 centimeter) magnetic separator was delivered to J. M. Huber by Magnetic Corporation of America in 1973. In the second quarter of 1975, Engelhard Minerals and Chemicals placed the fourth PEM 84inch unit in operation.

Description of HEMF Equipment

These PEM 84-inchseparators, which are the largest HEMF machines now in service, are far more efficient than any magnetic separation equipment that has been available in the past. The nucleus of the HEMF units consists of a canister which is (84-inch) 213 centimeters in diameter by 50 centimeters (20 inches) high and filled with compressed mats of 430 magnetic stainless steel wool. In essence, the unit is a powerful magnetic filter. The canister is surrounded by hollow conductor copper coils which energize the

FIGURE 5-

Long Coil Separator. Schematic cross section in elevation (top) showing coil taller than canister and pole pieces fitting into the coil cavity. Fluid flow is parallel or axial to the magnetic field.

entire void volume of the canister to 20 kilogauss. Water cooled copper coils in turn are surrounded by a steel magnetic return circuit 3.6 x 3.6 x 2.4 meters (12 x 12 x 8 feet). Efficiency of the magnetic circuit is such that power consumption is only 400- 500 kilowatts per hour.

Description of Process
HEMF units are operated in the manner of depth filters on a batch or cyclic basis. The filter bed matrix is magnetized and the slurry is pumped through to yield a filtered or nonmagnetic product. When the matrix is saturated, water is fed into the filter at a comparable rate to displace the nonmagnetic product.
The magnet is then de-energized and the bed flushed with high velocity water to remove magnetic contaminants. Processing of product is resumed after reenergizing the magnet. In normal operation, the duty or product cycle is 70 80 percent. Flow velocity can be varied from 102 centimeters (40 inches) per minute to 25 centimeters (10 inches) per minute, corresponding to a slurry retention time of 30 seconds to 120 seconds in a magnetic field. Filtration rate at 30 seconds retention corresponds to 90 liters per 0.09 square meter per minute (24 gallons per square foot per minute).

46

Utilization in Kaolin
HEMF equipment is employed to brighten kaolin by extraction of submicron feebly magnetic contaminants. Removal of only 1 - 2 percent of discolored ferruginous contaminants, based on original kaolin, typically brightens kaolin 2- 4 General Electric (GE) Brightness units . In extreme cases, brightness improvement may be as low as 1 or as much as 30 G E units, depending on the nature of mineral contaminants in the kaolin .
These improvements can be utilized to (a) mine lower quality crude kaolin, thereby extending reserves (b) reduce the consumption of leaching chemicals (c) produce super-brightness clays (d) allow use of hard media to delaminate kaolin without discoloration9 (e) produce clay products having superior end use performance in ceramics, glass, and catalysts because of reduced mineral contaminants.
Figure 7 shows the response of a good quality Georgia kaolin (80 percent minus two microns) as a result of HEMF processing. Even though the Fe203 content was reduced only 10 percent and the Ti02 content only 25 percent the brightness improvement of 3 points is commercially significant. A retention time of 60 seconds was sufficient to raise unbleached brightness from 84.7 to 86.6. Normally, this would require treatment of this kaolin with from 1.7 to 3.6 kilograms (4 to 8 pounds) of zinc or sodium dithionite

HEMF TREATMENT OF HIGH QUALITY GEORGIA# 2 KAOLIN

BRIGHTNESS

l . Cl

,_

z

~

~

,_:::E
~

--.z
0 u

1,0

-..
z.....,,.
t:
~
:1!!
.::.:.1.
~

fJ

t.U

._,.

MINUTES RETENTION TIME

FIGURE 7

FIGURE 6- 84-lnch Magnetic Separator with D. C. power supply.

leaching chemical per ton of clay. Magnetic separation allows this brightness level to be reached without leaching reagents. In practice, 0.5 to 0.9 kilograms (1 to 2 pounds) of leaching chemical may be used merely to improve shade of kaolin . In any case, there would be a savings of leaching regent ranging from 0.9 to 3.2 kilograms (2 to 7 pounds) per ton of clay. HEMF treatment of kaolin under these conditions costs less than 0.9 kilograms (2 pounds) of dithionite reagent.
With 2.0 minute retention time, brightness of this clay reached the 88 range. This can be converted to an 89 brightness after leaching, which allows the clay to meet premium or super brightness specifications and command a significant increase in price.
To illustrate the power and versatility of HEMF, submarginal kaolin samples from lone, California, were subjected to magnetic separation as shown in Figures 8, 9, 10, 11, and 12. As a result of HEMF processing, it was possible to convert all of these clays from unusable submarginal deposits to usable kaolin reserves .
Figure 8 shows the response of a submarginal kaolin to HEMF at 20 kilogauss, with 0 to 2 minutes retention time . The original G E brightness of this clay was about 73. Crude selection normally requires a minimum brightness of 78- 79 incommercial kaolin processing. After 30 seconds retention, the brightness of this sample rose to 79. This was accompanied by a decrease

47

1n iron content from 0.58 to 0.4 7 and a dramatic re duct ion in Ti02 content from 1.9 to 1.2. After 2 minutes retention, brightness of this clay was 82 and the final Ti02 analysis was 0.8 or less than half of the original.
Figure 9 shows a submarginal kaolin of 75 bright ness with an unusally high Ti02 value of 3.36. After 30 seconds magnetic retention, brightness reached 78 and the Ti02 content was reduced to a value of 2. 70. After 2 minutes retention, and brightness of the clay was about 82, allowing it to qualify for use as a paper filler without further processing. An interesting by product of this treatment was reduction of K2) from 0.30 to 0.22 and finally to 0.19. This represents a 37% decrease in mica content and has a very desirable improvement in the fired properties of this clay in ceramic outlets.
Figure 10 shows the dramatic reduction of Ti02 in a sideritic clay. The original Ti0 2 content of 2.38 decreased to 1.1 and finally to 0 .71 after two minutes retnetion. G E brightness increased 12 points from 62 to 74. Potassium oxide content of this clay was reduced from 0.21 percent to 0.14 percent after two minutes retention.
Figure 11 shows the unusual and spectacular response of a very discolored kaolin which would never be considered a useful reserve prior to the advent of high extraction magnetic filtration. After one minute retention time, brightness rose from 49 to 70. Fe203 content decreased from 2.3 to 0. 7 and Ti02 content

HEI\I f TRI::A I ~1LN1 OF MARGINAL KAOLIN 20 KILOGAUSS

'I I.

"'' .

1z i-z< 2.U -<
1z u 0
lR

-., lr.
z"....'.
1-
~
-:"..l.',l
~,.

1.4J .
-.....__

____ Fc20J

H

~--;......:_

I. U

.! .

MINl I S lc . . 'ION IIW

FIGURE 8

decreased from 1.9 to 0.8. After 4 minutes retention, this kaolin achieved a remarkable brightness of 79
and Fe2o3 and Ti02 contents of 0.5 and 0.4 K20
content decreased from 0.3 to 0.18 after 4 minutes, indicating significant removal of mica. As a result of magnetic treatment, an utterly unusable kaolin was converted to filler brightness. Subsequent chemical leaching of this whole fraction clay resulted in an 89 brightness which would put this clay in the premium brightness range.
Figure 12 shows the response of a gray lignitic kaolin (containing pyrite) to magnetic separation. Although the brightness increase was only 14 points over a retention span of 4 minutes and the final product submarginal with respect to brightness, the results are nevertheless significant. The Fe203 content decreased from 2% to 0.5% and the Ti02 reduction was parallel. K20 content was halved from 0.28 to 0.15.
The most significant aspect of this experiment was the striking reduction in the sulfur content of this clay. The initial sulfur value was 0.72 percent and the final sulfur content after 4 minutes retention was 0.08 percent. This represents a 90 percent reduction in the sulfur content of this clay through the nearly complete removal of micron and submicron pyrite by high ex traction magnetic filtration. This experiment has enormous significance in that it demonstrates removal of pyrite in a very difficult, fine-particle size range. This and other results indicated that magnetic separation

HEMF TREATMENT
OF HIGH Ti02 KAOLIN 20 KILOGAUSS

3. 0
...
z<z
...~
< 2.0 z
8
<#!.!

BRIGHTNESS

90
80
.,
...z - 70 w11'1 J:
- 60 5:2 Ill: a:l w - 5 0 \.?

1.0
-

- ~u

K20

1.0

2 , 11

MINUTES RETENTION TIM E

FIGURE 9

48

HEMF TREATMENT OF SIDERITIC KAOLIN 20 KILOGAUSS

3. 0
1z i-z< 2. 0 -<
1z u 0
"*
1.0

BRIGHTNESS

80 Ill Ill
z 70 101 1::c-:
~
= - 60 ell: 101 \,:1
- so
- 40

0

1. 0

2.0

MINUTES RETENTION TIME

FIGURE 10
of pyrite from coal is feasible and that high extraction magnetic filtration of coal is a viable method of coal desulfurization.
Removal of iron stained anatase, siderite, and pyrite as shown in Figures 8-12, as well as removal of hematite, tourmaline, and other ferruginous minerals from kaolin has enormous significance to the benefication of other industrial minerals ranging from asbestos to zircon.
Nature of Magnetic Fractions Removed
Mineral slimes extracted by HEMF from kaolin include iron-stained anatase and quartz, hematite, mica, tourmaline, siderite, and pyrite. Depletion of iron values in clay varies from 10 to 90 percent of original iron content. Bulk magnetic susceptibility of magnetic fractions is as low as 35 x 10-6 emu/cc, and iron content of magnetic fractions is usually less than 5 percent (expressed as Fe203).
Figures 11 and 12 show HEMF response of kaolins containing relatively high amounts of Fe203 as well as Ti02 while Figures 7, 8, 9, and 10 show effects on kaolins with a much higher content of Ti0 2 than Fe203. In all of these cases, iron-stained anatase is the principal mineral contaminant extracted by HEMF .
Magnetic separation in kaolin processing is a comparatively low cost and versatile technique. By variation of retention time, it is possible to achieve

differing degrees of magnetic extraction at acceptable variations in costs. The HEMF process is not only low in operating cost, but also furnishes high yield (98-99 percent) of kaolin product based on feed.
Economics
The economics of magnetic separation on large (84-inch) HEMF equipment are extremely attractive compared to that of small prototype units. This is due to efficient design coupled with the fact that the capacity of HEM F separators increases with the square of the diameter while the amount of copper conductor in creases linearly with the diameter. Operating costs for an 84-inch HEMF unit are estimated in Table 1.
Processing of Other Industrial Minerals by HEMF
HEMF processing has been applied to a series of 40 industrial minerals in a study conducted under the direction of H. H. Murray at Indiana University. This study was conducted on a 20 kilogauss mobile PEM 5-inch pilot plant unit leased from the Aquafine Corporation. A report of this work shows that HEMF is an economical benefication technique for many of the industrial minerals evaluated.
Figure 13 shows the degree of iron removed from a group of industrial minerals after HEMF processing at (B) 30 seconds (C) 60 seconds (D) and 120 seconds

HEMF TREATMENT OF DISCOLORED KAOLIN 20 KILOGAUSS

BRIGHTNESS

1.0
K2 o

TI0 2 MINUTES RETENTION TIME


:o
60 ""'z-'''
~ 50 ~ .....
o ~

FIGURE 11

49

TABLE 1 -Estimated Operating Cost of 84-inch Pem Separator

Amortization of installed separator ($1 ,600,000) over 10 years (80,000 hours) .. _ .
Magnet Power (400 KW @ 2r/. KWH (switching magnet off for flushing easily offsets rectification losses) . . . . . . . _
Pumping and Flushing Power (200 KW@ 2r/. KWH) ..... .
Labor (incl. benefits)
Maintenance . _ .. __ .. . .. . TOTAL

Cost per hour
$20.00 $20.00

General Purpose Use- Cost per Ton 66 tons per hour (30 seconds
retention)

Specialty Purpose Use- Cost per Ton 15 tons per hour (120 seconds
retention)

$0.30

$1.33

$0.30

$1.33

8.00
4.00 5.50
- 4-.0-0
$41.50

0.12
0.06 0.08 0.06 $0.62

. 53
.27 .37
- -.27
$2.77

HEMF TREATMENT OF LIGNITIC KAOLIN 20 KILOGAUSS z>z <
::;;:
<{
>z -
8
'*'
MINUTES RETENTION TIM[
FIGURE 12

70

60 V"J

Vl

zUJ

- so

>--
X

>2

""U""J
-10 1.;1

o Nevada bentonite

0.8

~ Montana talc

a California hectorite

~ Diatomaceous earth

r- ~o-~ ,.;.;
-:= l-.
6

u Texas white bentonite .;.. Sepioi ite
B_ ru_c_it_e_-o...__

0.4 ~ ~ / l

~X

"' XI/----::---"---X

0.2

RETENTION

TIME

0

SECONDS

FIGURE 13

30

60

120

50

0 .1 0
0 .08
<l.l
u.... 0.06

o Gloss sene x Silica flour o Calcium carbonate

<J
u
~ 0.04 ~

.

