DEPARTMENT OF ENVIRONMENTAL QUALITY  
AIR QUALITY DIVISION  
AIR POLLUTION CONTROL  
(By authority conferred on the director of the department of environmental quality  
by sections 5503 and 5512 of 1994 PA 451, MCL 324.5503 and 324.5512, and  
Executive Reorganization Order No. 1995-18, MCL 324.99903)  
PART 10. INTERMITTENT TESTING AND SAMPLING  
R 336.2001 Performance tests by owner.  
Rule 1001. (1) The department may require the owner or operator of any source  
of air contaminant to conduct acceptable performance tests, at the owner's or operator's  
expense, in accordance with R 336.2003 under any 1 of the following conditions:  
(a) Prior to issuance of a permit to operate.  
(b) The source is determined to be in violation of R 336.1301 and the potential  
emissions exceed 100 tons per year.  
(c) The owner or operator of the source has not submitted an acceptable  
performance test, in accordance with R 336.2003, that demonstrates that the source is in  
compliance with the department's rules and with the conditions specified in the permit to  
install.  
(d) The source of air contaminant is located in an area designated as  
nonattainment for 1 or more air pollutants, and more than 12 months have expired  
since the date of the last performance test for such designated nonattainment  
pollutants.  
(e) The source of air contaminant has potential emissions in excess of 100 tons per  
year, is located in an area designated as attainment for 1 or more air pollutants, and  
more than 36 months have expired since the date of the last performance test for such  
designated attainment pollutants.  
(f) After completion of a compliance program.  
(2) Performance tests required by subrule (1) of this rule shall be conducted  
within 60 days following receipt of written notification from the department, unless  
otherwise authorized by the department.  
(3) For a performance test required by subrule (1) of this rule, the owner or operator  
shall submit a site-specific test plan not less than 30 days before a performance test for  
approval of the department. The plan will include test program summary, test  
schedule, and the quality assurance measures to be applied.  
(4) Not less than 7 days before performance tests are conducted, the owner of a  
source of air contaminant, or his or her authorized agent, shall notify the department, in  
writing, of the time and place of the performance tests and who shall conduct them. A  
representative of the department shall have the opportunity to witness these tests.  
Page 1  
Courtesy of www.michigan.gov/orr  
(5) Results of performance tests shall be submitted to the department in the format  
prescribed by the applicable reference test method within 60 days after the last date of  
the test.  
History: 1980; 2002 AACS; 2009 AACS.  
R 336.2002 Performance tests by department.  
Rule 1002. (1) The department may conduct performance  
tests  
in  
accordance with R 336.2003 at any source of air contaminant, on behalf of the state, at  
a reasonable time and at the state's expense. During the conduct of such tests, the  
department may obtain samples of any air contaminant and samples of any material  
entering or exiting the source or aircleaning device for the purpose of evaluating  
pollutant emissions with respect to process operating conditions.  
(2) The department shall provide written notification to the owner or operator  
of a source of the department's intent to conduct performance tests pursuant to  
subrule (1). Within 30 days of receipt of such notification, the owner or operator  
shall provide, and bear the expense of, performance test facilities as specified by the  
department, including the following:  
(a) Sampling ports adequate for reference test methods applicable to the source.  
(b) Safe sampling platforms as required.  
(c) Safe access to sampling platforms.  
(d) A suitable power source within 50 feet of any sampling location  
designated by the department. Upon request, additional time for installing the required  
performance test facilities may be authorized by the department for special situations.  
(3) The owner shall not be responsible for providing sampling instruments  
and sensing devices.  
(4) Results of performance tests shall be furnished to the owner or operator, or  
both, in the format prescribed by the applicable reference test method within 60 days  
following the last date of the test.  
History: 1980 AACS; 2002 AACS.  
R 336.2003 Performance test criteria.  
Rule 1003. (1) Performance tests shall be conducted and data reduced  
according to the reference test methods listed in R 336.2004, unless the department  
does any of the following:  
(a) Specifies or approves, in specific cases, the use of  
with minor changes in procedures or equipment.  
(b) Approves the use of an equivalent method.  
a
reference test method  
(c) Specifies or approves the use of an alternative method if an applicable  
reference test method does not exist for a specific air contaminant or source of air  
contaminant.  
(2) A performance test shall consist of a minimum of 3 separate samples of a  
specific air contaminant conducted within a 36-hour period, unless otherwise  
authorized by the department. Each of the 3 separate samples shall be obtained while  
Page 2  
Courtesy of www.michigan.gov/orr  
the source is operating at a similar production level. For the purpose of determining  
compliance with an applicable emission limit, rule, or permit condition, the arithmetic  
mean of results of the 3 samples shall apply. If a sample is accidentally lost or  
conditions occur in which 1 of the 3 samples must be discontinued because of  
forced shutdown, failure of an irreplaceable portion of the sampling train, extreme  
meteorological conditions, or other circumstances beyond the owner's or operator's  
control, then compliance may, upon the approval of the department, be determined using  
the arithmetic mean of the results of 2 samples.  
