V-/EPA
United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-82-038 Sept. 1982
Project Summary
Environmental Assessment of a
Low-Emission Oil-Fired
Residential Hot Water
Condensing Heating System
C. Castaldini
The report gives results of tests to
evaluate multimedia emissions from a
condensing hot water residential
heating system equipped with a low-
emission oil-fired burner manufactured
by Maschinenfabrick Augsburg-Nurn-
berg (M.A.N.) of West Germany. Tests
included continuous monitors for flue
gas criteria pollutant emissions and
laboratory analysis of samples utilizing
gas chromatography (GC), infrared
spectrometry (IR), liquid chromatog-
raphy (LC), and gas chromatography/
mass spectrometry (GC/MS) for
organics; and spark source mass
spectrometry (SSMS), atomic absorp-
tion spectrometry (AAS). and ion
chromatography (1C) for trace metals
and anions. Flue gas concentrations of
NOx. SO2, and CO averaged 76. 156,
and 40 ppm, respectively, corrected
to zero % O2. Sulfate and copper were
the primary pollutants in the tank
water discharge, about 1,000 and
500 mg/l, respectively. Concentra-
tions of copper and other trace metals
in this water were attributed to
leaching of heat transfer surfaces
immersed in acidic (pH = 3.0) water.
Organic emissions measured 3.5
mg/dscm in the flue gas and 0.1
mg/l in the waste water. Biological
tests indicated moderate mutagenic
response of the flue gas and moderate
toxicity of waste water to mammalian
cells.
This Project Summary was devel-
oped by EPA 's industrial Environ-
mental Research Laboratory, Research
Triangle Park, NC, to announce key
findings of the research project that is
fully documented in a separate report
of the same title (see Project Report
ordering information at back).
Introduction
A number of low-emission, high-
efficiency residential systems and
burners have been developed recently.
This report describes the results of
extensive emissions testing of one of
these units. The flue gas was analyzed
for criteria pollutants as well as
noncntena organic and inorganic
species. Since the unit was a condensing
hot water heater, water tank composition
was also determined.
The residential heater tested repre-
sents an innovative European design
utilizing a condensing flue gas system
and a high-efficiency low-emission
burner. The burner, shown in Figure 1,
is manufactured by Maschinenfabrick
Augsburg-Nurnberg (M.A.N ) of West
Germany. It utilizes a finely atomized
distillate oil and recirculated hot com-
bustion gases mixed with fresh air to
complete combustion of the fuel in the
burner pipe. The fuel oil can be pressur-
ized to 2.1 MPa (about 300 psi) and is
atomized by a 60° hoHow-cone nozzle
-------
Nozzle
Seal
Damper
Figure 1. M.A.N. residential oil-fired
burner.
delivering about 0.53 ml/s (0.5 gal./hr).
The combustion of the fuel in the mixing
tube produces a stable blue flame The
recirculation of the combustion gases
also causes NO* emissions to be 40-
50% lower than those from a conven-
tional high-pressure atomizing burner
widely used for residential oil-fired
furnaces. Because the M.A.N. burner
recirculates the combustion gases
internally within the burner pipe where
combustion is completed, retrofit instal-
lation on existing residential heating
systems is possible. Although other
blue flame burner designs have been
developed and implemented in the U.S.,
the retrofit capability of the M.A.N.
design has made it attractive as a
potential technique for reducing NOx
emissions from existing residential
units.
The firebox of the furnace is completely
immersed in water. The water level
reaches approximately 2 cm (less than 1
in.) below the top of three exhaust
pipes. This water level is controlled by
positioning the condensed water drain
spout. Before the combustion products
exit the furnace exhaust duct, they
pass over the water surface andthrough
a series of baffles and heat exchanger
tubes. The cooling water, which absorbs
heat from the furnace and carries it to
the residence, enters through a heat
exchanger tube near the top of the
furnace and then flows through the
immersed copper coils before it exits.
Condensation of the water in the flue
gas begins when cool water meets
combustion products on their way out of
the tank, condensing practically all the
water produced by combustion of the
fuel.
Condensing heating systems such as
this can achieve thermal efficiencies
exceeding 95% under normal cyclic
operation. This high thermal recovery is
a significant improvement over cyclic
efficiencies of conventional residential
heating systems, normally about 75-
80%.
Tests were performed with the unit
operating in a typical cyclic mode. Cycle
frequency of the burner was controlled
by adjusting the setting of the tank
water thermostat and the cooling water
flowrate. A thermostat setting of
approximately 54°C (129°F) and a
cooling water flowrate of 107 ml/s (1.7
gal /hr) resulted in burner cycle fre-
quencies of 11-14 min on, 22-25 min
off.
The sampling and analysis procedures
used in this test program conformedto a
modified EPA Level 1 protocol for the
gas and liquid discharge streams. Flue
gas measurements made at the exit of
the furnace at about 1 m (3 ft) from the
base of the uninsulated exhaust pipe
included:
• Continuous monitors for NO, NO,,
CO, C02, 02, and total unburned
hydrocarbons (TUHC).
• Source Assessment Sampling
System (SASS) train sampling.
• EPA Method 5 for solid and con-
densable particulate mass emis-
sions
• EPA Method 8 for S02 and S03.
• Grab sample for onsite analysis of
Ci to C6 hydrocarbons by gas
chromatography.