0.02 l::=====::,- J:

ODO
RETENTION 0 TIME SECONDS

30

60

120

FIGURE 14

=-~

....
CD
w 0 80

78 <> Calcium carbonate
o Nevcaa bentonite
A ArKansas novaculite

76 ~------L-------L-----~

RETENTION 0
TIME

30

60

120

SECONDS

FIGURE 16

* 1.6

-Tennessee bell clay

~~:--Kentucky ball clay

1.4 x............

...............

........ ..................

~x--_Il02

1.2

x,

~

c~

''-,,,,''x,,_Ti 02 ]>:

~(j) 1.0 a<.!.)
j 0.8

----K----~--

--------~--~~~~~~=~~---- - 0

Fe

........ ........ ........
' >:
------

0.4 ~~'-K-----<>----<>----=-J~

60 RETENTION L_________L _________ L--------~120

TIME

0

30

SECONDS

FIGURE 15

92 ~ ~ ~:I I ....] X
!
I
90

VI
-VI Q) c::
~
. ~ .... 88
OJ
w
{!)

86

x Whiting

o Silica flour

84

RETENTION

TIME

0

SECONDS

FIGURE 17

30

60

120

51

retention time at 20 kilogauss. The untreated sample is shown by A. Of this group, diatomaceous earth, sepiolite, and Montana talc show signigicant reductions of iron content.
HEMF processing of glass sand was successful in reaching optical glass standards after 60 seconds retention time (Figure 14). Treatment of Kentucky and Tennessee ball clay achieved only small reductions in iron content, but significant reductions in Ti02 and K resulting from removal of deleterious anatase and mica, (Figure 15).
Even though Nevada bentonite (Figure 13), and calcium carbonate showed only slight removal of iron, these removals resu Ited in sharp and significant upgrading of brightness (Figure 15).
Likewise, Montana talc, whiting, and silica flour showed significant increases in brightness as a result of HEMF processing. Texas white bentonite gave a spectacular brightness response going from 83 brightness to 91.5 after 30 seconds, and finally to 93 brightness after 120 seconds retention time, (Figure 16).
Utilization in New Areas
The kaolin separation problem represents the most difficult application of magnetic separation ever commercialized. This technology can now be applied advantageously to beneficiation of other industrial minerals, concentration of metallic ores, desulfurization of coal, and clean up of effluent streams, Programs on magnetic separation of metallic ores, and desulfurization of coal are currently in progress at Indiana University under the direction of Dr. H. H. Murray. An NSF sponsored study on desulfurization of solvent refined coal is also under way at Auburn University under the direction of Dr. Y . A. Liu .

References
lannicelli , J., "Assessment of High Extraction Magnetic Filtration", National Science Foundation, (RANN) ISRU Washington, D .C. 20550 - NSF/RA/ G-74/009 PB 240-8805/WN - National Technical Information Service, Springfield, Virginia 22121, 1974.
lannicelli, J, Millman, N, Stone, W.J .D., "Process for Improving the Brightness of Clays", U. S. Patent 3.471,011, October 7,1969.
lannicelli, J., "High Extraction Magnetic Filter Separator", West German Patent 2, 111,986, October 21, 1971.
lannicelli, J., "High Extraction Magnetic Filter Separator", British Patent 1,347.396. February 20, 1974.
lannicelli, J., "High Extraction Magnetic Filter Separator", Canadian Patent 935,126, October 21, 1971 .
Marston , P.F., Nolan , J.J., Lontai , L.M ., "Magnetic Separator and Magnetic Separation Method", U. S. Patent 3,627,678, December 14, 1971.
Kolm, H.H ., "Process for Magnetic Separation", U.S. Patent 3,676,337, July 11, 1972.
lannicelli, J., 1976, "High Extraction Magnetic Filtration of Kaolin Clay and Mineral Slimes " in Filtration & Separation, January /February.
Whitley, J.B., lannicelli, J., "Method for Producing Mineral Products", U. S. Patent 3,667,689, June 6, 1972.
Murray, H.H ., lannicelli , J., "A Survey- Beneficiation of Industrial Minerals By High Inten si ty Wet
Magnetic Separation", NSF Project # GT 44219 -
Report in Progress.

52

Factors for Converting English Units to International System (SI) Units

The following factors may be used to convert the English units published herein to the International System of Units (SI).

Multiply English units

By

To obtain Sl units

Feet (ft)

.3048

metres (m)

Miles (mil

1.609

kilometres (km)

Square miles (mi2)

2.590

square kilometres (km2)

Gallons (gal)

3.785

litres (I)

Million gallons (106gal)

3785 3.785x10-3

cubic metres (m3) cubic hectometres (hm3)

Gallons per minute (gal/min)

.06309 .6309 6.309x10-5

litres per second (1/s) cubic decimetres per second (dm3/s) cubic metres per second (m3/s)

Million gallons per day (Mgal/d)

43.81 .04381

cubic decimetres per second (dm3/s) cubic metres per second (m3/s)

53

DEFINING A COMMERCIAL DIMENSION STONE MARBLE PROPERTY
Lance Meade Vermont Marble Company
Proctor, Vermont

A few years ago there was a popular song that asked -"Where have all the young men gone?" This of course referred to the lack of young men at the end of World War I. Today through, the producer of Dimension Stone Marble, Granite, Slate or Limestone can be heard singing "-Where have all the building projects gone?"
There are a few mid-western banks putting up an addition or two, and an occasional insurance company refurbishing a lobby, but for the most part, the area in the construction industry that uses a large volume of dimension stone is in a very depressed state of activity. What is more, the domestic industry has gone through a decade of devastating Italian competition. With a diminished market and cheap imports, it is a quiet time for most domestic stone producers.
There are areas in the third World Nations, however, who have invested development funds in resource studies and have been told that marble lies waiting to be marketed in their hills. They have been reading up on the amount of cubic feet of marble that Italy has exported to the big market place of North America in the recent past, and how many dollars have been flowing back into that greatly subsidized Italian industry. These developing nations have wanted to enter the marble mining industry. These nations have since learned that it takes more than a Unido-financed report to make a potentially commercial marble deposit located far from population centers and transportation into a reality.
The question remains: what makes a marble property commercially viable? The answer could be either a good government subsidy program or an extensive marble deposit that could be economically extracted and marketed in an acceptable form to return a profit to the owners.
Discussion, here, is limited to the type of stone deposit that would yield enough uniform material to satisfy the contemporary styled large structures that were being built prior to the economic slow-down. This would exclude the many small multi-colored deposits of decorative stone that are available in the world. Traditionally marble has been used in structures that use cubic pieces of stone that have been sawn from blocks. These cubic pieces normally are as thick as they are high in dimension. This style of architecture has been duplicated somewhat in using a thinner 1 1/4" veneer and fastened directly to the structural steel of contemporary buildings. Marble also is presently being used in an 1 1/4" or a 7 /8"

thick veneer combined with steel and glass to give an aesthetically pleasing building.
To satisfy the architects of these kinds of structures, the stone used has to meet certain requirements. The deposit has to be extensive enough in uniformity to supply consistent material that is marketable within a very restricted range of suitable color and texture.
The geological parameters of these kinds of deposits have been very adequately described in past reports of the Bureau of Mines, the U.S. Geological Survey and individual articles. A coherent core is needed when initially investigating the property by core drilling as well as a very favorable waste-to-ore ratio with a maximum utilization of the entire stone deposit. The deposit must yield consistently uniform blocks that will produce slabs of similar color and markings and be acceptable to an architect. An abundance of cracked, cross-clouded or variably colored blocks that will fall apart in the gang saws and yield unmatchable slabs in the shops is obviously undesirable.
As Robert Power stated in his 1972 article in "Mining Engineering, "... most cut dimension stone is sold by the cubic foot by vertically integrated companies". A well planned mill with maximum gang-saw space utilization and a well-run, efficient shop-layout to minimize handling and maximize utilization of stock is essential for an economically viable property.
With the filling of the above parameters, the dimension stone producer can expect to have a well defined, highly feasible dimension stone property. That is of course if he is a domestic producer and the E.P.A. and MESA and O.S.H.A. regulations don't impose such costly alterations to his original plan of production that his original cash flow analysis becomes negative rather than positive. Or, if he is on foregin soil and his operations aren't confiscated for the good of the people; or if he is in Canada and isn't taxed into oblivion . If he can surmount these perils and the economy really is on the upswing, and we start building again -using stone to conserve energy and create beautiful structures and use up our accumulated block and slab inJentories, then maybe we can justifiably say we have a commercial dimension stone property.
Many authors have contributed to the various techniques that can be used in the exploration of mineral deposits in general, and a few authors have been more specific in covering dimension stone. Some of these sources are conveniently listed in the bibli-

54

ography at the back. In any case, at the expense of being repetitive, I would suggest the following:
Phase I - Site Selection and Production Analysis
A. Determining the potential deposits to supply graded marble of uniform color, texture and size of blocks.
1. Literature Search Overview the entire list of deposits and determine the best sources based upon the above criteria.
2. Literature Search and On-Site Visits Review the present active quarry areas to determine the extent of present producers expertise and sources of material.
3. Select target areas of investigation, looking for those sources of natural stone that best answer the architect's design characteristics on current and projected buildings.
4. Select target areas based upon geographical locations of upcoming large contracts and building projects.
5. Review the transportation costs from the selected deposits to proposed building projects.
6. Select the Mill and Finishing Facility location in conjunction with domestic transportation and best location for importing of supplies and materials.
B. Determining the quarry feasibility of each individual target area.
1. Determine the geological uniformity and structures.
2. Determine horizon of most marketable material.
3. Determine potential percent of market ability of various horizons
4. Determine the percentage waste due to unsoundness, poor color, and nonuniformity of marble.
5. Study and determine the overburden and waste rock removal costs.

