Analysis of Metals in Cement Kiln Dust
Using the Lithium Fusion Method

Robert J. Schoenberger Ph.D., P.E.
Principal
Imagineering Associates
Post Office Box 620
Uwchland, Pennsylvania 19480

Charles E. Buchanan, Jr.
President
Roan Laboratories, Inc.
Post Office Box 145
Holly Hill, SC 29059

ABSTRACT

The analysis of metals using hot plate. microwave and lithium borate fusion digestion has been investigated for four samples cement kiln dust. Results of analysis show that the standard hot plate digestion yields the lowest results or recovery of metals. Microwave digestion generally shows a slightly higher recovery of metals, but the significance of the difference can not be calculated until more samples are analyzed. Because of the presence of silica and alumina, the fusion method shows significantly higher recovery for chromium, nickel, zinc, potassium, calcium, and iron. The fusion vaporizes some constituents; lead, sulfur, vanadium and therefore the method is not usable for those constituents. The impact on cadmium is unclear and more investigation is needed.

INTRODUCTION

The manufacture of Portland and other cements has long been a primary part of the manufacturing sector in both developed and developing nation states. In the United States the manufacture of cement and the use of cement in construction have been an integral part of the construction industry. Cement is used widely in industrial, commercial and infrastructure projects. Operation of the cement manufacturing plants, as well as other heavy industry such as iron and steel, results in a product which is chemically changed from the raw material input. This change is the result of heating the raw materials to temperatures which generally exceed 2600 degrees F. To maintain quality control for cement products, the raw materials are pulverized to very fine but discrete particle sizes and thoroughly mixed in controlled chemical proportions. The fuel burning to achieve the high temperatures combined with the fine particle size of the raw materials results in the potential release of particulate matter which is defined as kiln dust. Particulate emission standards are recognized as necessary by EPA and New Source Performance Standards are given in 40 CFR Part 60, Subpart F¹.

In the last two decades two non-air quality regulations and one world wide economic impact caused a revolution in the methods of operation and cement clinker production.

These three actions caused manufacturing facilities to seek lower cost fuel, and to explore tile methods for recycle and reuse of non-virgin materials which could be used as substitution for the raw materials. While raw material substitution remains a goal for the 1990's, it is energy recovery from waste which caused the largest change in operation requirements.

Energy Recycling
Depending upon the type of cement production process (wet or dry), characteristics of the raw materials, etc., between 3 and 6 million Btu are required for each ton of cement clinker. When RCRA implementation regulations were promulgated⁴ cement kiln dust was considered to be a large volume, low toxicity waste identified for more detailed study prior to adoption of snore specific regulations regarding its management. Concurrently, regulations were promulgated and implemented to regulate the kiln dust management for cement plants which used hazardous waste as fuel⁵. It is known that waste fuel can contain trace contaminants of inorganic constituents, including metals which are frequently defined as the BIF metals.

The potential for elevated concentrations of the BIF metals in either the stack gas or the kiln dust, requires that testing be completed for verification that the emission of metals in excess of risk based standards are not exceeded. In addition the management and disposal of wasted dust is impacted by the analysis for metals and organics. The RCRA regulations grant a temporary exclusion from regulation as a hazardous waste under a portion of tile Act known as the Bevill Amendment. To quality for tile Bevill exclusion, it is necessary that the dust be tested to verify one of two alternatives:

BIF Metal Requirements
Cement kilns burning hazardous waste fuel must routinely analyze for metals in several waste products, and these include:

1. The fuel prior to its being burned.
2. Dust which is wasted from the kiln and before recycle or disposal.
3. Draft Mufti-Metals Train catch for regulated metals.

In addition there are specialty testing requirements which may be required on an intermittent basis. Included in this category are analysis of all metal inputs; raw materials such as slurry, coal or other virgin fuel, and any specialty additives.

For facilities which must analyze for metals in slurry and coal, the selection of analytical method could well determine the quantity of metals which are input to the kiln from sources other than the waste fuel. Under BIF requirements, the analytical methods for determination of metals are found in SW-846⁶. These method include Method 3050A, and Method 3051. Method 3050A is commonly called the "Hot Plate" method. The method consist of evaporating the sample several times with nitric acid until about 30 percent of the original volume remains. The sample is then treated with hydrogen peroxide, and after effervescence stops the final evaporation cycle is completed with hydrochloric acid. The time frame is about four ours and constant attention is needed to assure that the material does not evaporate to dryness. Method 3051 is the microwave method and it consists of adding nitric acid to the sample and microwave the sample. The microwave temperature must reach at least 175 degrees C for at least 5.5 minutes. This method allows 12 samples to be run concurrently, and those twelve samples cam be completed in about 45 minutes.