(3) All performance tests shall be conducted while the source  
contaminant is operating at maximum routine operating conditions, or under such  
other conditions, within the capacity of the equipment, as may be requested by the  
department. Other conditions may include source operating periods of startup,  
of air  
shutdown, or such other operations, excluding malfunction, specific to certain  
sources. Routine operating conditions shall also include those specified within a permit  
to install or a permit to operate. The owner or operator shall make available to the  
department such records as may be necessary to determine the conditions of source  
operation that occurred during the period of time of the performance test.  
(4) For any source that is subject to an emission limitation calculated to 50%  
excess air, the multipoint, integrated sampling procedure of method 3 shall be used for  
gas analysis. For all other sources that require a determination of the molecular  
weight of the exhaust, any optional sampling procedure of method 3 may be  
used. Alternatives or modifications to procedures are subject to the approval of the  
department.  
(5) For reference test methods 5B and 5C, the minimum volume per sample shall  
be 30 cubic feet of dry gas corrected to standard conditions (70 degrees Fahrenheit,  
29.92 in. Hg.). Minimum sample time shall be  
60 minutes, which may be  
continuous or a combination of shorter sampling periods for sources that operate in  
a cyclic manner. Smaller sampling times or sample volumes, when necessitated by  
process variables or other factors, may be approved by the department.  
History: 1980 AACS; 2002 AACS.  
R 336.2004 Appendix A; reference test methods; adoption of federal reference  
test methods.  
Rule 1004. (1) The following federal reference test methods, described in the  
provisions of 40 C.F.R. part 60, appendix A (2007), are the reference test methods  
for performance tests required pursuant to the provisions of this part:  
(a) Method 1 - Sample and velocity traverse for stationary sources.  
(b) Method 1A - Sample and velocity traverses for stationary sources with small  
stacks or ducts.  
(c) Method 2 - Determination of stack gas velocity and volumetric flow rate  
(type-S pitot tube).  
(d) Method 2A - Direct measurement of gas volume through pipes and small  
ducts.  
(e) Method 2C - Determination of stack gas velocity and volumetric flow rate  
in small stacks and ducts (standard pitot tube).  
Page 3  
Courtesy of www.michigan.gov/orr  
(f) Method 2D - Measurement of gas volumetric flow rates in small pipes and  
ducts.  
(g) Method 3 - Gas analysis for the determination of dry molecular weight.  
(h) Method 4 - Determination of moisture content in stack gases.  
(i) Method 5 - Determination of particulate matter emissions  
from  
stationary sources  
(j) Method 6 - Determination of sulfur dioxide emissions from stationary  
sources.  
(k) Method 7 - Determination of nitrogen oxide emissions from stationary  
sources.  
(l) Method 8 - Determination of sulfuric acid mist and sulfur dioxide  
emissions from stationary sources.  
(m) Method 9 - Visual determination of the opacity of emissions from  
stationary sources.  
(n) Method 10 - Determination of carbon monoxide emissions from  
stationary sources.  
(o) Method 10B - Determination of carbon monoxide emissions from  
stationary sources.  
(p) Method 18 - Measurement of gaseous organic compound emissions by gas  
chromatography.  
(q) Method 21 - Determination of volatile organic compound leaks.  
(r) Method 24 - Determination of volatile matter content, water content,  
density, volume solids and weight solids of surface coatings.  
(s) Method 24A - Determination of volatile matter content and density of  
printing inks and related coatings.  
(t) Method 25 - Determination of total gaseous nonmethane organic emissions  
as carbon.  
(u) Method 25A - Determination of total gaseous organic concentration using a  
flame ionization analyzer.  
(v) Method 27 - Determination of vapor tightness of gasoline delivery tank using  
pressure-vacuum test.  
(w) Method 29 - Determination of metals emissions from stationary sources.  
(x) Method 30A - Determination of total vapor phase mercury emissions from  
stationary sources (instrumental analyzer procedure).  
(y) Method 30B - Determination of total vapor phase mercury emissions  
from coal-fired combustion sources using carbon sorbent traps.  
(2) The reference test methods listed in subrule (1) of this rule are adopted by  
reference in this rule. Copies of the test methods may be inspected at the Lansing  
office of the air quality division of the department of environmental quality. A copy  
of title 40 of the Code of Federal Regulations, part 60, appendix A, may be obtained  
from the Department of Environmental Quality, Air Quality Division, P.0. Box  
30260, Lansing, Michigan 48909 7760, at a cost at the time of adoption of these rules  
of $67.00; from the Superintendent of Documents, United States Government  
Printing Office, P.O. Box 979050, St. Louis, Missouri 63197-9000, at a cost at the  
time of adoption of these rules of $57.00; or on the United States government printing  
office internet web site at http://www.gpoaccess.gov.  
Page 4  
Courtesy of www.michigan.gov/orr  
(3) All alternatives that are subject to the approval of the administrator in  
the adopted federal reference methods are subject to the approval of the department.  
(4) Determinations of compliance with visible emission standards for  
stationary sources shall be conducted as specified in reference test method 9 or other  
alternative method approved by the department, with the following exceptions:  
(a) Visible emissions from a scarfing operation at  
a
steel manufacturing  
facility shall be determined as specified in reference test method 9A, which is  
described in R 336.2030.  