C. Determine costs of quarrying and plant construction of Mill and Finishing Shop to determine competitive costs with established sources in the country and current importers.
1. From selected sites outline various operating plans based upon geological parameters of deposit, available labor, equipment, power sources.
2. Determine finishing characteristics of production stone and develop design and layout of Mill and Finishing Shop according to block size and slab handling facilities needed.
3. Preliminary Costs :
a. Procurement of required operating permits.
b. Preliminary road construction and water supply, core drilling, test block, extraction.
c. Quantify volume available from deposit measured against projected demand volume and volume of waste needing to be moved.
d . Determine potential productivity factors, establish time parameters to develop deposit into full production cycle (establish site preparation costs and project potential quarry production costs, design plant facilities in conjunction with volume supply and projected demands).
4. Site Preparation Costs:
a. Prepare roadway systems, quarry buildings, water and power supply.
b. Procurements to set up quarry equipment.
c. Establish plant site and !nstall buildings, procure and install Mill and Finishing equipment.
d . Establish sources of supply of various supplies (abrasives, cutting tools, handling equipment) .

55

5. Development Costs:
a. Removal of overburden and waste
rock
b. Develop Keyways across produc tion layers -open production zones.
c. Initiate gang sawing and prelimi nary coping and finishing.
d. Review production scheduling to establish anticipated lead times for production cycles.
Phase II - Market Analysis
D. Determine a five year and ten year pro jected building projects within the country.
1. Determ1ne present suppliers and potential suppliers, both domestic and foreign.
a. Study past, current and projected volumes of variety of stone of each.
b. Determine present pricing and cost parameters of present suppliers.
2. Determine quantities of concrete, structural steels, and natural stones anticipated to be used within this time frame.
3. Determine the quantities of materials that will be imported versus that which can be supplied domestically. Identify the factors that will alter present development trends. Deter-

mine who the present Architects and Engineering Consultant Groups are. Determine the primary contractors operating and potential contractors operating within the market area.
Bibliography and References
Anon. 1970, "Mineral Resources Development with Particular Reference to the Developing Countries". United Nations Publication, E.70, II.B. 3., 74 pp.
Barton, W.R., 1968, "Dimension Stone", Information Circular 8391, U.S. Bureau of Mines, 147 pp.
Bowles, 0., 1956, "Granite as Dimension Stone," Information Circular 7753, U.S. Bureau of Mines, 18 pp.
Bowles, 0., 1956a, "Limestone and Dolomite," Information Circular 7738, U.S. Bureau of Mines, 29 pp.
Bowles, 0., 1958, "Marble," Information Circular 7829, U.S. Bureau of Mines, 31 pp.
Currier, L.W., 1960, "Geologic Appraisal of Dimension-Stone Deposits," Bulletin 1109, U.S. Geological Survey, pp. 7-14.
Havard, J.F ., 1970, "Mineral Project Evaluation," Mining Magazine, Vol. 123, No.4, October 1970, pp. 326-329.
Hockman, A., 1953, "Physical Properties of Currently Produced Marbles," Circular LC1 010, National Bureau of Standards, 13 pp.
Power, W.R., 1973, "An Evaluation of Building Dimension Stone Deposits," Mining Engineering, Vol. 24, No. 6, June, pp. 42-44.
Power, W.R., 1975, "Dimension and Cut Stone," Industrial Minerals and Rocks," 4th ed., S.J. Lefond ed., AIME, New York, pp. 157-174.
Shadmon, A., 1972, "Stone in Israel," Ministry of Development, Natural Resources Research Organization, Stone Technology Center, Technion, Haifa, 64 pp.
Winkler, E.M., 1973, Stone: Properties, Durability in Man's Environment, Springer-Verlag, New York, 230 pp.

56

SOUTHEASTERN CERAMIC RAW MATERIALS
J. L. Pentecost School of Ceramic Engineering Georgia Institute of Technology
Atlanta, Georgia

Traditionally, the ceramic industry has used primarily mineral raw materials, often native materials with little beneficiation other than grinding and sizing. While many new ceramic products depart widely from this tradition, a large segment of the industry is still closely tied to its mineral raw material supplies.
Consider first the three large tonnage consumers of minerals for producing ceramics: the brick industry, the glass industry, and the portland cement industry. These industries thrive nationwide largely because clays, limestone, and sand deposits are abundant. The southeast is particularly fortunate in its native clays, shales, limestone, and sand deposits. The brick industry in this area, for example, has been the leader in innovative production techniques. There are also approximately 20 cement plants in the southeastern U.S. with some of the most modern technology for dust control and close-cycle material handling represented in these plants.
The glass industry deserves special note. Over 70 percent of most compositions is silica sand, which must be a minimum of 99 percent Si02 with less than 300 parts per million Fe2 03 and 3 parts per million Cr, with a minimum of 85 percent plus 100 mesh, 0 percent plus 20 mesh. These are specifications for a material selling for $3.50 per ton at the plant. These specifications are necessary for ordinary clear container glass, which is most often used; for colored glasses the specifications are relaxed somewhat. To avoid confusion, most plants purchase only a single grade of sand.
The glass industry developed in the north and midwest, near population centers, using the St. Peter Sandstone (Illinois) and the Oriskany Sandstone (Virginia, Pennsylvania, and West Virginia). Both of these are excellent deposits and have 5erved the industry well. Twenty-five years ago there were only a very few glass plants in the southeast so most of the glass we used was shipped into the area. With the growth of the soft drink industry, along with "no return" bottles, rapid expansion of southeastern markets occurred. This resulted in the development of excellent glass sand deposits in southern Georgia and in Florida, and today a wide variety of containers and glass products are made in the south for southern markets.
When considering the value of brick, glass, and portland cement on a per pound basis, the pressure for cheap raw materials is obvious. Brick sells for just 1 per pound. To mine the clay, do at least minimal benficiation, maintain constant water content, formulate the desired colors, shape, stack, dry, fire to 2000 F, package,

move this heavy product a couple of times and sell the finished product profitably at $25/ton requires profound engineering and very large facilities. It is generally uneconomical to process less than 400 tons of brick per day.
Portland Cement is similar in value. The clay and limestone must be ground together, fused to a clinker at 2800 F (somewhat higher temperature than brick). ground to 325 mesh fineness, and sold for just over 1 per pound also.
By comparison, glass containers sell for about 10 per pound, with the cost of chemicals (soda ash), the accurate forming, and the inspecting all reflected in this cost. Even fiberglass insulation is only 17-24 per pound, and fiberglass filaments for textiles is approximately 35 per pound. With such value on a per pound basis, it is not surprising that these are large industries.
Other segments of the ceramic industry using pri marily mineral raw material include the tile industry, sanitary ware industry, and the dinnerware industry. Kaolin clays, ball clays, talc, silica, pyrophylite, and feldspar are among the major raw materials utilized by these industries. Low iron content and low alkali impuri ties are usually desirable. The refractories industry also depends on native clays, magnesite, (often from seawater) dolomite, chromite, silica and alumina minerals. While much of the technical ceramic industry uses re fined raw materials, the chemicals used are usually directly related to processed mineral raw materials.
Some of the specific minerals and their uses in the southeast deserve mention. Georgia and South Carolina's abundant kaolin clay and North Carolina's primary kaolin are widely used in dinnerware, refractories, sanitary ware, and tile. Ball clays from Kentucky and Tennessee are also necessary in most compositions for improved farming. In these same compositions are feldspars from North Carolina (about 70 percent of native production) and Monticello, Georgia. Feldspars are also important constituents in glass, with aplite and nepheline syenite also included in this group of fluxes. About 500 thousand tons are used annually in ceramic, glass and porcelain. Ceramists usually distinguish be tween high potassium and high sodium feldspars for specific uses. High potash feldspars are generally less available than soda or mixed feldspars.
Limestone for cement is produced widely since there is relatively little restraint on the iron and trace impurities it may contain. In contrast, limestone used in glass and ceramics must meet rigid specifications con

57

cerning iron content (less than 0.1 percent Fe2 03 ). Ceramic talc, a major constitutent in steatite
electrical insulators and wall tile, is produced primarily in Texas, California, Montana, and New York. Ground pyrophylite is also used in these products. Though Georgia has proven talc deposits several times the size of those being worked elsewehere, its iron content is too high for ceramic uses and would have to be removed. While such processing is practical, these deposits remain to be exploited as ceramic raw materials. Spodumene from Kings Mountain, North Carolina, is used as a major constitutent of many low expansion ceramics. As it occurs, this spodumene is too high in iron until bene ficiated by hot chlorine leaching, volatilizing the FeCI3 .
Ti02 is a constituent used in enamels and glazes to opacify the glass. It is a major constituent in special dielectrics and piezoelectrics. Ti02 is refined from southeastern deposits of heavy mineral sands; currently only Florida deposits are being mined.
The ceramic industry is also a large consumer of fuel, primarily oil and natural gas. All ceramic products are heated above 1000 and often above 2000 F. The continued availability of these fuels is important, but conversion to coal could be accomplished in much of the industry, if this became a necessity.
THE FUTURE
As energy costs increase, some product displace ments will occur due to resulting cost differentials; however, most ceramic producers are confident that there will be a reasonable balance and ultimately, due to bountiful raw materials, ceramic products will fare well in spite of their energy component. The southeastern market continues to grow as income levels increase and

markets increase. This optimism provides the impetus in Georgia, South Carolina, and Florida for the continued education of ceramic engineers to serve this industry. Such trained personnel are essential for continued growth of the industry, and the ceramic industry pro vides excellent employment opportunities for the future.
The ceramic industry is a demanding customer of mineral suppliers; (this has been so in the past and will continue in the future). Improved mineral raw materials will be necessary for our improved products and dynamic technology, As we must control our products more closely, raw material will receive increasing scrutiny. We will need new material in large quantity and, with transportation costs an ever present concern, nearby mineral deposits will be developed with the lower transportation cost providing part of the differential needed to support the beneficiation of lower grade deposits and to justify the capital investment.
It is most disturbing to see the ceramic industry press mineral suppliers to a point of cost competition where inadequate research in beneficiation and exploration can be supported out of the price obtained for the mineral. This is ultimately destructive to both industries which are interdependent and ought to work together for improved ceramic products.
REFERENCES
S. J. Lefond, ed 1975, "Industrial Minerals and Rocks," 4th edition, AIME, New York, New York.
E.J. Kliff et. al., 1973, "Ceramic raw material," Charles H. Kline & Co., Fairfield, New Jersey.
U.S. Bureau of Mines, 1972, "Minerals Yearbook," Volume I.

58

ECONOMIC GEOLOGY OF THE GEORGIA MARBLE DISTRICT
W. Robert Power Georgia State University
Atlanta, Georgia

Introduction
The Georgia Marble District, located in Pickens and Gilmer Counties, Georgia, has produced marble for more than one hundred and thirty years. For the first one hundred years practically all production was dimension stone, but since World War II a wide variety of crushed and ground marble products have been produced and sold. These include terrazzo and other decorative chips, agricultural lime, fillers, extenders, and similar products. Today the crushed and ground products account for a higher volume and dollar value than does dimension stone.
The Georgia Marble District includes three outcrop belts of the Murphy marble in Pickens and Gilmer counties, Georgia (Fig. 1). The largest is a hook-shaped area that occupies the valley of Long Swamp Creek at Tate, Georgia, and its East Branch near Marble Hill, Georgia . It was called the Marble Hill Belt by Bailey (1928) and the Tate Belt by Fairley (1965, p. 23). In this paper it is called the TateMarble Hill Belt. A second belt lies on the east side of Long Swamp Creek about two miles east of Jasper. It was called the Long Swamp Creek Belt by Bailey (1928)'and by Fairley (1965). However, the ridge from which the stone is mined is locally known as Cove Mountain and the stone is known commercially as Cove Mountain stone. Therefore the outcrop belt is referred to in this report as the Cove Mountain Belt.
The third belt lies on the east side of Talona Creek at Whitestone, Georgia. It lies partly in Pickens and partly in Gilmer County. It is referred to as the Whitestone Belt in this report.
Parts of the three belts are now being mined by the Georgia Marble Company. Crushed and ground products are produced from all three. Dimension stone is produced from the Tate-Marble Hill Belt only.
Regional Geology
The Murphy Marble.lies within the Murphy Marble Belt, a sequence of lower Paleozoic metasedimentary rocks occupying a long synclinal trough that extends from Bryson City, North Carolina to Canton, Georgia. The sequence was first described by Keith (1907) who described the rocks near Murphy, North Carolina. The marble and enclosing schists were traced southward into Georgia by LaForge and Phalen (1913), Bailey (1928), Hurst (1955), and Fairley (1965). Power and Forrest (1973) discussed the stratigraphy and correlations within the Murphy Belt.
The Murphy Marble over Iies the Brasstown Formation (of Hurst, 1955) which consists of thinly bedded mica schist and feldspathic quartzite. The Brasstown Formation makes excellent flagstone and is quarried locally for that purpose.
There is some confusion regarding the nomenclature of rocks overlying the Murphy Marble. Keith (1907) described the Andrews Formation as a thin calcareous schist that

overlies the marble. Hurst (1955) redefined the Andrews

Formation to include several hundred meters of pelitic schist

above the calcareous schist. Fairley (1965) introduced a new

name -the Marble Hill Hornblende Schist- for a thin calc-

schist that overlies the Murphy Marble at Tate and assigned

overlying garnet-mica schist to the Andrews Formation .

