COMPARISON OF METALS
SPIKING SYSTEMS AND METALS
SELECTION AMONG THE CEMENT KILNS CONDUCTING COMPLIANCE TESTS
President - Combustion Performance Services
Operations Consultant - Gossman Consulting, Inc.
President - Gossman Consulting, Inc.
Presented at the AWMA International Specialty Conference on Waste Combustion in Boilers and Industrial Furnaces March, 1993
This paper compares the various metals spiking equipment systems
regards to: injection points, injection method, and evaluates each
the guidance provided in the Technical Implementation Document. It also
compares the metal compounds selected for the various tests against the
selection criteria detailed in the Technical Implementation Document. A
short discussion of Cr+3 and Cr+6 compound
and EPA's guidance on chromium in cement kilns is included as well.
Between March and September of 1992 cement and aggregate kilns
about 40 Certification of Compliance tests in conformance to the BIF
Each of these kilns spiked their system with additional metals
This paper compares these activities to the guidance given in the
Implementation Document, the BIF Regulation and to each other.
METALS LIMITS IN KILNS
The BIF regulation does not require metals to be spiked into the HWF for the COC test. However BIF made it clear that "what you test is what you get". 40 CFR §266.103(c) states that the operating limits for the ten BIF metals are set during the Certification of Compliance (COC) test for feedrate of each metal in the feedstreams (a) total, (b) total HWF, and (c) total pumpable HWF.
The hazardous waste fuel is very unlikely to have each of the metals at the desired concentration, the highest input rate for the test. Also it is important to remember that the EPA limits the input of each of the metals on an instantaneous basis or on an hourly (2 to 24 hours for some metals) rolling average basis. That is, the input of each metal from all sources must not exceed the limit established during the COC test. This means that the variability of the metals in the raw feed and fuel to the kiln must be accounted for. However, not all metals require the Tier III emissions testing.
Using data generated for the Precompliance Certification, metals retention data for the kiln, and the limits set in Tier I and Tier IA, it is generally possible to limit this to 5 metals - the four carcinogenic metals and lead. In some cases additional metals are added because of state limits, a poor dispersion coefficient or naturally higher levels of a metal in the raw feed or coal.
The quantity of each metal to be spiked is set by an additional set of criteria. The trial burn plan must be conservative enough to pass the test - at $300,000+ per test there is realistically only one opportunity. Conversely, the plan should not unduly restrict your HWF feed rate or metals concentrations. To "walk the line" on this requires a good model of the cement production process; metals retention data, raw feed metals concentrations, metals concentrations in the waste dust, particulate emission rates etc. To this should be added a couple of caveates:
Once these variables have been addressed, the input rates of the
metals can be calculated.
METALS COMPOUND SELECTION
The EPA included guidance on compound selection in the Technical Implementation Document (TID) as follows:
What is interesting here is that the guidance is at odds with EPA's and industry's actual field experience. In the list of chemical compounds listed in the TID there are only a couple that are organic based and these are water soluble acetates. The rest are water soluble inorganic salts.
Table I lists the compounds used in all of the COC tests. The most commonly used compounds are in bold; most of these compounds are inorganic salts with lead octoate being the exception. Compound selection for the COC tests was, for the most part, consistent with the list in the TID, but not what the EPA had hoped for. There are metal-organic compounds available for chrome and antimony as well as lead, as noted above. These materials were used in only a few tests however.
USE OF SURROGATES
In the COC test reports there were a couple instances of the use of
surrogates in place of the specified metal. That is, the facility
permission from the Regional EPA office to substitute emission data on
one metal for another. It is reported in the COC that this permission
granted. The TID however in Section 3.4.1 states; "Compliance
limits would be based on compliance test data using only the toxic
regulated by the BIF Rule." Additionally, in the BIF regulation proper
40CFR§266.106(d) states; (1) "Conformance with the Tier III metals
controls must be demonstrated by emissions testing to determine the
rate for each metal,..." and (6) "Under Tier III, the metals
controls must be implemented by limiting the feed rates of the
metals to levels during the trial burn or the compliance test." It
is reasonable to believe that the words "each" and "individual"
in the BIF text limit the test to the specific metal, not a surrogate.
Table I. Metal compounds used in all of the COC tests.
|Frequency of Use
|Listed in the TID?
|Sodium Antimony Tartrate
|Arsenic (Compound not stated)
CHROMIUM COMPOUND SELECTION
In the TID the EPA also included guidance for the selection of the Chromium compound that should be used in the test. The EPA believed, based on a trial burn at an incinerator, that hexavalent chrome should be fed to the kiln if Cr+6 emissions rates were to be tested where feed rates for Cr+6 were to be established.
