[Frank R. Clay, inventor of test Method 306A for sampling chromium air emissions, wrote two letters to EPA regarding various technical issues within the test protocol. These letters, as well as EPA’s response, appear below]

[Frank Clay's first letter]

1714 Park Drive
Raleigh, North Carolina 27605

9 January 1997

Mr. Lalit Banker
Mail Drop 13
U.S. EPA
Research Triangle Park, N.C 27711

Dear Lalit,

Since leaving the Agency, I have done some compliance sampling for chromium emissions for the electroplating industry. During this sampling. I have encountered some problems with the chromium NESHAP that are confusing for the regulatory agencies involved. There are some deletions and additions that need to be made, as well as some procedures that could be revised to facilitate sampling. This letter will discuss primarily additions and deletions. Knowing that correcting the FEDERAL REGISTER is a monumental task, I would appreciate a reply in writing concerning the Agency position on the items that I shall discuss. Your reply will help eliminate confusion that may occur on future tests.

The first problem concerns Cr+6 recovery and analysis and can be found on page 4982, FEDERAL REGISTER / Vol. 60, No. 16 / Wednesday, January 25, 1995 / Rules and Regulations. The section of concern is: Section 5. Procedure.

In section 5.1(c), the paragraph is confusing. The procedure says that the pH of the first impinger should be checked at the end of the run using a pH indicator strip. The pH should be greater than 8.5. If the pH is not greater than 8.5, reagent should be added to the solution until it is 0.5 N and the sample should be rerun (what exactly does rerun mean? - rerun the sample or rerun the pH indicator test?) It would be difficult to obtain a pH of 8.5 at the end of a test if 0.1 N sodium bicarbonate were used in the impingers since 0.1 N sodium bicarbonate has a pH of 8.4. I don't recall that we ever checked the pH of the first impinger on our source tests for the NESHAP, and I doubt that anyone is actually doing this now. It would take a lot of concentrated reagent to raise a solution that started out as a 0.1 N solution to a 0.5 N. solution. This should be looked into by a chemist to see if it is correct or if it is necessary at all. A typical outlet test will have a gas sample volume of about 90 cubic feet (2.55 cubic meters) and will contain no more than 38.2 micrograms of chromium catch to be acceptable.

Section 5.2.3 under sample recovery is also an area of concern and deals with IC/PCR analysis for hexavalent chromium. This paragraph says that the sample must be filtered immediately following recovery to remove any insoluble matter. During the source tests that produced the data for the chromium NESHAP, we never filtered the sample immediately after recovery. The IC/PCR analytical method was originally developed for analyzing samples from boilers and industrial furnaces, and these combustion sources would contain insoluble particles. A plating source employs no combustion in the process and the filtration step is unnecessary. Furthermore, the sample is filtered prior to analysis at the lab.

There are a couple of other misconceptions about the IC/PCR analysis. Some people who have collected samples for Cr+6 analysis think that the samples must be refrigerated immediately after recovery and that they should be delivered to the laboratory within 24 hours. This may be a result of reading the boiler and industrial furnace method which was originally Method 0013. During the development of the chromium NESHAP, however, samples were not refrigerated nor were they shipped to the laboratory within 24 hours. This would be impractical and is not pointed out in the chromium NESHAP.

Another area not mentioned in the NESHAP is the production rate for the source test. Those of us involved with the chromium standard agreed that the test should be conducted at the average production rate. On a recent test in Florida, the state regulatory people did not know what production rate should be used during the test so the owner then operated the process at 150% of normal. The production rate is not specified in the regulation nor is any mention made of how to handle breaks in the plating process when parts are added to or taken from the tank.

The last item that has caused some confusion concerns Method 306- A and is on page 4986 under section 1: Applicability and Principle. Paragraph 1.1 ends with the sentence, "This method is applicable under ambient moisture, air, and temperature conditions." This sentence has been considered by some agencies to mean that electroplaters using scrubbers as a control device could not be tested using Method 306-A. This was never the intent, and if scrubbers had presented a problem, it would have been mentioned in the method. Any electroplating process with a scrubber that cannot show compliance with Method 306-A will not show compliance with Method 306. To avoid confusion, the last sentence in paragraph 1.1 should be deleted.

I would appreciate your comments on the items mentioned above and I will look forward to your reply.