Power and Forrest (9173) pointed out that the original

definition of the Andrews Formation was for a thin calc-

schist that overlies the Murphy Marble and corresponds to

the Marble Hill Hornblende Schist. The succeeding garnet

mica schist should therefore be called the Mineral Bluff

Formation following Hurst (1955).

Whatever the nomenclature all writers have agreed upon

the succession of lithologies. The succession with names used

....

in this report given in parentheses is from oldest to youngest:

(1) thinly bedded mica schist and feldspathic quartzite

(Brasstown Formation). (2) marble (Murphy Marble),

(3) calc-schist (Marble Hill Hornblende Schist). (4) garnet

mica schist (Mineral Bluff Formation).

Fairley (1965, 1969) reported that the Murphy Marble

is not a continuous layer throughout the Murphy Marble

Belt, but rather is discontinuous and locally interfingers

with the enclosing schists. Power and Forrest (1973) agreed

and interpreted the marble as original reefs.

Fairley (1965) distinguished between relatively pure

calcite marble and dolomite marble, and impure micaceous

marble and calc-schist. He included the former within the

Murphy Marble, but considered the latter to be a calcareous

facies of the Mineral Bluf (i.e. Andrews) Formation. Bayley

(1928) had included both within the Murphy Marble.

The Murphy Marble as defined by Fairley occurs

principally in two areas named the Tate Marble Hill Belt

and the Cove Mountain Marble Belt. All commercial

production of marble from the district has come from these

two areas plus a third area outside the Tate Quadrangle at

Whitestone, Georgia.

Tate-Marble Hill Belt

The Tate-Marble Hill Belt crops out in a hook-shaped area that occupies the valley of Long Swamp Creek at Tate, Georgia and its East Branch near Marble Hill, Georgia (Fig. 1 ). It is the area first exploited for marble and has produced practically all the dimension stone in the district. Today dimension stone is produced from the main valley of Long Swamp Creek at Tate; crushed and ground marble is produced from Marble Hill along the East Branch or "barb" of the hook.
Lithology

The marble is a medium to coarse-grained calcite marble that is magnesian or dolomitic in places. The dolomitic marble is generally finer grained. Individual crystals in the calcite marble typically are five to ten millimeters across

59

lWHITESTONE 1 r --,
r -'-- J ' - - - -- - _j

JASPER

TATE

,..._ .,

- - - ----1

'--- -, 1.-.1

I PIC~Sl::!..a,. ~:.....__ ___

C HE"ROKI!E CO.

I

I

I
I

5

0

5

10 K,.,., .

~e==~F~3~E3=t========~======~

FIGURE 1. Location map of Georgia Marble District. Marble outcrops shown in solid black. 60

whereas crystals in the magnesian marble are typically less than two or three millimeters. The marble contains varying amounts of accessory minerals, chiefly graphite, mica, amphibole, pyrite, and epidote. These typically occur disseminated in layers or "veins" that give the marble its character as ornamental stone . A number of commercial varieties have been named based on color and nature of the veining. Mapping these varieties has proved useful in working out the local stratigraphy and structure of the marble. The more important varieties are described below.
Cherokee. Cherokee marble has a white background and gray markings. The markings are caused by disseminated graphite, black mica, and finely disseminated pyrite. They occur as layers or bands which are in places straight, but more commonly are disrupted and contorted, appearing as irregular veins and blotches. Markings make up as little as one percent to as much as fifteen percent or more of the rock. Cherokee marble is the principal variety of building stone quarried from the district.
Messotint. Messotint marble has a gray background with dark gray or black markings. Accessory minerals and the character of markings are the same as for Cherokee marble . The gray background is caused by finely disseminated graphite and pyrite.
Creole. Creole marble has a white background with black markings . Accessory minerals are the same as for Cherokee marble, but more heavily concentrated. The amount of dark colored marking is much greater, commonly exceeding 50 percent of the stone.
Silver Grey. Silver Grey marble is a uniform gray variety in which the gray color comes from finely disseminated graphite and pyrite. It is used extensively for monuments.
Etowah Pink. Etowah Pink marble has a light to deep pink background with green to greenish black markings. Accessory minerals are ch iefly green biotite, but include amphibole and epidote. The pink color comes from hematite that is disseminated along grain borders. The markings are shaped similarly to those in Cherokee marble. They generally make up 10 to 30 percent of the stone.
Golden Vein. Golden Vein marble has a white back ground with golden brown markings caused by accessory phlogopite and less commonly, colorless to pale tan amphibole. The more highly prized Golden Vein has only the golden brown phlogopite set in a white background of calcite crystals. However, in places the layers containing the phlogopite also have a grayish background color, probably caused by finely disseminated graphite. In places Golden Vein type marble contains patches or layers with green mica and amphibole rather than tan and brown. This makes a striking looking stone, but it is not common enough to have been quarried and marketed as a separate variety.
White Georgia. White Georgia is nearly pure white marble with only a trace of color resulting from widely scattered grains of phlogopite and rare colorless or pale tan amphibole . It is the most highly prized marble in the district, but is relatively rare.
Rosepia. Rosepia is a fine-to medium-grained magnesian marble that is pale pink in color with amber colored streaks or "veins." The amber color is caused by disseminated grains of phlogopite.

Stratigraphy
The stratigraphy of the marble is best displayed at Marble Hill (Fig . 2) where the layering is regular, the dip rarely exceeds 20 degrees, and several deep core holes have penetrated the entire formation . The stratigraphy was first described by Reade (1965).
Detailed mapping and core drilling in the New York Mine (Fig. 2) has resulted in the recognition of many units, some of which are very thin and of very limited aerial extent. Other units grade into each other along the stratification and probably reflect local facies changes in the original sediments . In order to summar ize the vertical and horizontal changes it is convenient to distinguish five units which will be called: (1) magnesian marble, (2) middle schist, (3) quartz sulfide marble, (4) silicate marble, and (5) graphitic marble or "Cherokee" type. All units are recognized at Marble Hill, but only the silicate marble (4) and graphite marble (5) have been recognized at Tate.
The contact between marble and the underlying schist (Brasstown Formation) is abrupt and apparently conformable. The schist is a medium to coarse-grained feldspathic, quartz, biotite schist.
Magnesian Marble . The magnesian marble is the lowermost marble unit. It is characterized by high magnesian content and dolomite layers. It consists of alternating layers of fine-grained white, gray, and tan dolomitic marble and coarse-grained white to light gray marble. Some layers contain abundant sulfide minerals. Rosepia marble comes from this unit. The thickness is variable, ranging from about 18 to 50 meters under the New York Mine, but it thins out and disappears to the southeast.
Middle Schist. A thin layer of hornblende, biotite, calcite schist overlies the magnesian marble . In hand specimen or drill core it is hardly distinguishable from the Marble Hill Hornblende Schist which overlies the Murphy marble. The middle schist ranges in thickn ess from about one and onehalf to five meters ( 5 to 16 feet).
Quartz, Sulfide Marble. Quartz-bearing, gray marble containing disseminated pyrite overlies the middle schist. Pyrite also occurs in fractures that cut the regular stratification . No commercial stone comes from this unit.
Silicate Marble. The silicate marble unit contains the commercial varieties Etowah Pink, Golden Vein, and White Georgia. It is white, light gray, or pink marble with accessory silicate minerals (mica, amphibole, and epidote) that generally are disseminated in layers giving the rock a striped appearance. The stripes or layers commonly range in thickness from one to about fifteen centimeters (-5 to 6 inches). Colored layers contain anywhere from a few percent to as much as thirty percent or more accessory minerals . In extreme cases dark colored layers approach calc schist in composition.
The chief accessory mineral is mica which may be golden brown phlogopite or green biotite. Amphibole ranges in color from very pale tan to bright green. Epidote is less common, but is present in some layers. Quartz, pyrite, norbergite fuchsite, and chlorite occur rarely. Where the background color is white the accessory minerals are generally pale tan or brown in color. Where the background color is pink the

61

accessory minerals are green. Green mica and amphibole occur less commonly in white marble.
The silicate marble is more than 75 meters (250 feet) thick in the New York Mine. The lower half is much darker or more heavily striped than the upper half of the unit. The upper half contains an elongate lense or prism-shaped body of almost pure white calcite marble. The commercial varieties Golden Vein and White Georgia are produced from this area. Pink marble is common in the lower half of the unit and becomes more abundant west of the New York Mine.
Pink silicate marble is also exposed in the central part of Long Swamp Valley at Tate, where it is quarried as Etowah Pink. Brown, striped Golden Vein marble occurs on both sides of the Etowah Quarry.
The silicate marble is thickest in the New York Mine, but it thins rapidly westward and is less than 100 feet thick a few hundred feet west of the mine .
Gra phitic (Cherokee) Marble. The graphitic marble contains disseminated py rite, black mica, and hornblende as well as graphite. Cherokee, Messotint, Creole, and Silver Grey Marble all come from this unit. The unit is less than 30 meters (100 feet) thick at the New York Mine, but it thickens westward and at Tate it is the dominant marble exposed in the valley. The structure at Tate is too complicated to work out detailed stratigraphy, but the darker varieties (Creole and Messotint) tend to be concentrated along the west margin of the valley. They grade eastward into typical Cherokee Marble and finally into silicate marble (Etowah) in the east central part of the valley. Cherokee Marble has been quarried from both Marble Hill and Tate, but Creole, Messotint, and Silvery Grey, from only Tate .
Transi tion Mar bl e. The Murphy Marble is overlain by the Marbl e Hill Ho rn blende Schist. However, there is a transition zone about 15 meters (50 feet) thick between typical Cherokee type marble and typical hornblende schist. This consists of alternating layers one to ten centimeters (4 inches) thick of white marble and calc-silicate schist. The calc silicate layers contain hornblende, biotite, epidote, and calcite. They become increasingly silicate rich and darker in color upsection and grade into hornblende schist containing only minor epidote, biotite, calcite, and other minerals.
Structure
Layering. Layering is well developed in the marble at Marble Hill. The silicate marble shows it particularly well where colored layers containing abundant accessory minerals alternate with white layers of pure white marble. Similar layering is less spectacularly displayed by other units . Individual layers range in thickness from a centimeter (.4 inches) or less to several meters (feet). They can be traced tens of meters (feet) in the mine workings and groups of layers can be correlated several thousand meters (feet) by core drilling.
The present mineral assemblages are metamorphic in origin, but they undoubtedly reflect original chemical differences in sedimentary layering .
Layering is also present at Tate, but it is much distorted by deformation . Tight folds a meter (3 feet) or less in amplitude are common. Layers are discontinuous so that it is seldom possible to trace a unit more than a few meters

