Review of Compliance Test Methods for Determining
Formaldehyde Emissions from IC Engines and Turbines with
Potential Application to Other Combustion Sources
INTRODUCTION
This is a review of techniques that are being used for measuring formaldehyde emissions
from IC engines and turbines. Since the source operators could use the procedures for
data gathering in the EPA rule-making process, a clear understanding of the strengths and
deficiencies of these methods is necessary. Although the comparisons are based upon
engines and turbine measurements, the observations are potentially applicable to other
combustion sources.
The following methodologies are reviewed:
1. The extractive FT-IR method
2. The aqueous acetylacetone impinger method (NCASI or Celanese Method).
3. The DNPH impinger methods (CARB 430 and EPA Method 0011).
4. The dry DNPH method (the Ashland Method).
SUMMARY
The extractive FT-IR is currently the best of the EPA recognized methods for measuring
formaldehyde concentrations in engine combustion exhaust gas. FT-IR might also be the
least expensive method where formaldehyde, other aldehydes, and criteria pollutant
emissions will be measured over a range of operating conditions or on multiple units at
the same site. But measurement costs are expensive when only one or a few
formaldehyde measurements at a single source are required. The acetylacetone method
can be applied at a low cost and it has shown promise as an accurate and reliable method.
The Gas Research Institute (GRI) and Radian plan to perform an EPA Method 301
validation for the acetylacetone method in the coming months and expect EPA approval
for its use at natural gas fired units within six months to a year. In the meantime, we
recommend that comparative tests using this method should be conducted whenever the
opportunity exists within EPA and ICCR source group source test plans. The DNPH
impinger and the dry DNPH methods are less reliable than the other two methods for
engine combustion gas measurements.
DTSCUSSTON
August 28, 1998, FORMALD2.WPD
-------
1. The Extractive FT-IR Method
The extractive FT-IR is currently the most accurate and reliable measurement technique.
It has been reviewed by the EPA and has been approved for natural gas fired combustion
sources. However, the extractive FT-IR method is presently relatively expensive and
complex procedure. Field measurements require a trailer or van with expensive
equipment and highly trained personnel. Once on the site, however, the instrument and
the personnel can make many measurements quickly. We believe that FT-IR will likely
be the least expensive alternative for field tests where formaldehyde and other aldehydes
will be measured over a range of operating conditions along with criteria pollutant
emissions (CO, NO, N02), or if there are multiple units at the same site. The highly
trained testing personnel at the site can recognize test interference problems, make on site
adjustments in measurement procedures, and provide the operators with real time
emissions data.
The FT-IR measurement procedure might be simplified to specifically address
measurement of formaldehyde and criteria pollutant emissions from combustion sources.
In the future, field-ready instruments might be pre-tuned to give direct concentration
readings. This will enable the use of less specialized instrument operators and lower the
testing costs. However, since such instruments are not yet available, the cost for
measuring only formaldehyde emissions at a single field source will likely be high relative
to the other three methods. Accurate cost comparison data are needed and these will
depend upon the specific test program details.
2. The Aqueous Acetylacetone Method ("AAA Method")
This is a method that was published in the National Council of the Paper Industry for Air
and Stream Improvement, Inc., Technical Bulletin No. 684 and has been widely used by
the Celanese Corp. The source gas is drawn through midget impingers containing chilled
water. The formaldehyde concentration in the impingers is determined by the AAA
procedure. This involves the reaction of the acetylacetone with formaldehyde to produce
a colored derivative which is then measured by colorimetric analysis. The sampling
procedure is inexpensive and measurements can be carried-out by most environmental
contract laboratories.
The GRI has recently funded Radian to carry out a series of comparative studies
between this technique and the FT-IR on emissions from a gas fired internal combustion
engine and got excellent agreement. No evidence was found of interference from
nitrogen oxides or acetaldehyde. The GRI plans to proceed with EPA 301 validation test
and obtain EPA approval within 6 months to a year.
After validation is obtained, we believe that this method will be an excellent alternative
method to the already validated FT-IR method when only formaldehyde concentrations
August 28, 1998, FORMALD2.WPD
-------
are of interest and no criteria pollutant data are needed. For such applications, it will be
an accurate, precise, and least-cost alternative.
3. The DNPH Impinger Methods
Most of the current formaldehyde emissions data have been gathered by DNPH impinger
methods. The DNPH method was initially developed for analysis of ambient air and vent
streams and subsequently applied to combustion gases without extensive validation tests.
The GRI research program has identified inaccuracies with the method when they tested
IC engines using the FT-IR and DNPH impinger methods. The problem was initially
traced to a competitive reaction of nitrogen dioxide with DNPH that depleted the DNPH
reagent and appeared to be limited to measurements on sources with relatively high
nitrogen dioxide concentrations. Subsequent laboratory tests indicated that high NO
levels may also be a problem. However, it is incorrect to assume that all data collected
by DNPH procedures are invalid. GRI and Radian are in the process of analyzing data
from a laboratory study on DNPH method interferences and intend to present their
findings in a GRI Topical Report planned for release in October, 1998.
With the advent of the AAA Method, the potential need for the DNPH methodology for
combustion source measurements is greatly diminished since the AAA Method will be
suitable for testing on all IC engine types.
4. The Dry DNPH Method (Ashland Method).
This method consists of sampling the emissions gases using a proprietary sorbent resin
that has been coated with DNPH. It has been used by Radian in a series of tests
sponsored by GRI and the American Petroleum Institute (API) together with the
validated FT-IR method. Measurements were made on both engine and turbine emissions.
Although special sampling procedures were used to cool the sorbent tubes to achieve the
proper moisture content on the DNPH tubes, poor agreement was found in both
instances. Because the analyses of the tube had to be carried out at a designated
laboratory, costs for this procedure were higher than that for the AAA Method.
No further tests of this method are planned in GRI or API sponsored research programs.
August 28, 1998, FORMALD2.WPD
------- |