Analysis of Problem
The cement manufacturing process is a high temperature activity which changes the chemical constituents which are present in the final product. Although the kiln dust and slurry have not been subjected to the high temperature of clinker, some of the components are chemically complex and analysis of the metals is not always complete or accurate. This problems is especially prevalent in kiln dust where a sizable portion of the dust has been exposed to the very elevated temperatures at which clinker is formed. The di- and tri-calcium silicates and aluminates incorporate metals into a matrix from which leaching is difficult or incomplete. A preliminary comparison of the hot plate and microwave techniques offered the first clue that comparison of the results could yield different quantification. If addition, a question was raised about whether other methods of analysis could be used to determine the metals in dust and slurry.

A review of candidate techniques identified a fusion method which could be used for dust or cement clinker. The method is described in correspondence⁷ and requires that the sample be pulverized so that it passes a No. 100 sieve, and then the sample is mixed with lithium metaborate and heated to an elevated temperature to form a fusion. In the evaluation of the method, temperatures of 800 and 950 degrees C were used.

OBJECTIVES

The objective of this investigation was to determine whether metal quantification for kiln dust was consistent among three methods: hot plate digestion, microwave digestion and lithium borate fusion. Differences between the methods are to be evaluated.

PROCEDURES

Four cement kiln dust samples were collected from four geographically separate manufacturing facilities. The samples are from four plants which are burning hazardous waste fuel, and all four kilns are wet process. The plants arc located in the east and midwest.

All samples were treated in the same manner. First, the samples were placed in a muffle furnace to determine the Loss on Ignition for each of the four samples. The loss for each of the four samples was: A-14.85%, B-22.53%, C-21.33% and D-24.16%. While the loss is not directly related to analysis of metals, it is an indicator or degree of calcining for the dust and for presence of free carbon, most likely from the unburned fuel.

Each sample was digested using four methods: hot plate, microwave, fusion at 800 degrees C and fusion at 950 degrees C. The filtrate was then analyzed using a Thermal Jarrall Ash, Model 61E, ICP with 24 element simultaneous analysis.

RESULTS

Tables 1,2,3 and 4 present the results of digestion and analysis or each of the four samples, and for each of the four treatment systems. Also on each of the tables is a compilation of the results for each treatment system compared with the hot plate method as the base of comparison.

It can be seen that for all dust samples there arc no data for arsenic, selenium, antimony and thallium because all data for all four parameters was below the detection limit. TCLP data fur all four samples also showed that the metal parameters were all below the detection limit except for barium. No failures for TCLP were noted.

Tables 5,6,7 and 8 compare the results of each digestion method for the individual samples of kiln dust. The results show that there is a significant difference among the various digestion techniques. The hot plate results have been chosen as the base line for comparison of results and they have been assigned the arbitrary value of 100 in tables 1 to 4. The microwave digestion consistently indicated higher metal concentrations than for the hot plate, but the difference is only in tile ten to twenty percent range. The fusion method was conducted at two temperatures and radically differing results were found for the two temperatures.

It can be seen that at the 950 degree C temperature the technique results in the loss of lead and sulfur. Vanadium and cadmium also appear to volatilize partially at this elevated temperature. On the other hand, use of the fusion technique results in significant greater quantities of nickel, chromium and beryllium. In the fusion method silica and aluminate, are solubilized in the digestion, and it is believed that the increased solubilization results in release of trace metals. This same concept also applies to raw mix or slurry, although to a lesser degree than is found the partially formed clinker.

REFERENCES

1. Code of Federal Regulations. Protection of the Environment. National Archives and Records Administration, Volume 40, Part 60, Subpart F; U.S. Government Printing Office, Washington, D.C.

2. Resource Conservation and Recovery Act, P.L. 94-580, October 21, 1976.

3. Pollution Prevention Act of 1990.

4. Code of Federal Regulations, Protection of the Environment. National Archives and records Administration, Volume 40, Part 262, Section 261.4(b)(8).

5. Code of Federal Regulations, Protection of the Environment. National Archives and Records Administration, Volume 40, Pan 266, Subpart Ii, February 21, 1991.

6. U.S. Environmental Protection Agency, Publication No. SW-846

7. William G. Hime, Erlin, Hime Associates, December 26, 1972 to Charles E. Buchanan, Penn-Dixie Cement Company, Nazareth, PA, personal communication.

Table I
Metal Extraction for Plant A Dust

Table II
Metal Extraction for Plant B Dust

Table III
Metal Extraction for Plant C Dust

Table IV
Metal Extraction for Plant D Dust

Table V
Hot Plate Comparison for Microwave Dust Samples

Table VI
Hot Plate Comparison for 950 Degree Fusion Dust Samples

Table VII
Hot Plate Comparison for 950 Degree C Fusion

Table VIII
Particle Size Distribution of four Dust Samples