Sincerely,

Frank R. Clay

---

[Frank Clay's second letter]

1714 Park Drive
Raleigh, North Carolina 27605

7 March 1997

Mr. Lalit Banker
MD-13
U. S. EPA
Research Triangle Park, NC 27711

Dear Lalit,

It was good to see you this past January at the 1997 AESF/EPA Conference in Orlando, Florida. Perhaps the two things that I enjoyed most about the conference were flying to and from the conference in a twin Cessna 340, and talking to Bruce Jordan about airplanes.

When I returned home, I had a phone call from a consultant at Network Environmental in Kalamazoo, Michigan. He had recently performed a Method 306-A test in Kentucky and told me that he had recovered the nozzle and tubing using deionized (DI) water. He said that the state observer wondered if the DI water was acceptable for this part of sample recovery. The analysis done was for total chromium.

I told the consultant that I no longer represented the Agency and suggested that he contact Gary McAllister. I also told him that I didn't think that using DI water in place of a basic reagent would make any difference in the results if a total chromium analysis was made on the sample.

This phone call resulted in an interesting question. Is the re any need to use a basic solution as a reagent if a total chromium analysis is to be made?

In the beginning of the chromium project, we were using the diphenylcarbazide colorimetric analytical method and were looking for hexavalent chromium only. A basic solution (0.1 N NaOH) was used for the train reagent in order to stabilize the hexavalent chromium and keep it from converting to the trivalent form. As time went by and the control devices became more efficient, the colorimetric method was no longer sensitive enough (10 microgram per liter minimum detection limit (MDL)). Three more sensitive analytical methods were adopted. They were inductively coupled plasmology (ICP), graphite furnace atomic absorption (GFAA), and ion chromatography with a post column reactor (IC/PCR).

The first method, ICP, determines total chromium (hexavalent chromium plus trivalent chromium) and has a MDL of 4 micrograms per liter at Triangle Labs, and a MDL of 7 micrograms per liter at Research Triangle Institute. GFAA , has a MDL of 1 microgram per liter and determines total chromium. The third method, IC/PCR, has a MDL of 0,5 microgram per liter and detects hexavalent chromium only. For the IC/PCR analysis, the sample is collected in basic solution to prevent the hexavalent chromium from converting to trivalent chromium. At the present time, all samples, regardless of analysis, are collected in a basic solution.

If the analysis is to be for total chromium, does it matter if some of the hexavalent chromium converts to the trivalent form between the time that the sample is taken and the time that it is analyzed? The total of the two should be the same regardless of the sizes of the fractional parts. Would not DI water work just as well as a basic solution in this case, or is there some other reason that there is a need for the basic solution?

There is a test report that may be helpful in resolving this issue The report is from Greensboro Industrial Platers in 1986, and I believe that several combinations of DI water and basic solutions were analyzed, as well as analyzing split samples at different time intervals.

If DI water can be used in place of a basic solution for a total chromium analysis, it would benefit bath the platers and the consultants in terms of time, money, and effort spent to show compliance. Furthermore, an answer to this question by the Agency would demonstrate support for the Common Sense Initiative that has thus far worked so well between the electroplating industry and the U. S. Environmental Protection Agency.

In addition to the question from the consultant in Kalamazoo, Michigan, other consultants have recently contacted me about the use of Method 306-A for sampling control devices that have multiple inlets. Some of the inlets do not carry chromic acid mist , and the procedure for sampling under these conditions is found in the FEDERAL REGISTER / Vol. 60, No. 16 Wednesday, January 25, 1995/ Rules and Regulations, beginning on page 4970, Section 63.344 Performance test requirements and test methods. At the present time, only Method 306 is allowed for sampling control devices where some of the inlet ducts do not Carry chromic acid mist and provide what is considered to be dilution air.

Method 306-A will also work for sampling these locations. The reason it was not recommended initially was that it was felt that the calculations necessary for this testing were too complicated for most electroplates to perform. Specifying Method 306 in these circumstances insured that the sampling would be done by a consultant. As it turns out, many consultants are using Method 306-A for compliance sampling of electroplating facilities where multiple ducts are not a problem.

Method 306-A will work using the guidelines given for Method 306 where multiple ducts are concerned, It will still be necessary to determine the total sample time for the test as specified in the Method 306 procedure and to divide this time by 24 to get anew point base time for use with Method 305-A. (The point base time for a normal two hour Method 306-A test is 5 minutes.) The new point base time is then used in the calculations that determine the sample times in minutes and seconds for each sample point. After sampling, the procedures for multiple ducts outlined for Method 306 can also be used for Method 306-A.

Allowing the use of Method 306-A as well as Method 306 for sampling control devices with multiple inlets would not only save the industry money, it would be another demonstration of support for the Common Sense Initiative that has thus far been so beneficial to the electroplating industry.