(feet). Distorted and disrupted dark layers make the typical "veining" of the marble . Abundant barb-shaped veins probably are the hinges of small disrupted folds. These typically plunge about 15-20 degrees down the valley to the southeast.
Foliation . Foliation resulting from the sub-parallel alignment of micas and amphiboles in dark colored layers is almost invariably parallel to the layering. Petrographic studies by Fairley (1965) showed flattened calcite crystals and parallelism of calcite twin lamellae.
Folds. Several orders of folds exist in the area. The largest or first order folds include an open south plunging syncline at Marble Hill and a tight overturned southeastplunging anticline at Tate. These two folds produce the hookshaped outcrop pattern of the belt.
At Tate second order folds ten to fifty meters (30 to 150 feet) across can be seen in quarry walls. These are flowage (passive slip) folds that are greatly thickened along the axes or hinges of anticlines. The axial planes generally dip steeply east. Some of the best quality dimension stone produced in the area occurs in the thickened hinge areas of these folds and the quality stone can be followed down plunge to the southeast.
Third order folds probably show the nature of the deformation best. Individual layers are drawn out, thinned, and commonly disrupted along the limbs, but are greatly thickened at the axes where they form hook or barbshaped features in the rock. The plunge of the third-order folds is the same as that of first and second order folds.
In the New York Mine at Marble Hill the structure is a very broad, shallow syncline that plunges about 15 to 20 degrees south. Second and third order folds have not been found within the mine; however, an elongate lense or prismshaped body of Georgia White Marble lies with its long axis down the plunge. Correlation of layers or units is easier and more consistent southward down the plunge than in the eastwest direction. Immediately east of the mine there appears to be a steep anticline that is overturned to the west and thrust faulted from east to west. There is further evidence from drill holes and mapping that a zone of tight folding exists immediately west of the mine.
Faults. Two faults have been recognized from mapping and core drilling. Both are east-dipping thrust faults that apparently grade along strike into overturned anticlines. One occurs southwest of the New York Mine and the other along the east side of the valley of Long Swamp Creek at Tate, east of the dimension stone quarries (Fig. 2).
Structural Interpretations. Previous workers have proposed two different structural interpretations to account for the distribution of marble and enclosing schists in the Tate-Marble Hill Belt. Bailey (1928) proposed that older rocks (his Carolina Gneiss and Roan Gneiss) had been thrust over the marble from the east leaving the marble at Marble Hill as a partial window. Bailey (1928, p. 55, 56, 122) recognized no clear criteria for the fault within the Tate Quadrangle, but apparently considered it a necessary extension of a fault previously mapped in the Ellijay Quadrangle by LaForge and Phalen (1913).
Fairley (1965) interpreted the structure as a syncline overturned to the west and later subjected to a second period of deformation that twisted the eastern limb of the

62

N

A
TATE DIMENSION STONE QUARRIES

NEW YORK MINE
I
I I
l
?.

' ,!'
A
?

Marble Hill Hornblende Schist Murphy t1arb 1e FIGURE 2. Geologic map and section of the Tate-Marble Hill Area.
63

fold into the sickle or hook-shaped map pattern displayed at Marble Hill. He further suggested that the marble passes laterally into schist by facies changes and interfingering so that, in effect one limb of the syncline is exposed as marble, the other as schist.
Earlier structural interpretations of the marble were part of larger pictures that involved regional patterns, the mapping of schist units, and long range correlations. The difficulty of the mapping and correlation is attested to by Bailey's statement (1928, p. 56) that "there seems to be no way of distinguishing between rocks belonging in the Great Smoky Formation and those belonging in the Carolina Formation ." Lithologic descriptions of the various schist units sound monotonously alike and it is difficult to find clear-cut criteria for distinguishing them. Except for the marble and the overlying hornblende schist there seems to be no good marker bed within the Murphy Belt rocks in the Tate-Marble Hill area.
In this report a structural interpretation of the Murphy Marble and overlying Marble Hill Hornblende Schist is given independently of the enclosing schists. It is assumed that the marble and overlying hornblende schist are derived from original sediments that had lateral continuity within, but not necessarily outside the Tate-Marble Hill area.
The interpretation is summarized in the cross-section (Fig. 2). Long Swamp Valley at Tate is underlain by a south plunging anticline. The hinge of a second anticline is exposed at the Amicalola Quarry about one mile east. The area between is underlain by a broad syncline interrupted by minor folding. Both anticlines are overturned to the west and are cut by thrust faults that strike and dip sub-parallel to the axial planes of the folds.
A key to this interpretation is the rocks at the Amicalola Quarry. Bailey's map (1928) shows the marble at this quarry to be completely surrounded by hornblende schist. Fairley's map (1965) shows the hornblende schist on the west side only. Recent core drilling at the Amicalola Quarry shows that the horn biende schist does indeed crop out on both sides of the marble and further that the schist on the west side dips eastward demonstrating that the fold is overturned .
A small valley runs north from the Amicalola Quarry to join the west branch of Long Swamp Creek at Marble Hill. Core drilling on the west side of this valley shows marble at least 200 meters (650 feet) lower in elevation than marble on the east side at the Amicalola Quarry. This can only be explained by a steep fold or fault, or both.
Economic Geology
The Georgia Marble Company operates mines and quarries for the production of dimension stone and a wide variety of crushed and ground stone products. Dimension stone has been produced for more than a hundred years. Significant production of crushed and ground products began with the formation of the Calcium Products Division in 1947. Today the production of crushed and ground products exceeds that of dimension stone in both volume and value.
Dimension Stone. The Georgia Marble District is famous for dimension stone which was used for the state capitol buildings at Rhode Island, Minnesota, the New York

Exchange and many other famous buildings. Since 1947 the production of dimension stone at Marble Hill gradually diminished and finally ceased. Today all dimension stone is produced from open quarries in Long Swamp Valley at Tate.
Most production is of Cherokee Marble, but Messotint and other graphitic varieties are also produced. There is no current production of Etowah Pink, Golden Vein, or White Georgia as dimension stone; however, these varieties are produced in a unique process for split-face ashlar.
Dimension blocks averaging about 16 tons are cut by Iine drilling using conventional quarry bars on flat floors. Blocks are broken free with wedges and feathers (Power, 1975) and moved across the quarry floor with tractors and loaders to where they can be lifted out with fixed derricks. The blocks are then graded and sent either to storage, the mill, or to waste. Some of the "waste" blocks can be partially salvaged for split-face ashlar, but many are a total loss and must be dumped.
Quarry blocks may be rejected for color or, more commonly, structural unsoundness. Most dimension marble for buildings is cut into thin slabs (7/8 inch to 1-1/4 inch thick) for use as veneer or in curtain wall construction. No cracks, fractures, or other weakness along which the slabs can break or part can be tolerated. It is important to the producer that such weakness be discovered before money is wasted in sawing the block . The result is that commonly more than one-half the blocks removed from the quarry are rejected. When this loss is combined with kerf cuts and other inevitable losses it is common that only 25 percent or less of the stone removed from the quarry is actually sent to the mill.
Some of the rejected blocks can be partially salvaged for split-face ashlar. Blocks are sawed into slabs of modular thickness (generally 2-1/2 inches or 5 inches thick). which are then split into oblong slabs that can be laid brick-like in courses with the broken face of the slab showing.
The price of split-face ashlar is such that it is uneconomic to quarry blocks solely for the production of split-face and therefore production is viable only from blocks that are "waste" as dimension stone. However, the demand for Etowah Pink, and White Georgia ashlar is so high that Georgia Marble Company has developed a special technique for producing these varieties even though there is no current production of dimension stone in these varieties.
Circular diamond saws mounted in a special frame cut a series of parallel grooves or kerfs in the quarry floor spaced 2-1/2 inches or five inches apart and about four inches deep. The stone between the kerfs is broken out and becomes split-face ashlar. Pink stone is produced at Tate; the White Georgia at Marble Hill.
Calcium Products Division. The Calcium Products Division of Georgia Marble Company produces crushed and ground marble at Marble Hill. Two abandoned and one active underground mine are in a lense or prism-shaped layer of stone approximately 60 meters (200 feet) thick and 600 meters (2000 feet) wide (east-west) that dips southward about 18 degrees. The mines are of the room and pillar type . The active New York Mine is multi-level with stacked pillars (Fig. 3).
The older workings have random pillars and varying room height. Current mining plans call for 12 meter (40 foot)

64

B

B'

FIGURE3. North-south cross section through New York Mine. 65

pillars on 24 meter (80 foot) centers lined up along strike. Initial openings are 7.5 meters (25 feet) high and plans call for lifting floors until the total room height is 36 meters (60 feet). Succeeding levels will be spaced 30 meters (100 feet) apart thus leaving 24 meters (40 feet) of stone between levels. Because of the dip of the stone only two-thirds of each level will directly underlie the level above . Access to deeper levels will be by minus 10 percent ramps. All haulage at the present time is by diesel trucks.
The stone varies in color and quality both across and along the layering, but chiefly across . Each level will extend from foot wall to hanging wall thus ensuring accessibility to all grades of stone.
Stone from the mine is delivered to one of three primary crushers feeding three plants or plant compexes. One complex manufactures a complete range of products requiring high r~flectivity and whiteness; a second produces dry-ground products with less stringent color requirements; the third produces screened products only and has the least stringent color requirements . Because the footwall rock is generally off-color marble it is always possible to obtain additional "no color standard" stone by extending any level into the foot wall.
Products are produced and sold on the basis of size and brightness (i.e. whiteness). The division produces marble in sizes ranging from less than ten microns to boulders weighing several tons. Boulders and pebble-size particles are sold for decorative landscape purposes. Finer screened products are used in synthetic marble, floor tile, welding rods, aggregate, and other uses. Ground products are used as fillers and extenders in paint, rubber, plastics, putty, chewing gum, paper, and a host of other products. As grinding and classification techniques improve new uses are found .
Ground products are produced dry in roller or tube mills and wet in ball mills. Generally the finer ground products require higher brightness. The higher brightness is achieved partly by controlling the feed, but partly also by the grinding itself because there is an inverse relationship between particle size and brightness. Price is also inversely related to particle size, the finest and brightest material commanding the highest price.
Cove Mountain Belt
The Cove Mountain Belt lies on the east side of Long Swamp Creek about 3 km (two miles) east of Jasper, Georgia. Dolomitic marble crops out along a 5 km (three mile) strip on the east valley wall and dips about 20 to 30 degrees east. The layer reaches a maximum thickness of nearly 100 meters ( 300 feet) in the middle of the belt, but thins to the north and south finally pinching out.
Schist typical of the Brasstown Formation encloses the marble . Bailey (1928) and Fairley (1965) suggested that the belt is an infolded, isoclinal syncline because it is enclosed by similar schist both above and below the marble. Fairley also found possible graded bedding on the east side of the belt suggesting overturned beds.
Core drilling shows calcite marble in the central layer at several places. This is consistent with the sync Iina I hypothesis if the original stratigraphic suggestion was similar to that in the Tate-Marble Hill Belt, that is, calcite