The difficulty is, there is no known (let alone EPA approved) method of determining Cr+6 in an organic matrix. In addition to this, while checking out various chrome compounds for use in spiking solutions some interesting information came to light. There were papers written in the 1960's on the use of chrome compounds in ceramics, primarily papers on chemical stability under different temperature and oxygen conditions similar to the temperature and oxygen conditions found in cement kilns. Essentially what they say is some Cr+3 would oxidize to Cr+6 in the burning zone of the kiln, then decompose back to Cr+3 as it cools. Some of this Cr+6 that is so formed enters the clinker before it leaves the burning zone. This is consistent with the fact that cement chemists frequently see 30 to 40 ppm of Cr+6 in the clinker.
Another difficulty is that there are no Cr+6 organic compounds, only inorganic salts most of which are reactive with organic materials. There is one compound, lead chromate, a pigment, which might be present in wastes, though less so each year due to restrictions on the use of lead in paints. Solutions of Cr+6 salts react with organic materials and with most of the common metal salts used in the COC tests. It is highly unlikely that the Cr+6 salt solution was still Cr+6 by the time it reached the kiln burner nozzle. In combination, it would seem that EPA's guidance regarding chrome compound selection is technically questionable.
METHOD OF SPIKING
The TID included guidance on the method of spiking as well. For spiking liquid HWF, the TID recommends the solutions be injected into a flowing HWF line to the kiln upstream of an in-line mixer and as close to the burner nozzle as possible. It is suggested that the feed rate of these solutions be metered separately from the HWF. For solid HWF, the spiking materials should be added as preweighed packets or bottles. Table II tabulates the various metals spiking methods and frequency of use during the COC tests.
Some kilns used more than one method, consequently the numbers add up to more than the number of kilns tested. Direct injection of metals solutions into a flowing HWF line to the kiln was the most frequent method. There is no way of knowing if there was an "in-line mixer" installed downstream of the input point, or if the location of the input point was "as close as possible to the burner nozzle". Neither of these items are mentioned in the COC reports. In metals spiking operations conducted by CPSI and ACT, the solutions were injected downstream of the return loop to the storage tank. This was usually downstream of the HWF flow control station. As for the "in-line mixer" it is reasonable to believe that the configuration of the burner nozzle assures thorough mixing. Air atomization is also often used.
Table II. Various metals spiking methods and frequency of use during the COC tests.
|Frequency of Use
|Direct injection of metal solution into a flowing HWF line to kiln
|Direct injection of metal solution into a non-fuel lance in the burning zone
|Blend metal salts with stored HWF
|Preweighed containers of dry metal salts inserted with solid HWF at timed intervals
|Direct injection of salts into coal mill or coal line
|Direct injection of dry metal salts into a dry solid (powder) HWF lance into the burning zone
Except for the one facility that blended the metal salts into the HWF in storage tanks, the metal solution input rates were monitored separately as suggested by the TID. Generally these rates were monitored by differential change in weight of a container on a scale, or less frequently by a differential change in volume in a container. Solid HWF was spiked in the method suggested in the TID i.e., by discrete packets added at timed intervals.
In comparison to the TID there are two methods that are questionable. Injection of metal solutions into a non-fuel lance is not a method included in the TID. Additionally, this method does not conform to the intent of the 40 CFR § 266.103(c). Metals spiked in this manner would not be part of either the "Total HWF" or "Total Pumpable HWF" feedstreams. Consequently for Tier II and III metals, the "Maximum Metals Feed Rates" reported on Compliance Certification Form 5 for these streams should not include the spiked metals. The same is true for metals injected into the kiln via the coal mill/coal line.
As a secondary point, unless the kiln consumes only solid HWF there was no requirement to spike the solids with metals. The "Total HWF" and the "Total Pumpable HWF" can have the same values. This provides the facility with the flexibility to consume all the metals via the liquid HWF.
Comparisons on the efficacy of one method of injection over another were not able to be performed. The COC reports did not generally contain this type of information or the information was not readily available. Few of the facilities reported problems with their systems. Based on what little there was in the reports and in conversations, there were some equipment problems. Those reported were generally associated with the chemicals causing plugging, corrosion and leaks. Some of the COC reports provide enough data to evaluate the stability of the flow rate into the system during the test and stabilization periods. Most of these show fluctuations of ± 10% or less. A few showed much larger fluctuations generally associated with plugging or other equipment problems.
1. EPA, "Burning of Hazardous Waste in Boilers and Industrial Furnaces - Final Rule", Federal Register, Volume 56, No. 35, Feb 21, 1991.
2. EPA, Technical Implementation Document for EPA's Boiler and Industrial Furnace Regulations, EPA530-R-92-011, March, 1992.
3. D.G. Gossman, D.L. Constans, Commercial BIF Compliance Test
Gossman Consulting, Inc., Hampshire IL, 1993.
TEN KEY WORDS
Certificate of Compliance (COC)
Technical Implementation Document (TID)