At the present time, there is a problem with sampling control devices with multiple inlet ducts using the currently specified procedure. Although the procedure will work, the problem is the increased amount of time required to obtain the sample. The time calculation is based upon the amount of time required to pull the same volume of chromic acid mist through the sampling train as if the dilution air did not exist. For example, if a control device had three inlet ducts

that had identical areas, and only one duct contained chromic acid mist, the areas of all the ducts would be summed (the numerator) and ratioed with the inlet area that carried chromic acid mist. (denominator). The ratio would be 3. A standard 2 hour run would be multiplied by three to give a total sample time of six hours.

The lengthy sample times increase the cost of testing using either Method 306 or Method 306-A, A good sampling team can obtain three runs (two hours each) in a single day using either method. Longer sampling times require longer field times and more days spent at the job site.

There is a way to sample these installations that would require only the standard two hour sample time. This approach is based on obtaining enough sample to achieve 5 times the minimum detection limit of the analytical method used. This would require calculating an allowable concentrate ion value, which, when considering the dilution air, would naturally be less than the 0.015 milligrams per dry standard cubic meter allowed from facilities where dilution air is not present.

A typical Method 306-A test consists of about 85 dry standard cubic feet (DSCF), and 550 milliliters of recovered sample solution. A good control device will have an emission concentrate ion number of about half the emission standard or 0.0075 milligrams per dry standard cubic meter (mg/dscm). The chromic acid mist collected in 85 DSCF is equal to 0.02832 cubic meters per cubic foot x 85 cubic feet x .0075 milligrams/ DSCM. This is equal to 0.01805 milligrams, and multiplying this number by 1000 gives micrograms of catch. The catch would be 18.05 micrograms.

The GFAA analysis has a minimum detection limit of one microgram per liter. Five times this amount is 5 micrograms per liter. A 550 milliliter sample would have to contain 550ml./1000ml./liter x 5 micrograms/liter = 2.75 micrograms in the sample catch to meet the five times the minimum detection limit required. This is [1/(18.05/2.75)1 or 1/6.56 of what would be caught in a normal sampling run of two hours with no dilution air and at half the allowable concentration. It also means that almost 85% dilution air could be tolerated in a sampling run and still produce reliable emission data.

The above example is what would be typically expected, but no allowance is made for smaller cubic feet and higher sample solution volumes. Consider a sample of 60 DSCF instead of 85, and a recovered solution of 600 milliliters. The volume of 60 cubic feet converted to cubic meters is 60 x 0.02832 = 1.699 cubic meters. At 0.0075 milligrams/DSCM, the micrograms of catch is 12.74. If the minimum catch is 5 micrograms per liter, then 600 milliliters would need to contain 600/1000 x 5 or 3 micrograms. A3 microgram catch is a little less than one quarter of what would be expected with no dilution air, meaning that the system could have over 75% dilution air and still provide a reliable emission value.

These two examples have used GFAA as the analytical procedure. The IC/PCR method is twice as sensitive, and has the effect of doubling the amount of dilution air that can be tolerated. It is not likely that there will be a control device used to control chromic acid emissions that has 7 inlets with dilution air and only one duct with chromic acid mist.

To improve both methods, I recommend the following:

1. Determine the diameters of all inlet ducts.
2. Divide by two to get the radius of each duct.
3. Square each radius value.
4. Sum all radii² values. (Use this for the denominator).
5. Sum all the radii² of chromic acid ducts. {Numerator)
6. Multiply Cr radii² /total radii² by 0.015 Mg/DSCM. This will give the maximum allowable concentration.
7. Use 0.1 N NaOH in impinger(s) and for sample recovery.
8. Sample with either method for 2 hours. Obtain a minimum sample volume of 60 DSCF.

The ratio of the radii² of the chromic acid mist ducts divided by the sum of all the radii² should not be less than O.12. If it is, use the IC/PCR analysis. If the ratio is less than 0.06, consult the Administrator. The use of 0.1 N. NaOH makes it possible to perform the IC/PCR analysis if necessary.

This procedure will be just as accurate as the procedure that is currently in the FEDERAL REGISTER. It has additional advantages of requiring less time to obtain the sample, and there is no need to calculate a mass emission rate. To increase the accuracy, however, a velocity traverse of each inlet duct would give the total Dry Standard Cubic Feet per Minute (DSCFM) flow rate of the ducts. If the flow rates were added instead of the areas, a more accurate number might be obtained if the flows among the ducts were not linear.