marble overlying dolomitic marble. At the south end of the belt the layer is split by a ten
meter (30 foot) layer of schist. This could be interpreted either as interfingering of original layers or as an overlying schist in the core of a syncIine. Because the marble and schist clearly interfinger at the northern end of the TateMarble Hill Belt several kilometers (miles) south (Fairley, 1969) the author prefers this interpretation for the interlying schist at the Cove Mountain Belt.
Lithology
The marble is distinctly layered in most places. White to light gray, fine-grained, dolomitic marble is interlayered with brown mica schist and gray, medium to coarse-grained, calcite marble. Accessory minerals include brown mica, amphibole, and quartz.
The dolomitic marble is most abundant. It occurs in layers ranging in thickness from a few centimeters to nearly ten meters (30 feet). The thickest and most persistent layers are near the upper contact of the formation and are the only layers mined commercially.
Brown, calc-mica schist occurs in paper-thin laminae and layers up to a meter (3 feet) in thickness. In places brown mica is also disseminated in the dolomitic layers giving them a brown or tan tint. Schistosity is parallel to layering and in places the rock splits readily into thin slabs along planes of schistosity.
Gray, medium-to coarse-grained marble is interlayered with dolomitic marble in the central part of the belt. This rock strongly resembles the graphitic marbles of the Tate area.
Mines and Quarries
Dimension stone was quarried from the Lincoln quarry at the north end of the belt in the nineteenth century (McCallie, 1907). The quarry was inactive when Bailey (1926) studied the area and has not been worked since. It is now overgrown by vegetation.
The Georgia Marble Company operates an Lmderground mine in the belt due east of Jasper. Only white dolomite marble is mined. It is crushed and screened to make terrazzo and other decorative chips. Waste material is sold for agricultural limestone.
The mine is room and pillar type and follows a layer averaging 6 to 7.5 meters (20 to 25 feet) thick down dip. Stone is hauled by truck to a crushing and screening plant at the mine portal. Finished products are then truck-hauled to a bagging and shipping plant on the railroad south of Jasper.
Whitestone Belt
The geology of the Whitestone area was described by Power and Reade (1962). The geologic map (Fig. 4) is from that report.
The Murphy Marble crops out in the valley of Talona Creek and along the east wall of that valley. It dips steeply eastward and is complexly faulted. It is overlain by quartzbiotite schist that is tentatively correlated with the

66

0
FIGURE 4- Geologic Map of the Whitestone Belt. (Power and Reade, 19.62)
67

Brasstown Formation and overlies chlorite-sericite phyllite that possibly correlates with the Mineral Bluff Formation. If these tentative correlations are correct the section is overturned to the west and is probably the overturned limb of a syncline.
The marble is divisible into two units: a medium to finegrained dolomite marble and a coarse-grained calcite marble.
The dolomite marble overlies (but may be stratigraphically lower than) the calcite marble and is in contact with the quartz-biotite schist. Many thin interlayers of talc-phlogopite schist occur near the contact. The main body of the unit is a more-or-less massive, white to gray, fine-grained dolomite marble that is mined for decorative chips and agricultural limestone . Near the contact with the coarse-grained calcite marble, the dolomite is increasingly siliceous and contains numerous pods and laminae of white talc and scattered pale tan to colorless tremolite crystals. The dolomite marble appears to be 150 to 200 meters (500 to 600 feet) thick, but some of this thickness may result from repetition due to faulting.
The coarse-grained, calcite marble closely resembles dark Cherokee or Messotint Marble of the Tate area. It is several hundred meters (feet) thick and underlies most of the flood plain of Talona Creek at Whitestone.
The marble crops out along the east wall of the valley for a distance of more than 3 km (two miles). Most of the area has been open to mining due to the opening of eight connecting mines. The mines are random room and pillar type. Pillars are generally twelve to twenty meters (38 to 64 feet) in diameter with rooms about the same size. Practice is to take up floors after initial openings are made thus increasing ceiling height. Some ceilings in old, abandoned sections are more than 50 meters (150 feet) high, but current practice is to limit ceiling heights to 20 meters (64 feet).

The Georgia Marble Company is currently the sole producer in the area. The company operates three plants and four mines for the production of decorative chips (terrazzo, landscape chips, etc.), agricultural limestone, and dry-ground fillers for use in latex and similar products.
References Cited
Bailey, W. S., 1928, Geology of the Tate Quadrangle, Georgia: Georgia Geological Survey, Bull. no. 43.
Fairley, W. M., 1965, The Murphy Syncline in the Tate Quadrangle, Georgia: Georgia Geological Survey, Bull. no. 75.
- - - - -- , 1969, Stratigraphy and structure of the Murphy Marble Belt in parts of northern Georgia: Georgia Geological Survey Bull. no. 80, p. 89-120.
Georgia Geological Survey, 1976, Geologic Map of Georgia: Atlanta, Georgia Geol. Survey, 1:500,000.
Hurst, V. J., 1955, Stratigraphy, structure and mineral resources of the Mineral Bluff Quadrangle, Georgia: Georgia Geological Survey, Bull. no. 63.
Keith, A., 1907, U. S. Geological Survey Atlas, N antahala Folio: no. 143.
LaForge, L., Phalen, W. C., 1913, U.S. Geological Survey Atlas, Ellijay Folio: no. 187.
McCallie, S. W., 1907, Marbles of Georgia: Georgia Geological Survey, Bull. no. 1.
Power, W. R., Forrest, J. T., 1973, Stratigraphy and paleogeography in the Murphy Marble Belt: Amer. Jour . Science, v. 273, p. 698-711 .
Power, W. R, Reade, E. H . (1962). The Georgia Marble District: Guidebook no. 1, Georgia Geological Survey.
Reade, E. H., 1965, the geology of the Tate-Marble Hill area [abs.] : Georgia Acad. Science Bull. 23, p. 69.

68

GEOLOGY OF KYANITE
Dennis Radcliffe Department of Geology
Hofstra University Hempstead, New York

Introduction
Kyanite has the chemical formula AI 2 Si0 5 with an equal ratio of alumina to silica (AI 2 0 3 Si02 ). The theoretical chemical analysis is 63 percent Al 2 0 3 , 37 percent Si0 2 . Kyanite, andalusite and sillimanite are all polymorphs of AI 2 Si05 and occur under varying pressure and temperature conditions in regionally metamorphosed rocks (Figure 1).

Crow'"'
--~ - - - ,."_.,_~pal.~..\l

._.,.

L __

BARS 20,000

STABILITY FIELDS OF A~ Si05 POLYMORPHS

KYANITE

8[01t81A

..... ~ 10,000
"!":!''
"-

SILLIMANITE

FIGURE 2 - location of the principal kyanite quartzite deposits.
Industrial Use of Kyanite

0

500'

IOOO'C

FIGURE 1- Pressure and temperature condition for kyanite formation.
Kyanite occurs frequ ently in metamorphic rocks such as gneiss, schists and quartzites. In these rocks, kyanite normally does not exceed 5-10 percent of the total mineral abundance, but under certain conditions it may reach 40 percent. Rocks containing 30 percent kyanite are not uncommon from an exploration viewpoint. However, kyanite is produced commercially from only 2 small districts - Graves Mountain, Georgia (C-E Minerals), and Baker and Willis Mountain, Virginia (Kyanite Mining) (Figure 2). This is due not just to the relatively small market (estimated at 100,000 TPY), but to the specifications demanded by the consumer end use (Table 1). Kyanite also exhibits differing properties depending on the local geology of the deposit, and only certain classes of deposits meet industrial specifications. It is, therefore, necessary to thoroughly understand the end use in order to appreciate the commerciality of the deposits.

Kyanite has a range of uses, but the largest tonnage is consumed by the refractory industry, particularly in specialty refractories including : gunning mixes, mortars, castables and ramming mixes. In most of these applications (Table 2) the expansion characteristics of kyanite (Table 3) are extremely important during firing, as it can offset the shringkage of other phases and produce a stress-free volume-stable refractory. In a related application, the utilization of kyanite in mortars causes mortar expansion during firing. This then locks the refractory bricks tightly in place within the furnace or kiln.
The expansion characteristic of kyanite is related to the generation of mullite plus cristobalite (Figure 3) beginning at 2678 F. The resulting mullite and cristobalite not only has a greater volume than the original kyanite (Table 3) but is extremely refractory. The refractoriness is measured as the softening and deforming point of a cone shaped aggregate. This is known as the pyrometric cone equivalent (PCE). For pure kyanite a PCE of 40 can be attained but the impurities in the commercial concentrate reduces the PCE to the 36-38 range.
Clearly the least amount of impurities is most desirable in any commercial concentrate. In the case of kyanite, any impurities must not tend to lower the softening and melting temperature of kyanite. Thus

69

ORE*

TABLE 1 - Typical composition of ore and specifications of kyanite concentrates.

CONCENTRATE**

Analysis

Characteristics

Al 2 0 3 Si02 Fe20 3 Ti02 CaO MgO Na2 0+K2 0 L.O.I.

34 5 52.3
2.5 1.7 0.9 1.4 4.7 1.8

Al 2 0 3

Si02

(Total)

Fe
2

0

3

Ti02

CaO

56.6 40.0
1.0 (acid soluble) 1.0 0.5

MgO

0.5

Na 2 0+K2 0 0.1

L.O. I.

0.3

ca. 91 % Kyanite ca . 5% Quartz ca. 2% Rutile and Fe minerals ca. 1% Hydrous silicates

PCE:
S.G.: raw calcined
Hardness: Shape:

36-37
3.6 3.0 5-7
Bladed

* Kyanite Quartzite contaminated with schist **From C-E Minerals Sales brochure - Mineralogy calculated from analysis

tOO 9H 90

80

10

60

3801l Melting pc>nt of silrcd (cristobdlrte) 1723'C 3133'F Melting pornt of dlumind (corundum) 2050'C, 3722'F

40

30

282

31i00-

:i4oo
3344 3362

liquid

mullite solid

solution plus liquid-- -~

1900

I 1840

r- J200 crislobalite
3133 plu1 liquid

mullite plus liquid*

- --msoululltiitocnsolid

1800

1100

3000 2903 2800
ms
2600 -
.F
240 S.5 10

cristobalitc plus mullitc

u

c:

0 u
-(I)

-Q)
c:
(.1

c:

>..

0

.X

tridymitc plus mullitc

--"

(..I.)

ai

::J

> a..

0

~ t

20

30

~0

50

so

10 718

corundum

1600

plus mullite

solid solution

1500
c

1400

. Weigill pet cenl AI103 I

80

90

100

FIGURE 3- Phase diagram for the AI 2 0 3 -Si02 system.
70

GENERAL EXTRACTIVE METALLURGY OF KYANITE

ORE

~~
PRIMARY CRUSHING
and SCREENING
'if 11h"
STOCKPILE
..,
'
RODMILL

~

CLASSIFIER

+35 mesh

,~ -35 mesh

DES LIMING

-,.