(b) Visible emissions from a coke oven pushing operation and fugitive coke  
oven visible emissions shall be determined as specified in reference test method 9B,  
which is described in R 336.2031.  
(c) Visible emissions, fugitive and nonfugitive, from basic oxygen furnace  
operations, hot metal transfer operations, and hot metal desulfurization operations  
shall be determined as specified in reference method 9C, which is described in R  
336.2032.  
(5) Determinations of particulate emission rates for stationary sources shall  
be conducted as specified in 1 or more of the following reference test methods:  
(a) Reference test method 5B, which is described in R 336.2011.  
(b) Reference test method 5C, which is described in R 336.2012.  
(c) Reference test method 5D, which is described in R 336.2013.  
(d) Reference test method 5E, which is described in R 336.2014.  
(e) "Standard Methods for the Examination of Water and Wastewater," (14th  
edition), section 208C, as described and modified in R 336.2033.  
(6) Determinations of total gaseous nonmethane organic emissions as carbon,  
using the alternate version of federal reference test method 25 incorporating the  
Byron analysis, shall be conducted as specified in R 336.2006.  
History: 1980 AACS; 1985 AACS; 1989 AACS; 1993 AACS; 1998-2000 AACS; 2002 AACS; 2006  
AACS; 2009 AACS.  
R 336.2005 Reference test methods for staterequested tests of delivery  
vessels.  
Rule 1005. The following reference test method shall be used to detect  
gasoline vapor leaks by a combustible gas detector:  
(a) Principle. A combustible gas detector is used to indicate any incidence  
of leakage from gasoline delivery vessel tanks  
and  
vapor control systems. This  
qualitative monitoring procedure is an enforcement tool to confirm the continuing  
existence of leak-tight conditions.  
(b) Applicability. This method is applicable to determining leak-tightness  
of gasoline delivery vessel tanks during loading without taking the delivery vessel  
tank out of service. The method is applicable only if the vapor control system does  
not create back pressure in excess of the pressure limits of the delivery vessel tank  
compliance leak test. For vapor control systems, this method is applicable to  
determining leak-tightness at any time.  
(c) Apparatus and specifications. The following apparatus shall be used:  
Page 5  
Courtesy of www.michigan.gov/orr  
(i) Manometer. Liquid manometer, or equivalent, capable of measuring up to 0.9  
pounds per square inch (24.9 inches of water) gauge pressure within 0.003 pounds per  
square inch (0.1 inches of water) precision.  
(ii) Combustible gas detector. A portable hydrocarbon  
associated sampling line and probe which complies with all of  
provisions:  
gas  
analyzer with  
the following  
(A) Safety. The device is certified as safe for operation in explosive  
atmospheres.  
(B) Range. The device shall have a minimum range of 0 to 100% of the lower  
explosive limit (LEL) as propane.  
(C) Probe diameter. The sampling probe shall have an internal diameter of 0.625  
centimeters (1/4 inch).  
(D) Probe length. The probe sampling line shall be of sufficient length for  
easy maneuverability during testing.  
(E) Response time. The response time for full-scale deflection shall be less than  
8 seconds for a detector with a sampling line and probe attached.  
(d) Test procedure. The following test procedure shall be complied with:  
(i) Pressure. Place a pressure tap in the terminal, plant, or service station  
vapor control system as close as possible to the connection with the delivery vessel  
tank. Record the pressure periodically during testing.  
(ii) Calibration. Calibrate the combustible gas detector with 2.2% propane,  
by volume, in air for 100% lower explosive limit response.  
(iii) Monitoring procedure. During loading or  
unloading,  
check the  
periphery of all potential sources of leakage of the delivery vessel tank and of the  
terminal, plant, or service station vapor collection system with a combustible gas  
detector. The check shall comply with the following procedure:  
(A) Probe distance. The probe inlet shall be 2.5 centimeters from the potential  
leak source.  
(B) Probe movement. Move the probe slowly (2.0 centimeters per second).If there  
is any meter deflection at a potential leak source, move the probe to locate the point of  
highest meter response.  
(C) Probe position. As much as possible, the probe inlet shall be positioned  
in the path of (parallel to) the vapor flow from a leak.  
(D) Wind. Attempt, as much as possible, to block the wind from the area being  
monitored.  
(iv) Recording. Record the highest detector reading and location for each  
incidence of leakage.  
History: 1981 AACS; 1989 AACS; 2002 AACS; 2006 AACS.  
R 336.2006 Reference test method serving as alternate version of federal  
reference test method 25 by incorporating Byron analysis.  
Rule 1006. When using the alternate version of federal reference test method 25  
incorporating the Byron analysis, the procedures in method 25, which are described in  
R 336.2004, shall be followed, except that all of the following parts in method 25 are  
amended to read as follows:  
Page 6  
Courtesy of www.michigan.gov/orr  
1.2 Principle. An emission sample is withdrawn from a stack at a constant  
rate through a stainless steel absorber tube packed with porasil; the gaseous portion of  
the sample is pulled past a battery-operated sampling pump into a tedlar bag. After  
sampling is complete, the contents of the tedlar bag are analyzed on an automated gas  
chromatograph (GC), and the sample in the porasil packed tube is heated to remove all  
components for analysis on the GC. The GC separates CO, CO2, and CH4 from  
the nonmethane organics (NMO), then converts the NMOs to methane for analysis.  