By recalculating the allowable concentrate ion rather than recalculating the sample time, it is possible to perform two hour testing with either Method 306 or Method 306-A as long as the dilution air does not exceed 87% of the total. It also makes it possible to determine an allowable concentration value for outlet sampling where dilution air is present.

Knowing that the Agency is dedicated through the Common Sense Initiative to making it as easy and inexpensive as possible for platers to comply with the regulation, I would appreciate a reply on using Method 306-A for sampling facilities where multiple inlet ducts (not all containing chromic acid mist) are employed. In addition, I would also like your comments concerning the method that I have outlined for calculating a lower concentration instead of a longer sampling time. This will save time and money in compliance sampling, and there will be no loss in data quality. It is superior to the procedure currently in the FEDERAL REGISTER, and it should be more than satisfactory as an alternative sampling method for showing compliance.

At the present time, I know of three plating facilities that have a single control device with multiple inlets that have not been tested. I am sure that there are many more. Using Method 306-A for the compliance test would be of benefit to both the Agency and the electroplating industry.

I apologize for writing such a long letter, but I did not have the time to write a short one. I await your timely response.

Sincerely,

Frank R. Clay

cc : F, Altmayer, D. Bell, K Hankinson, B. Hunt, B. Jordan, R. Mitchum, R. Segall, L. Zitko

[EPA Memorandum]

 
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
OFFICE OF AIR QUALITY PLANNING AND STANDARDS

JULY 9, 1997

MEMORANDUM
Subject: Response to Frank Clay Comments
From: C. B. Oldham for Gene Riley
Emission Measurement Center, EMAD (MD-19)
To: Lalit Banker
Organic Chemicals Group, ESD (MD-13)

The following is offered in response to two letters sent to Mr. Lalit Banker from Mr. Frank Clay. Mr. Banker forwarded the letters to the Emissions, Monitoring, and Analysis Division for technical review and comments. Specific issues and technical recommendations are listed for each letter in the following summary.

In the summary, you will see where we plan to either make revisions to Methods 306 and 306A regarding these issues or develop guidance to respond to the issues. For process and operational issues, we recommend revisions that you may want to make to Subpart N, the Chrome NESHAP. Please feel free to use the contents of this memo to prepare your response to Mr. Clay's letters. If you require additional information or have concerns regarding our course of action, please do not hesitate to contact me at (919) 541-5239.

cc: W. Lamason (MD-14)
P. Mulrine (MD-13)
C. Oldham (MD-19)
R. Segall (MD- 19)
S. Wyatt (MD-19)
 

---

[EPA Reply to Frank Clay’s two letters]

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
OFFICE OF AIR QUALITY PLANNING AND STANDARDS

AUG. 18, 1997

Mr. Frank Clay
1714 Park Drive
Raleigh, North Carolina 27605

Dear Mr. Clay:

I have enclosed the response from Mr. Gene Riley of the Emission Measurement Center to the questions in your two letters you wrote to me. In response to one of your questions regarding the criteria to be used to establish the facility production rate for the compliance test and to address plating process interruptions during the compliance test which Mr, Riley deferred to the Emission Standards Division, I have included the statement from the General Provisions section 63.7(e). This section relates to the conduct of performance tests and parts of which states that the performance tests shall be conducted under such conditions based on representative performance (i e., performance based on normal operating conditions) of the affected source. Operations during periods of startup, shutdown, and malfunction shall not constitute representative conditions for the purpose of a performance test. As such, no performance test should be conducted during process interruptions or malfunctions. Also, it maybe advantageous for the source to test during a range of process operations that will ensure compliance during all operating conditions.

If you have any further questions, please do not hesitate to contact Mr. Riley or me.

Thank you for your interest and patience.

Sincerely,

Lalit Banker
Environmental Engineer
Organic Chemicals Group

Enclosure
cc: P. Mulrine (MD-13)
G. Riley (MD-19)
S. Wyatt (MD-13)

 

ISSUES AND RESPONSES

I. Frank Clay Letter To Lalit Banker Dated January 9, 1997

Issue 1- If the pH is not greater than 8.5, reagent should be added to the solution until it is 0.5 N and the sample should be rerun (what exactly does rerun mean? - rerun the sample run; rerun the pH indicator test?). See Reference Method 306, Section 5.1c.