MICA FLOTATION

It'
PYRITE FLOTATION
,if KYANITE
- - CONDI.'l'JON I NG QUARTZ SUPPRESSION

PETROLEUM SULFONATE
or PINE OIL

KYANITE FLOTATION
'
DEWATERING
It DRYER

ACID or BASIC CONDITIONER

I
BULK SHIPPING
BIN

MAGNETIC 'SEPARATION

t

I

I

BAGGING PLANT

CALCINING KILN

I
GRINDING PLANT

FIGURE 4 General schematic for kyanite mining and beneficiation .
71

TABLE 2 - Summary of industrial applications of kyanite.
Uses of Kyanite
1. REFRACTORY SPECIALTIES Monoliths, Castables, Plastic Ramming Mixes, Cements, Mortars Kyanite constitutes 10 to 40% of mixture and offsets shrinkage of clay binder
2. Ceramic Bodies
3. Foundry Molds
4. Brake Shoes
5. Welding Rods
6. Catalytic Converters
7. Aluminum-silicon Alloys. Future Use?
alkalis, iron and other fluxes should be either absent or of very low concentration. Alkalis are particularly harmful in refractories, and their occurrence more than any other factor restricts the types of deposits which can be utilized by industry.
The shape of kyanite crystal is also important, and the sharp rectangular outline of the kyanite cleavage fragments lends itself to bonding with aluminous clays and other ingredients of refractory grogs. The effect .is similar to the strength imparted by lath-shaped plagioclase feldspars to ophitictexture basalts.
The largest size of domestic kyanite sold in the U. S. is 35 mesh. Many refractory manufacturers would

prefer a coarser size, but the availability of size is controlled by metallurgical flotation technology. It is just not possible to float kyanite grains which are larger than 35 mesh. Since domestic kyanite must always be beneficiated by standard flotation metallurgy, a 35 mesh upper limit is the effective cut off.
Kyanite also has significant uses because of its chemical, as well as size or shape, characteristics. These applications include mostly refractory ceramics, such as: spark plugs, the substrate of automobile catalytic converters, porcelain enamels, and a host of other uses, but generally in small tonnages. In this application kyanite is normally sold calcined and ground to 100, 200 ro 325 mesh. Calcination is a solid state subsolidus recrystallization into different phases. For kyanite this occurs in the range 2678-2903 F (Figure 3) when mullite and cristobalite are formed from kyanite in a process similar to metamorphism.
Specifications of Kyanite Concentrates
The commercial specifications of the Graves Mountain Kyanite are shown in Table 1. Of great importance is the low level of fluxing elements (alkalis, etc.) which otherwise degrade the refractory bodies. The effect of 1 percent Ti02 is debatable as it can be a slight advantage or a slight disadvantage depending on the application. Iron oxide has a negative effect in the refractory, in that it causes "rafting" and "spotting" within the body, and can lead to carbon monoxide poisoning. The lowest possible iron contents are most desirable, but 1 percent Fe2 0 3 is not serious. Lower levels of iron are achievalbe but not economically viable. Thus, it is quite possible to generate a 0.5 percent FE 2 0 3 concentrate, but the recovery loss is so large that price increases would be necessary and consumers are not willing to pay the price for the higher quality.

Maximum Volume Expansion of AI 2 Si05 Minerals on Inversion To Mullite and Cristobalite at 1300 C (2372 F)

3A1 Si0
25

3A1 2 0 3



2Si0
2

+

Si02

(63.0% A1 0
23

(71.8% Al 0 )
23

100% Kyanite----.- 88.4% Mullite

+ 11.6% Cristobalite

Andalusite Sillimanite Kyanite

Cell Volume (A03 ) 342.26 331.42 292.90

t:J. Volume~Mullite - 2.36 + 8.48 +47 .00

t:J. Volume~Cristobalite + 21.17 + 32.01
+ 70.53

L%t:J.~Volume
0.1 3.4 17.1

TABLE 3- Illustration of the expansional characteristics of kyanite.

(Realized volume expansions are less due to voids)

72

The alumina content is quite revealing as pure kyanite should contain 63 percent AI 20 3, yet most commercial concentrates assay in the range of 56 to 57. This relates to the difficulty of floating kyanite (silicate) from quartz (silica). The normal method is to add a silica depressant to prevent quartz from floating, but the depressant also acts on kyanite. Therefore, only a small amount of suppressant can be used resulting in the floating of some quartz together with kyanite . As a result all kyanite concentrates are contaminated with quartz, generally in the range of 5 to 10 percent.
Supply and Demand of the United States Market
The actual size of the U.S. kyanite market is very difficult to determine accurately because sales figures are held as industrial secrets by the two producing com panies. It is possible to estimate production based on the consuming industries, and Radcliffe (1976) estimated 100,000 TPY plus or minus 10 percent depending on the state of the national economy. However, this production figure includes export sales which is another "secret", variously estimated at 10 to 20 percent of total sales.
The future demands on the kyanite producers is also difficult to determine. While the U.S . steel industry is projected by economists to grow at an annual rate of 8 percent per year for the next 5 years, improvements in refractory technology tend to reduce the unit comsumption of refractory raw materials. The main hope for expanded demand is in the development of new applications. The most promising appears to be in the manufacture of AI-Si alloys by heating kyanite with coke and then converting the carbide to the alloying elements in an electric arc furnace.
The state of the supply precludes the entry of new companies into the field of production. The two current suppliers together have an approximate 100 percent excess production capacity. This overwhelming excess originated in 1973 when kyanite production fell well behind demand due to rapidly expanding sales in the pre-oil embargo economy. Both producers expected a continuation of strong and expanding sales and moved into a position to meet grossly exaggerated future sales forecasts. In addition to this excessive production capacity, both producers have very substantial ore reserves (Bennett, 1975, Espenshade, 1973, and Potter, 1975) suitable for at least 50 years of continuous operation.
Mining and Benefication of Kyanite Deposits (Figure 4)
Mining is by open pit, utilizing standard techniques. Ten to twenty foot benches are drilled with percussion equipment. The 6-inch holes are packed with an explosive mixture of fertilizer (ammonium

nitrate) and fuel oil, and detonated with dynamite.

The shot is loaded with shovels, transported by truck

to a primary crusher, normally backed up with a

secondary Symans Cone crusher. The minus 1 to 1Y.

inch rock is then stockpiled for feeding into the

flotation plant.

The ore is fed by conveyer into a rod mill in

closed circuit with a classification system which is

generally a rake. The classification is aimed at 28 mesh

but due to variation in ore characteristics may become

35 mesh. After desliming, the ore is conditioned (pine

oil and xanthate at Graves Mountain- amine collector)

for removal of pyrite, sericite and any other clay

minerals. Kyanite can be flotated either in an acid or

basic circuit. At Graves Mountain an acid circuit is

employed, and after removal of pyrite and sericite,

the slurry is reconditioned with sulfuric acid and

petroleum sulfonate followed by normal counter current

flotation circuitry.

The resulting kyanite concentrate contains clean

..--

grains and kyanite crystals coated with varying amounts

of iron oxide contaminants. Wet or dry magnetics re-

duce the iron content to a level of approximately

1 percent Fe2 0 3 resulting in a recovery loss of kyanite. To achieve lower levels of iron, it is necessary to process

very pure ore (an extravagence) or to magnetize the

remaining iron by heat treatment. This latter is costly

and in addition, the small gains in iron are offset by

excessive losses in kyanite grains which, at this level,

contain only very small patches of iron. The iron coating on the kyanite crystals through-

out the deposit is sufficiently tenacious such that a

ferrugenous pseudomorph after kyanite is suggested.

The nature of the iron contaminant is normally thought

to be low temperature goethite (hydrous iron oxide).

as it does not normally give an X-ray pattern. However,

recent Mossbauer studies have shown it to be an

aluminum-substituted hematite, perhaps supporting (?)

a hydrothermal (rather than weathering) origin for most

of the strong iron alteration products so prevalent in

many parts of the deposits.

Petrography and Metallurgy
Kyanite crystals vary in texture from one deposit to another, and an elementary analysis of petrographic characteristics (Figure 5) quickly leads to a prediction of metallurgical behavior and quality of concentrates. The following types of information can normally be obtained from a standard 0.03mm petrographic thin section:
(a) The grain size of the kyanite crystal and probable liberation point of the intercrystalline rock aggregate.
(b) The frequency of inclusions within the kyanite grain. For example, the kyanite grains in Figure 5A are essentially free of inclusions, whereas those in 5C are loaded with impurity inclusions. In this latter case a theoretic 100 percent recovery and a

73

FIGURE 5 Photomicrographs of kyanite ore. Rock sections are 0.03mm thick and show the distribution of kyanite (ky), quartz (qtz), quartz inclusions (qtz incl in SCI and biotite (biot) .

theoretic 100 percent kyanite grain concentrate would still only yield a substantially low 40-50 percent A1 2 0 3 compared to 2A which could be essentially 63 percent.
(c) The nature of the distribution of undesirable phases, such as iron-bearing micas (biotite). In Figure 5D the micas are attached to the kyanite grains, and a minimum 48 mesh grind is necessary to liberate the two phases. In Figure 58 a relatively clean 35 mesh kyanite concentrate might be predicted, however, the biotite (lower right) is included within the kyanite grain and may well contaminate the fi nal product.
Geology of Kyanite
Kyanite occurs principally in two petrologically distinct facies, both of which are metamorphic in origin. It occurs either in a gneiss (or schist) associated

with feldspar, biotite and garnet, or in a quartzite where it is associated with pyrite, rutile and pyrophyllite.
Every metamorphic petrologist is familiar with kyanite grade metamorphism of pelitic assemblages, but such rocks are generally low in kyanite (less than 10 percent) because the available alumina is complexed in other phases. The principle deposits of kyanite in North America were described by Radcliffe (1975) as follows:
(a) Kyanite-quartzite belt of the southeastern states (C-E Minerals and Kyanite Mining) .
(b) Kyanite-garnet-biotite gneisses of the Wanapitei-Crocan Lake-Timiskaming belt in Ontario-Quebec (Kyanite Mining). (c) Kyanite-biotite-garnet gneisses of the Clearwater River area (middle fork) in north-central Idaho (Ethyl Corporation). (d) Kyanite-quartzite near Ogilby, California.

74

CaO MgO
L.O.I. Mineralogy

TABLE 4- Origin of Kyanite Quartzite
Hypothesis of Metamorphism of Transported Sandy Kaolin Deposit
Typical Kyanite Ore
34.5
52.3
2.5
1. 7
0.9
1.4
4.7
1.8
KYANITE AI 2 Si0 5 QUARTZ lazulite pyrite ilmenite goethite rutile muscovite pyrophyllite

Typical Sandy Kaolin
34.2
47.5
1.0
1.5
0.2
0.4
1.1
14.0
KAOLINITE AI 4 Si 4 0 10 (0H) 8 QUARTZ apatite pyrite ilmenite goethite rutile muscovite kyanite and sillimanite

The Geology of Graves Mountain
Graves Mountain (Figure 6) was first described by Shepard (1859), Watson and Watson (1912), then by Prindle (1935), followed by Furcron and Teague (1945), more recently by Crickmay (1952) . However, Hurst (1959) provided the first comprehensive mapping and detailed analysis of the deposit. Espenshade and Potter (1960) provided the first regional analysis of the southeastern deposits, and Hartley (1976) gave the up-to-date account of the mining and processing operation. It is somewhat ironic that while Graves Mountain is one of the very best kyanite deposits, it is known primarily for its very fine, black lustreous multiple-twinned rutile crystals. These latter are illustrated in most modern mineralogy text books and also can be found in all good mineralogy museums anywhere in the world.
Graves Mountain is an elongate monadnock (Figure 6) reaching 900 feet above sea level and projecting 400 feet above the general level of the surrounding Piedmont. It is about 7200 feet long and 200 feet wide trending in a northeasterly direction, directly parallel with regional strike. The main ore zone how-

ever is about 800 feet wide. Graves Mountain occurs within gneisses and schists of the Little River Series which have been correlated with rocks of The Carolina Slate Belt and dated as Lower Paleozoic. These sequences are regarded as metabolcanics or at least interbedded metavolcanics and metasedimentary facies. Because of this association the Graves Mountain kyanite quartzite is frequently assigned as being volcanic in origin (Hurst, 1959, Espenshade, 1960, and Hartley , 1976), but this is not necessarily correct.
Graves Mountain itself has an average N70E strike and consists of interbedded and somewhat lensoidal kyanite quartzite and sericite schist (with or without kyanite). The individual quartzite units vary in thickness from 6 to 50 feet. The zones dip northwesterly at varying angles, generally 60-70. The quartzites are massive and the schistose rocks display cleavage parallel only with the apparent bedding direction of the quartzite bodies. Therefore, it is difficult to determine whether or not Graves Mountain is an inclined series of interbedded rocks, an overturned syncline or an over-turned anticline.
Examination of Graves Mountain rocks can

75

N

I I

I \
\

r

,,. .... --/ I
I
\

"" ....