2. Apparatus. The sampling system consists of a nonmethane organic (NMO)  
absorber tube, a sampling pump, and a sample bag (figure 25-1). The analytical  
system has 2 parts the oven for removing the sample from the absorber tube and an  
automated gas chromatograph (GC).  
2.1 Sampling. All of the following equipment is required, as shown in figure 25-  
1:  
2.1.1 Heated probe. 6.4-millimeter (mm) (1/4-inch (in.)) outside diameter  
(o.d.) stainless steel tubing with a heating system that is capable of maintaining a  
gas temperature at the exit end of not less than 129 degrees Centigrade (265 degrees  
Fahrenheit). The probe shall be equipped with a thermocouple at the exit end to  
monitor the gas temperature. The nozzle is an elbow fitting that is attached to the  
front end of the probe while the thermocouple is inserted in the side arm of a tee  
fitting that is attached to the rear of the probe. The probe is wrapped with a suitable  
length of high-temperature heating tape and then covered with 2 layers of glass cloth  
insulation and 1 layer of aluminum foil.  
2.1.2 Heated prefilter-only for stacks with possible particulate matter interference.  
A stainless steel filter holder with a 47-mm type A/E fiberglass filter without  
organic binder. The entire prefilter shall be maintained at 110 degrees Celsius. Note -  
if it is not possible to use a heating system for safety reasons, an unheated system  
with an instack filter is a suitable alternative.  
2.1.3 NMO absorber tube. 1/2-inch inside diameter (i.d.) stainless steel tube  
packed with porasil (thermally stable silica gel).  
2.1.4 1/4-inch o.d. teflon line that is 2 to 4 feet long.  
2.1.5 Battery-operated diaphragm sampling pump with kurz digital mass flow  
meter. Total flow is integrated electronically to measure flow with an accuracy of 1% at  
any flow rate. (Byron instruments model 90).  
2.1.6 Sample bag. 0.3-mil tedlar, 1/2-cubic foot capacity. The sample bag  
undergoes nitrogen purge cycle until analysis exhibits zero carbon content in the  
sample bag.  
2.2 Analysis. The following equipment is required:  
2.2.1 Sample recovery on the adsorber tube is done in a Byron model 75 oven in  
2 stages, each stage requiring a 0.3-mil tedlar bag that has a 1/2-cubic foot capacity.  
2.2.2 Analysis is done on a Byron model 401 gas chromatograph (GC) that meets  
all criteria specified in method 25, section 2.2.2.  
2.3 NMO analyzer. The NMO analyzer is a Byron model  
chromatograph (GC). (Remainder of 2.3 as stated in method 25)  
401  
gas  
2.3.5.2 Range. A full scale range of 1 to 10,000 parts per million (ppm) CH4.  
Signal attenuators shall be available to produce a minimum signal response of 10%  
of full scale.  
Page 7  
Courtesy of www.michigan.gov/orr  
3.1.1 Delete (dry ice is not required).  
4.1.1 Model 90 and model 75 flow meter calibration. The model 90 sample pump  
inlet is attached to the outlet of the model 75 oven. Air is passed through the system at  
the rate that will be used in sampling and for the total volume anticipated to be  
sampled. If the flow meters on the 2 instruments do not agree within 0.01 liters,  
then adjust the meter on the model 90 until agreement is within 0.01 liters. After  
making any correction, run a full calibration again.  
4.1.2 Sample train assembly. Assemble the probe (prefilter if needed), adsorber  
tube, and teflon line to the inlet of the model 90. Attach a short (8 to 10 inches)  
flexible line to the outlet of model 90. Have a completely clean evacuated tedlar  
sample bag nearby for collection of sample to be analyzed.  
4.1.3 Pretest leak check. Stopper the inlet of the probe and place the flexible tube  
on the outlet of the model 90 in a small open container of water. Turn on the sampling  
pump. For a satisfactory leak check, bubbling should cease within 1 minute. If the leak  
check is unsatisfactory, tighten the fittings or change parts until a satisfactory leak  
check is obtained. 4.1.4 Sampling train operation. Place the probe and the front portion  
of the adsorption tube in the stack. If the stack has a temperature higher than ambient,  
allow time for the probe to heat before starting the sample pump. Start the model 90  
pump and adjust to the desired flow, usually about 90 ml/min. After about 0.1 liter  
of sampling, or equivalent to the volume of air that is displaced in the sample system  
before the flowmeter, remove the flexible tube from the outlet of the model 90 and  
install the evacuated tedlar bag. This assures that gaseous components are undiluted by  
the air originally in the sampling system. Record requested data on the data form during  
the sample time. The sampling is usually done for 1 hour with a total of 5 to 6 liters  
sampled. When sampling is complete, record the precise volume sampled. The  
process may require different sample times or sample volumes. (Sampling form is  
figure 25-8.)  