EPA Response - we agree that this instruction needs clarification. If the measured pH of the sample does not meet the specified criteria, the collected sample is to be discarded and a replacement sample collected (e.g. another test run conducted). However, it should be recognized that this pH requirement is applicable only for Cr+6 determinations and does not apply to total chromium samples. Additional clarification will be proposed in the Federal Register package scheduled for release this year.

Issue 2- Mr. Clay makes the point that the sodium bicarbonate sampling solution does not ever reach a pH of 8.5. Also, he states that the solution pH was not checked during performance of the NESHAP background testing and he doubts that anyone is actually checking the pH now. Mr. Clay also raises concerns regarding the normality increase of the sodium bicarbonate solution from 0.1 N to 0.5 N to accommodate the acidic emissions. Is it necessary to measure the impinger solution pH? Can a normality of 0.5 N be obtained with

the sodium bicarbonate solution?

EPA Response - Mr. Clay is correct in regard to the pH of the sodium bicarbonate; the pH of the sodium bicarbonate solution does not reach 8.5. This oversight will be corrected to read a pH of 8.0 in the proposed revisions, Again, this specified pH requirement is applicable for Cr+6 determinations only and does not apply to total chromium samples.

Methods 306 and 306A require that the pH be measured and verified during Cr+6 sample recovery. This procedural verification will still be required for any Cr+6 determinations. The specified pH range (>8.0) is cited by a 1989 paper entitled, "Sampling And Analytical Methodology For Measurement of Low Levels of Hexavalent Chromium From Stationary Sources"; S.C. Steinsberger, A.C. Carver, W.G. DeWees, J.E. Knoll, F.E. Butler, M.R. Midget; Proceedings of the 1989 EPA/A&WMA International Symposium on Measurement of Toxic and Related Air Pollutants, Air & Waste Management Association, Pittsburgh, PA 1989, pp. 308-313. Dr. Steinsberger' work indicates that collecting the Cr+6 at pH's above 8 prevents conversion of Cr+6 to Cr+3.

An EMC laboratory check of the sodium bicarbonate solutions of 0.1 N and 0.5 N yielded pH results of >8.0 for both solutions.

 

2

Issue 3 - Is it necessary to filter Methods 306 and 306A Cr+6 samples?

EPA Response - See attached EPA Letter dated May 22, 1997 from William Hunt to Frank Clay.

 

Issue 4 - Method 306 guidelines on sample storage and sample holding times we unclear. Should Cr+6 air samples be refrigerated during shipping and should the holding time prior to analysis be only twenty-four hours?

EPA Response - current plans are to revise Methods 306 and 306A to include the requirements for sample storage/shipping along with sample holding times, The following specifications were developed from discussions with EPA and consultant analysts as well as the review of SW-846 OSW Method 7199, "Determination of Hexavalent Chromium In Drinking Water, Groundwater, and Industrial Wastewater Effluents by Ion Chromatography", Section 6.3.

1- Samples to be analyzed for Cr+6 must be shipped and stored at 4 deg. C.

2- Samples to be analyzed for total chromium need not be refrigerated.

3- Samples to be analyzed for Cr+6 shall be analyzed within 14 days of collection.

4- Samples to be analyzed for total chromium shall be analyzed within 60 days of collection.

 

Issue 5 - What criteria should be used to establish the facility production rate for the compliance test? What criteria should be established to address plating process interruptions during the compliance test?

EPA Response - EMC defers to ESD on these issues.

 

Issue 6 - The Applicability and Principle section of Method 306A contains the following sentence: "This method is applicable under ambient moisture, air, and temperature conditions," Mr. Clay notes that this sentence is confusing to the state agency staff and should be deleted. In particular, the phrase has been interpreted by some to exclude the use of Method 306A when scrubber systems are used as controls at the electroplating sources. This text is not included in M306 and should not be included in M-306A.

 

3

EPA Response - We agree that the intent of the sentence can be misinterpreted. The key words of concern are "ambient moisture, air, and temperature conditions". Most electroplater sources with or without emission control devices usually operate their systems under conditions that are at or near ambient temperature and moisture. Either Method 306 or 306A can be used to measure chrome emissions from most of these chromeplater sources.

Method 306 is capable of measuring actual stack conditions during the testing: stack gas velocity, volumetric flowrate, stack moisture and stack gas temperatures. Whereas, Method 306A is designed to assume ambient moisture conditions (2 percent) and the average (beginning and end) and/or ambient conditions for the stack temperatures. We believe that a distinction between the two methods should be made and plan to propose revisions to Method 306A as follows: "This method is applicable under nominal ambient moisture, air, and temperature conditions; this applicability also applies to sources controlled by wet scrubbers.