I
___ ,\

=-=:-' -- o

4;oo

.! _- - : - : : -' - : '
F" E " T

Soo

' ' ' ,, .... -)

' \

/ I /

, ' ---'-'
I

__, --' \

I

\

......

;
, '
,
_, /'

,--...... ' .~',

- ' ,' .,~ ~ /

"\ l <o(j '

..,.-,

rI

\. ... - - / ' ~t;:)!!>

I 'I

, I / t,"' I

I

I
I

/!; n\\ ,/"' '- -' ; \

0 /

1 /

I

I ' "' u \\~.... I \

I

r"-\" I

1
.,./

\ _,.. 1 / I

;

1

/'
/

~tj,iP

I

.... I

1 /

/

I ,

-~ -- --......,,
--- r .. ,... ," /' .... /
'- - -
_-- -

'
',

.. :.__~f\fllfltr,.J(f\

I
t

/I I

,/ / ,
/ /,:,, '

- t~ -

- --

~

\ \

t -,

- -'1 I

- -----"'1

I

1

,~

- - - ,ai\1\"'e - \.. \' ~'~

,/1 / "I I

~

"~'"
'I~CI~ ma"i'-n--'--p?i-t

Q~' \
,';?~>11

~\

I(/

, v;_.

'G-.Uu~ j,'~~ }. ~ :~ ;..-

-.
-
\_\.\
~

.,

. .

I
I
I
_
-

I
~

, ,
/
' 1 ....

/

;

'1,' .., ., , 0o' ,' ~ - __ - - - - -- -

"

-

; 1

1

'' \ \

__ ,.,

- , ..

~ I

' '

'..1 ' \ -... ,<, .,\

' '

\ .,..

I

.-

\

-

I I I I I
\.,_ '
,--
'f\
rob able fault

I

I I

__________ , "

limit of kyanlte quartzite unit

FIGURE 6- Generalized geologic map of Graves Mountain. Opalescent refers to quartz porphyroclasts in a kyanite-poor pyrite quartzite band and lazulite refers to a lazulite-bearing kyanite pyrite rutile quartzite. These two units bound the main ore zone.

76

Many other deposits exist but most of these are eclipsed by the enormity of these first three deposits. Because of the restrictions imposed by the application of kyanite concentrates in refractories, the kyanite quartzites of the southeastern states are by far the most suitable deposits.
Kyanite quartzite occurs in a discontinuous belt from south central Virginia to northeast Georgia (Figure 2), a distance of almost 700 miles. No less than 14 deposits have been described by Espenshade and Potter (1960) in this belt. Ironically, the best deposits occur at the extremities of the belt. Kyanite Mining Corporation of Dillwyn, Virginia, operates five plants at their two mines located at Willis Mountain and Baker Mountain. Kyanite Mining are currently building a sixth plant at East Ridge, immediately east of Willis Mountain.
At the southern extremity of the belt, C-E Minerals, a subsidiary of Combustion Engineering, Inc., operates a shipping plant at Little River, and a processing plant at its Graves Mountain mine. The geology of the Graves Mountain deposit is fairly typical of the belt and a short description of this area illustrates the geology of this class of deposits.
Genesis of Graves Mountain
The origin of kyanite quartzite of the Graves Mountain type is generally thought to be metamorphism of volcanic rocks. A different origin is now suggested.
Hurst ( 1959) described the deposition of tuffaceous sediments and intercalated gravels along a zone passing through Graves Mountain. Espenshade and Potter ( 1960) describe the replacement of an igneous rock by aluminum and silicon, and Hartley (1976) described the metamorphism of vitric tuffs.
The evidence for a pre-metamorphic igneous origin is based on the occurrence of the rounded opalescent quartz grains in the kyanitepoor pyrite quartzite which occurs at the northwest margin of the main ore zone (Figure 6). The opalescent quartz is supposedly bipyramidal and hence igneous in origin. The grains are rounded and this is explained by magmatic resorption. It requires a fair amount of imagination to observe bipyramidal quartz. The present writer has never observed a quartz grain which is convincingly bipyramidal.
If this opalescent-quartz bearing quartzite were located elsewhere, such as the Blue Ridge of Virginia, it would be called a metaconglomerate which is much more plausible. It is even possible for a clay-bearing sand to contain large fragments of bipyramidal quartz, weathered from an igneous rock. Metamorphism of this sedimentary facies might produce the rock unit observed at the margin of the main Graves Mountain ore zone.
If we strip away the complications imposed on the rock series by secondary hydrothermal activity and

weathering, we find that the mineralogy of the Graves Mountain-type kyanite quartzites is essentially the same on a world-wide basis: kyanite, quartz, pyrite, rutile and lazulite. Occasionally substitute minerals, based on local variations of pressure and temperature, are also found, such as: andalusite for kyanite, goyazite for lazulte, and occasionally ilmenite. This mineral assemblage tends to be intercalcalated with a schistose assemblage characterized by a potassium-bearing mica (sericite), quartz, pyrite and generally some kyanite in small quantities. Whether or not kyanite occurs is based on the relative chemical mass balance of alkalis and alumina in this facies.
The rather distinct mineralogy must be significant, especially that of the quartzite. It is proposed that the kyanite and quartz crystallized from a sandy kaolinite sedimentary facies under metamorphic conditions. The occurrence of pyrite reflect the depositional environment, and the occurrence of rutile and lazulite represents a heavy mineral suite or possibly a phosphatic horizon
The late-Cretaceous, early-Tertiary Kaolin deposits of Georgia, Alabama and the Carolinas form a subcontinuous arcuate belt and are known to be deposited in coastal or lagoonal environments. The kaolinite was weathered from the exposed crystalline Paleozoic piedmont rocks and transported towards the ocean where it was deposited in a relatively low energy environment. Such environments are generally sandy, may be characterized by dark-colored pyrite-bearing clays in the estuarine marshes and lagoons and normally contain a small heavv mineral suite of ilmenite (FeTi0 3 ), rutile or anatase (Ti02 ), and apatite, (Ca5 (P04) 3 (0H), togehter with kyanite and sillimanite. Occasionally phosphatic sand horizons (associated with the near shore environment) also occur.
From an industrial viewpoint the occurrence of anatase and altered pyrite in the Georgia kaoli.11 deposits and their passage laterally or vertically vertically into either a sandy facies or a pyritic sericite clay facies is well known.
Metamorphism of this normal sedimentary facies will produce the exact mineralogy observed in the Graves Mountain-type of kyanite quartzites (Table 4). Further, it will also produce the interbedding with sericite pyrite schist.
If this hypothesis is correct then the occurrence of kyanite quartzites in the near linear belt extending from Graves Mountain, Georgia, through the Carolinas into Virginia may well represent a paleo-shoreline of lower Paleozoic age. It would also follow that the associated volcanic rocks which are prevalent in Georgia and parts of the Carolinas were deposited near-shore and possibly under shallow water conditions. This is somewhat similar to the China-Korea mainland and its association with the continental shelf of the Sea of Japan and the Japanese volcanic archepelago at present.

77

lead to descriptions of at least 8 distinct, important rock types. Detailed analysis shows however that they are only varients of the two basic rock types. The apparent difference results from later weathering and/or hydrothermal alteration.
The two original pre-alteration rocks of Graves Mountain were characterized by the omnipresence of pyrite (FeS2 ) and generally small quantities of rutile (Ti0 2 ). Weathering and/or hydrothermal alteration differentially leached pyrite which then generated staining iron-rich solutions. The staining solutions preferentially coat kyanite and cause serious metallurgical problems. The change in rock type is illustrated below:

blue kyanite pyrite quartzite + iron oxides
white sericite pyrite schist oxides

red kyanite quartzite red sericite schist+ iron

This four-fold rock division is further complicated by a zone of kyanite-pyrite-rutile-quartzite containing small amounts of lazulite, (MgFe)AI 2 (P04 )2 (0H) 2 . This latter occurs along the southeast margin of the bodies (Figure 6). On the northwest margin of the main ore body a zone of kyanite-poor pyrite quartzite contains bluish colored, rounded, opalescent quartz porphyroclasts (?) up to 10mm in diameter .
This series of rocks is further complicated by an intersecting network of hydrothermal massive quartz veins, resulting in localized pyrophyllite (Hurst, 1959) and locally recrystallized (?) kyanite. It can be observed that within at least 3 feet of the massive quartz veins, the kyanite in the kyanite-quartzite increases markedly in size. If this recrystallization process were operative it can also explain the selective occurrence of the large museum size (1OOmm diameter) back rutile crystals at the margin of the narrow (12-24") massive quartzite veins which are themselves thoroughly steeped in hydrous iron oxides. Rutile normally occurs as small (0.5mm diameter) red crystals. It is possible that the hydrothermal solutions dissolved pyrite (FeS2 ) from the quartzites. This could generate an iron-staining solution of sulfuric acid which would dissolve the smaller rutile crystals. The titanium later precipitated from solution as large, black iron-bearing rutile crystals at the margins of the quartz vein. This would explain

the juxtaposition of intense iron stammg and large rutile crystals, a fact borne out by direct field observations.
Bennett, P. J. and Castle, J. E., 1975, Kyanite and related minerals: Industrial Rocks and Minerals, 4th Edition, AI ME, pp. 729-736 .
Crick may, G. W., 1952, Geology of the crysta II ine rocks of Georgia : Georgia Geol. Survey, Bulletin# 58.
Espenshade, G. H. and Potter, D. B., 1960, Kyanite, sillimanite, and andalusite deposits of the southeastern states; U. S. Geol. Survey Prof. Paper.
-..,....--.,....-~ , 1973, Kyanite and related minerals; United States Mineral Resources: U. S. Geol. Survey Prof. Paper# 820, edited by Brobst & Pratt, pp. 307-312.
Furcron, A. S. and Teague, K. H., 1945, Sillimanite and massive kyanite in Georgia: Georgia Geol. Survey, Bulletin #51.
Hartley, M. E., 1976, Graves Mountain in Stratigraphy, Structure and Seismicity in State Belt Rocks along the Savannah River: compiled by T. M. Chowns. Georgia Geological Society Guidebook 16, Atlanta, pp. 42-52.
Hurst, V. J., 1959, The geology and mineralogy of Graves Mountain, Georgia: Georgia Geol. Survey, Bulletin # 68.
Potter, M. J., 1975, Kyanite and related minerals; Mineral Facts and Problems: U.S. Bur. Mines Bulletin# 667, pp. 579-590.
Prindle, L. M. , 1935, Kyanite and vermiculite deposits of Georgia: Georgia Geol. Survey, Bulletin # 46.
Radcliffe, D., 1975, Kyanite: Annual Commodity Review, Mining Engineering, V. 27, # 2, pp. 77-78. , 1976, Kyanite: Annual Commodity Review, Mining Engineering, V. 28, # 3, pp. 39-40.
Shepard, C. U., 1859, On lazulite, pyrophyllite, and tetradymite in Georgia : Am. Jour. Sci., V. 27, pp. 36-40.
Watson, T. L. and Watson, J. W., 1912, A contribution to the geology and mineralogy of Graves Mountain, Georgia: Virginia University Philosophical Society Bulletin, Science Series, V.1, pp. 201-221.

78