4.1.5 Post test leak check. Remove the tedlar bag and replace it with the flexible  
tube. Stopper the probe and operate the same as the pretest leak check specified in  
section 4.1.3. If the leak test is not acceptable, invalidate the sample.  
4.2 Sample recovery. The tedlar bag is ready for direct analysis on the GC. The  
adsorber tube shall undergo the following 2-stage preparation:  
4.2.1 Sample purge. The absorber tube is placed in the Byron model 75 oven  
with a clean tedlar bag attached directly to the tube. A volume of clean dry air is  
passed through the adsorber tube while holding the oven temperature at about 130  
degrees Celsius. The volume of air should be precisely the same as that sampled.  
This purge is necessary to remove any CO2 on the sample tube, and the elevated  
temperature is needed to assure CO2 removal from any absorbed water. The tedlar  
bag is now ready for direct analysis on the GC.  
4.2.2 Sample digest. The absorber tube, now free of CO2 and the lighter NMOs,  
is now attached to an oxidation catalyst, and another tedlar bag is attached to the outlet  
of the oxidation catalyst. A volume of clean dry air equal to that sampled is passed  
through this system while the temperature on the sample tube is brought up to 600  
degrees Celsius. If the sampled volume was less than 3 liters, a larger volume shall be  
used in the digestion to assure completion. Usually a multiple of precisely 1.5 or 2.0 of  
the sampled volume is sufficient. This third tedlar bag is now ready for direct analysis  
Page 8  
Courtesy of www.michigan.gov/orr  
on the GC. If anything other than CO2 is found in this bag, the model 75 oxidation  
catalyst is probably in need of replacement. In this case the test would be invalid  
and would have to be redone.  
4.3 Analysis. Each of the 3 bags is analyzed on the GC. Each bag should be  
analyzed as soon as possible after being filled. At the completion of analysis, the  
bags shall be cleaned by repeated fillings with either clean air or nitrogen. Before being  
used again, the bags shall be checked by filling with clean air and then analyzed  
on the GC to assure zero concentrations of all analyzed substances. All pertinent  
calibration, performance, and operational checks in sections 4.4 and 5 of method 25  
apply to the Byron system.  
6. Calculations.  
6.1 Nomenclature.  
C1 = Concentration of sample bag, ppm C, (NMO converted to methane).  
C2 = Concentration of purge bag, ppm C, (NMO converted to methane).  
C3 = Concentration of digest bag, ppm C, (CO2 converted to methane).  
C = Ppm C (NMO).  
6.2-6.4 (Delete).  
6.5 C1, C2, C3 calculated directly as: ppm C calibration gas x GC reading unknown  
= ppm C GC reading calibration gas  
unknown 6.6 C = C1 + C2 + C3 Delete  
figures 25.3, 25.4, 25.9, and 25.10 from method 25. Amend figures 25.1 and 25.8 from  
method 25 to read as follows:  
FIGURE 25.1  
SAMPLING TRAIN  
Page 9  
Courtesy of www.michigan.gov/orr  
FIGURE 25.1  
Page 10  
Courtesy of www.michigan.gov/orr  
History: 1993 AACS.  
R 336.2007 Alternate version of procedure L, referenced  
in R  
336.2040(10).  
Rule 1007. The alternate version of procedure L is as follows:  
1. Introduction.  
1.1 Applicability. This procedure is applicable for determining the input of  
volatile organic compounds (voc), measured as equivalent propane as measured by a  
flame ionization instrument. It is intended to be used as a segment in the development  
of liquid/gas protocols for determining voc capture efficiency (ce) for surface coating  
and printing operations.  
1.2 Principle. The amount of voc introduced to the process (l) is the sum of  
products of the weight (w) of each voc containing liquid (ink, paint, solvent, or  
similar material) used and its voc content (v), corrected for a response factor (rf) to  
allow the input to be calculated in terms of propane, the same calibration gas used in  
the gaseous voc measurements. A sample of each coating used is distilled to separate  
the voc fraction. The distillate is used to prepare a known standard for analysis by a  
flame ionization analyzer (fia), calibrated against propane, to determine its rf.  
2. Apparatus and reagents.  
2.1 Liquid weight.  
Page 11  
Courtesy of www.michigan.gov/orr  
2.1.1 Balances/digital scales. To weigh drums of voc containing liquids to within  
0.2 lb.  
2.1.2 Volume measurement apparatus (alternative). Volume meters, flow meters,  
density measurement equipment, or similar material, as needed to achieve the same  
accuracy as direct weight measurements.  
2.2 Response factor (rf) determination (fia technique). The voc distillation  
and tedlar gas bag generation systems apparatus are shown in figures 1 and 2. The  
following equipment is required:  
2.2.1 Sample collection can. An appropriately sized metal can to collect voc-  
containing materials. The can shall be constructed in such a way that it can be grounded  
to the coating container.  
2.2.2 Needle valves. To control gas flow.  
2.2.3 Regulators. For fia, calibration, dilution, and sweep gas cylinders.  
2.2.4 Tubing and fittings. Teflon and stainless steel tubing and fittings with  
diameters and lengths and sizes determined by connection requirements of the  
equipment.  
2.2.5 Thermometer. Capable of measuring the temperature of the hot water and oil  
baths to within 1 degree Celsius.  
2.2.6 Analytical balance. To measure plus or minus 0.01 mg.  
2.2.7 Microliter syringe. 10-microliter size.  
2.2.8 Vacuum and pressure manometers. 0 to 760 mm (0 to 30 in.) hg.U-tube  
manometer, vacuum or pressure.  
2.2.9 Hot oil bath, with stirring hot plate. Capable of heating and maintaining a  
distillation vessel at 110 plus or minus 3 degrees Celsius.  
2.2.10 Vacuum/water aspirator. A device capable of drawing a vacuum to within  
20 mm hg from absolute.  
2.2.11 Rotary evaporator system. Complete with folded inner coil, vertical  
style condenser, rotary speed control, and teflon sweep gas delivery tube with  
valved inlet. Buchi rotavapor or equivalent.  
2.2.12 Ethylene glycol cooling/circulating bath. Capable of maintaining the  
condenser coil fluid at minus 10 degrees Celsius.  
2.2.13 Dry gas meter. For the precise measurement of dilution gas volume. It  
shall be calibrated to a primary standard, either spirometer or bubble meter.  
2.2.14 Activated charcoal/mole sieve trap. To remove any trace level of organics  
picked up from the dry gas meter.  
2.2.15 Gas coil heater. Sufficient length of 0.125-inch stainless steel tubing to  
allow heating of the dilution gas to near the water bath temperature before entering  
the volatilization vessel.  
2.2.16 Water bath, with stirring hot plate. Capable of heating and maintaining  
a volatilization vessel and coil heater at a temperature of 100 plus or minus 5 degrees  
Celsius.  
2.2.17 Volatilization vessel. 50-milliliter midget impinger fitted with a septum top  
and loosely filled with glass wool to increase volatilization surface.  
2.2.18 Tedlar gas bag. Capable of holding 30 liters of gas, flushed clean with  
zero air, leak tested and evacuated.  
Page 12  
Courtesy of www.michigan.gov/orr  
2.2.19 Cylinder of compressed zero air. Used to supply dilution air for making the  
tedlar bag gas samples.  
2.2.20 Cylinder of compressed thc free N2. Used as sweep gas in the rotary  
evaporator system.  
2.2.21 Organic concentration analyzer. An fia with a span value of 1.5 times the  
expected concentration as propane; however, other span values may be used if it can  
be demonstrated that they would provide more accurate measurements. The fia  
instrument shall be the same instrument used in the gaseous analyses adjusted with  
the same fuel, combustion air, and sample backpressure (flowrate) settings. The system  
shall be capable of meeting or exceeding the following specifications:  
2.2.21.1 Zero drift. Less than plus or minus 3.0% of the span value.  
2.2.21.2 Calibration drift. Less than plus or minus 3.0% of span value.  
2.2.21.3 Calibration error. Less than plus or minus 5.0% of the calibration  
gas value.  
2.2.22 Integrator/data acquisition system. An analog or digital device or  
computerized data acquisition system used to integrate the fia response or compute the  
average response and record measurement data. The minimum data sampling frequency  
for computing average or integrated values is 1 measurement value every 5  
seconds. The device shall be capable of recording average values at least once per  
minute.  
2.2.23 Chart recorder (optional). A chart recorder or similar device  
is  
recommended to provide a continuous analog display of the measurement results  
during the liquid sample analysis.  
2.2.24 Calibration and other gases. For calibration, fuel, and combustion air,  
if required, contained in compressed gas cylinders. All calibration gases shall be  
traceable to NIST standards and shall be certified by the manufacturer to plus or  
minus 1% of the tag value.Additionally, the manufacturer of the cylinder should  
provide a recommended shelf life for each calibration gas cylinder over which the  
concentration does not change more than plus or minus 2% from the certified  
value. For calibration gas values that are not generally available, alternative  
methods for preparing calibration gas mixtures, such as dilution systems, may be  
used with prior approval.  
2.2.24.1 Fuel. 99.995% hydrogen, 40% hydrogen/60% helium, or 40%  
hydrogen/60% nitrogen. The fia manufacturer's recommended fuel shall be used. An  
attempt shall be made to avoid fuels with oxygen to avoid an oxygen synergism  
effect that reportedly occurs when oxygen concentration varies significantly from a  
mean value.  
2.2.24.2 Carrier gas. High purity air with less than 1 ppm of organic material (as  
propane) or less than 0.1% of the span value, whichever is greater.  
2.2.24.3 Fia linearity calibration gases. Low-, mid-, and high-range gas mixture  
standards with a nominal propane concentration of 20 to 30, 45 to 55, and 70 to 80%  
of the span value in air, respectively. Other calibration values and other span values  
may be used if it can be shown that more accurate measurements would be achieved.  
2.2.24.4 System calibration gas. Gas mixture standard which contains propane  
in air and which approximates the voc concentration expected for the tedlar gas bag  
samples.  
Page 13  
Courtesy of www.michigan.gov/orr  
3. Determination of liquid input weight. A capture efficiency test shall consist of  
not less than 3 sampling runs. Each run shall cover at least 1 complete production or  
processing cycle or shall be at least  
1
hour in duration. For automotive surface  
coating operations, the sampling time per run shall be based on coating a minimum of 3  
representative vehicles.  
3.1 Weight difference. Determine the amount of material introduced to the  
process as the weight difference of the feed material before and after each sampling  
run. In determining the total voc-containing liquid usage, account for all of the  
following:  
(a) The initial (beginning) voc-containing liquid mixture.  
(b) Any solvent added during the test run.  
(c) Any coating added during the test run.  
(d) Any residual voc-containing liquid mixture remaining at the end of the  
sample run.  
3.1.1 Identify all points where voc-containing liquids are introduced to the  
process. To obtain an accurate measurement of voc-containing liquids, start with an  
empty fountain, if applicable. After completing the run, drain the liquid in the  
fountain back into the liquid drum, if possible, and weigh the drum again. Weigh the  
voc-containing liquids to plus or minus 0.5% of the total weight (full) or plus or  
minus 0.1% of the total weight of voc-containing liquid used during the sample run,  
whichever is less. If the residual liquid cannot be returned to the drum, drain the  
fountain into a preweighed empty drum to determine the final weight of the liquid.  
3.1.2 If it is not possible to measure a single representative mixture, then weigh  
the various components separately, for example, if solvent is added during the  
sampling run, weigh the solvent before it is added to the mixture. If a fresh drum of  
voc-containing liquid is needed during the run, then weigh both the empty drum and  
the fresh drum.  
3.2 Volume measurement (alternative). If direct weight measurements are not  
feasible, the tester may use volume meters, flow  
rate  
meters,  
and density  
measurements to determine the weight of liquids that are used if it can be demonstrated  
that the technique produces results equivalent to the direct weight measurements. If a  
single representative mixture cannot be measured, measure the components separately.  
4. Determination of voc content in input liquids.  
4.1 Collection of liquid samples.  
4.1.1 Collect a 1-pint or larger sample of the voc-containing liquid mixture at  
each application location at the beginning and end of each test run. A separate sample  
shall be taken of each voc-containing liquid that is added to the application mixture  
during the test run. If a fresh drum is needed during the sampling run, then obtain a  
sample from the fresh drum.  
4.1.2 When collecting the sample, ground the sample container to the coating  
drum. Fill the sample container as close to the rim as possible to minimize the amount  
of headspace.  
4.1.3 After the sample is collected, seal the container so the sample cannot leak  
out or evaporate.  
4.1.4 Label the container to identify clearly the contents.  
4.2 Distillation of voc.  
Page 14  
Courtesy of www.michigan.gov/orr  
4.2.1 Assemble the rotary evaporator as shown in figure 1.  
4.2.2 Leak check the rotary evaporation system by aspirating a vacuum of  
approximately 20 mm hg from absolute. Close up the system and monitor the vacuum  
for approximately 1 minute. If the vacuum falls more than 125 mm hg in 1 minute, repair  
leaks and repeat.  
4.2.3 Deposit approximately 20 mls of the sample (inks, paints, or similar  
material) into the rotary evaporation distillation vessel.  
4.2.4 Turn off the aspirator and gradually apply a vacuum to the evaporator of  
within 20 mm hg.  
4.2.5 Begin heating the vessel at a rate of 2 to 3 degrees Centigrade per minute,  
maintaining the vacuum specified in 4.2.3. Care shall be taken to prevent material  
bumping from the distillation flask.  
4.2.6 Continue heating until a temperature of 110 degrees Centigrade is achieved  
and maintain this temperature for not less than 10 minutes or until the sample has  
dried in the distillation flask.  
4.2.7 Slowly introduce the N2 sweep gas through the purge tube and into the  
distillation flask, taking care to maintain not less than 125 mm hg vacuum at all  
times.  
4.2.8 Continue sweeping the remaining solvent voc from the distillation flask and  
condenser assembly for 10 minutes or until all traces of condensed solvent are  
gone from the vessel and the still head.  
4.2.9 Disassemble the apparatus and transfer the distillate to a labeled sealed vial.  
4.3 Preparation of voc standard bag sample.  
4.3.1 Assemble the bag sample generation system as shown in figure 2 and bring  
the water bath up to a near-boiling temperature.  
4.3.2 Inflate the tedlar bag and perform a leak check on the bag.  
4.3.3 Evacuate the bag and close the bag inlet valve.  
4.3.4 Record the current barometric pressure.  
4.3.5 Record the starting reading on the dry gas meter, open the bag inlet valve,  
and start the dilution zero air flowing into the tedlar bag at approximately 2 liters per  
minute.  
4.3.6 The bag sample voc concentration shall be similar to the gaseous voc  
concentration measured in the exhaust gas ducts. The amount of liquid voc required  
can be approximated using the equations in section 6, the gaseous voc measurement  
results in terms of propane, and an assumed response factor of 1.0. Let Cc3 equal  
the exhaust gas concentration in terms of propane and rf=1.0. Calculate Cvoc. Let  
bv = 20 liters and calculate ml, the approximate quantity of liquid to be used to  
prepare the bag gas sample.  
4.3.7 Quickly withdraw an aliquot (approximately 5 microliters) of sample  
from the distillate vial with the microliter syringe and record its weight from the  
analytical balance to the nearest 0.01 mg.  
4.3.8 Inject the contents of the syringe through the septum of the volatilization  
vessel into the glass wool inside the vessel.  
4.3.9 Reweigh and record the tare weight of the now empty syringe.  
4.3.10 Record the pressure and temperature of the dilution gas as it is passed  
through the dry gas meter, as shown in the figure 2 diagram.  
Page 15  
Courtesy of www.michigan.gov/orr  
4.3.11 After approximately 20 liters of dilution gas have passed into the tedlar  
bag, close the valve to the dilution air source and record the exact final reading on the  
dry gas meter.  
4.3.12 The gas bag is then analyzed by fia within 1 hour of bag preparation  
in accordance with the procedures contained in section 4.4. 4.4 Determination of voc  
response factor.  
4.4.1 Start up the fia instrument using the same settings as used for the gaseous  
voc measurements.  
4.4.2 Perform the fia analyzer calibration and linearity checks according to  
the procedure in section 5.1. Record the responses to each of the calibration gases and  
the back-pressure setting of the fia.  
4.4.3 Connect the tedlar bag sample to the fia sample inlet and record the bag  
concentration in terms of propane. Continue the analysis until a steady reading is  
obtained for not less than 30 seconds. Record the final reading and proceed with the  
calculation of the response factor.  
4.5 Determination of coating voc content as voc (vu).  
4.5.1 Determine the voc content of the coatings used in the process using EPA  
method 24 or 24a as applicable.  
5. Calibration and quality assurance.  
5.1 Fia calibration and linearity check. Make necessary adjustments to the air and  
fuel supplies for the fia and ignite the burner. Allow the fia to warm up for the period  
recommended by the manufacturer. Inject a calibration gas into the measurement  
system and adjust the back-pressure regulator to the value required to achieve the flow  
rates specified by the manufacturer. Inject the zero- and the high-range calibration  
gases and adjust the analyzer calibration to provide the proper responses. Inject the  
low and mid-range gases and record the responses of the measurement system. The  
calibration and linearity of the system are acceptable if the responses for all 4 gases are  
within 5% of the respective gas values. If the performance of the system is not  
acceptable, repair or adjust the system and repeat the linearity check. Conduct a  
calibration and linearity check after assembling the analysis system and after a major  
change is made to the system. A calibration curve consisting of zero gas and 2  
calibration levelsshall be performed at the beginning and end of each batch of  
samples.  
5.2 Systems drift checks. After each sample, repeat the system calibration  
checks in section 5.1 before any adjustments to the fia or measurement system are  
made. If the zero or calibration drift is more than plus or minus 3% of the span value,  
discard the result and repeat the analysis.  
5.3 Quality control. A minimum of 1 sample in each batch shall be distilled  
and analyzed in duplicate as a precision control. If the results of the 2 analyzed differ by  
more than plus or minus 10% of the mean, then the system shall be reevaluated and the  
entire batch shall be redistilled and analyzed.  
6. Calculations.  
6.1 Bag sample volume, Bv.  
(
) ( )  
) (  
MV TSTD PM  
(
=
BV  
) ( )  
T M PSTD  
Page 16  
Courtesy of www.michigan.gov/orr  
Where:  
Bv  
Mv  
= Bag sample volume in standard liters.  
= Indicated dry gas meter volume, in liters.  
TSTD = 2930K.  
TM  
PM  
= Meter gas temperature, in 0K.  
= Meter gas pressure, in mm Hg absolute.  
PSTD = 760 mm Hg.  
6.2  
6.3  
Bag sample voc concentration, as voc, Cvoc.  
Cvoc = Ml/Bv  
Where:  
Cvoc = Bag sample voc concentration, as voc, mg/std. liters.  
Ml  
= Weight of voc liquid injected, mg.  
Bag sample voc concentration, as propane, Cc3.  
Cc3 = Rc3*K  
Where:  
Cc3  
Rc3  
= Bag sample voc concentration, as propane, mg C3/std. liter.  
= FIA reading for bag gas sample, ppm propane.  
mg propane / std. liter  
K = Conversion factor, 0.00183  
Response factor, RF.  
ppm propane  
6.4  
6.5  
RF = Cvoc/Cc3  
Where:  
RF  
= Response factor, weight voc/weight propane.  
Total voc content of the input voc containing liquid, as propane, L.  
n
n
n
V IJ WIJ  
RF J  
V FJ WFJ  
RF J  
V AJ W AJ  
RF J  
L =  
-
+
J=1  
J=1  
J=1  
Where:  
L =  
Total voc content of liquid input, calculate as propane, kg.  
VIJ = Initial voc weight fraction of voc liquid J.  
VFJ = Final voc weight fraction of voc liquid J.  
VAJ = Voc weight fraction of voc liquid J added during the test.  
WIJ = Weight of voc containing liquid J at beginning of test, kg.  
WFG = Weight of voc containing liquid J at end of test, kg.  
Page 17  
Courtesy of www.michigan.gov/orr