• Bacharach smoke spot. The analy-
sis protocol for SASS train samples
included:
• Analyzing the filter catch, ashed
XAD-2 resin, and the first impinger
solution for 73 elements using
spark source mass spectrometry
(SSMS) and for Hg using cold-
vapor atomic absorption spectrom-
etry (AAS)
• Analyzing the second and third
impinger solutions for As and Sb
using furnace AAS techniques,
and for Hg using cold-vapor AAS.
• Extracting the XAD-2 sorbent resin
in a Soxhlet apparatus using
methylene chloride, concentrating
the extract to 10 ml, then deter-
mining the organic content of the
extract in two boiling point ranges:
100-300°C by total chromato-
graphable organics (TCO) analysis
and >300°C by gravimetry.
• Further concentrating the extract
to 1 ml and analyzing for the 58
semivolatile organic priority pol-
lutants by gas chromatography/
mass spectrometry
Water tank discharge samples collected
were subjected to inorganic analysis by
SSMS and AAS for Hg, As, and Sb; and
to anion analysis for chloride, nitrate,
and sulfate by ion chromatography.
They were a Iso extracted with methylene
chloride and subjected to the organic
analysis protocol noted above.
The XAD-2 sorbent resin extract was
also subjected to liquid chromatography
separation into seven polarity fractions
on silica gel to give compound category
composition information. In addition,
infrared spectra were obtained for the
gravimetric residues of all extract
samples (whole samples and liquid
chromatography fractions).
The XAD-2 sorbent extract and the
tank water discharge were subjected to
mutagenicity and toxicity evaluation
using the Level 1 Ames mutagenicity,
CHO cytotoxicity, and the whole animal
acute toxicity in rodent (RAT) bioassay
tests.
Summary and Conclusions
Table 1 lists flue gas emission levels
of CO, C02, NO, N02, TUHC, particulate,
SOX, and smoke in the flue gas measured
during the period of firing. During the
test there were peaks of CO and HC
emissions at the start and end of
burner-on times. The peak emissions at
the start of each cycle are included in
the reported emissions; however, the
effects of burner shut-off were not
included. Since the blower and the fuel
pump were shut off at the same time,
there was no forced air when the burner
was shut off. Thus, the combustion air
flowrate is unknown, and the CO and
HC emissions at the end of the firing
cycle cannot be evaluated.
Burner start-up peak emissions
averaged 150 ppm for CO and 1 5 ppm
for HCs. The NO started at zero and
reached approximately 70 ppm on the
average at 1.9% average excess O2.
Smoke emissions measured with the
Bacharach hand pump kit were zero
during the entire burner-on period. NO
emissions averaged 76 ppm at zero % O2
over the duration of the test. This level is
a 40% reduction from conventional
residential heating systems burning
distillate oil. Condensation of flue gas
moisture apparently removed all N02
from the flue gas. Analysis to determine
anions in the tank water and condensate
drain collected during the test showed,
in fact, that nitrates were absorbed in
the water Tank water nitrate levels
reached 7 mg/l. The nitrogen content of
the oil burned averaged 0.04%, making
it a relatively high nitrogen distillate,
leading to correspondingly high NO
emissions.
Sulfur species (SO2 and SOa) in the
exhaust gas were analyzed by EPA
Method 8. As expected, S02 was the
-------
only sulfur species found in the exhaust
gas. Both gaseous SO3 and any con-
densed-phase sulfate were absorbed in
the water, as indicated by the sulfate
content of the tank water, which
reached 990 mg/l. Chloride levels in the
tank water reached 1.0 mg/l.
Bacharach smoke emissions were
measured throughout the test program
at various time intervals after fuel light-
off Smoke numbers were consistently
zero throughout each firing period The
absence of smoke and relatively low CO
emission peaks during burner light-off
are attributable to the fuel oil delay valve
of the M.A.N. burner which prevented
ignition for approximately 15 sec after
the blower started.
Table 2 shows results of the organic
analysis of SASS samples by boiling
point range. 100 to 300°C (TCO) and
greater than 300°C (gravimetric). The
flue gas results indicate that 74% of all
the organic emissions were measured
by TCO analysis of the XAD-2 extract.
Still, total concentration of organic
matter m the flue gas measured only 3.5
mg/dscm. The water analysis results
indicate that some organic matter
condensed in the_water, however, the
total concentration measured in the
tank water discharge was less than 0.1
mg/l.
Infrared spectra of the gravimetric
residue of the XAD-2 extract and the
tank water discharge extract suggested
the presence of aliphatic hydrocarbons
and oxygenated species (carboxyhc
acids, esters, aldehydes, alcohols,
etc.).
The XAD-2 extract was subjected to
liquid chromatography separation.
Results of this fractionation combined
with infrared analysis of the gravimetric
residue of sample fractions suggested
that, of the 3.5 mg/dscm of organics
emitted by the furnace, about 92% are
aliphatic hydrocarbons and the remain-
ing 8% are oxygenated species. This
suggests that the bulk of the organics
emitted consist of unburned fuel; the
remainder is partially oxidized fuel.
Gas chromatography/mass spec-
trometry analysis of sample extracts
was performed to determine the 58
semivolatile organic priority pollutants.
Of these, only naphthalene and phe-
nanthrene or anthracene were detected,
as also shown in Table 2.
Results of a SAM/IA evaluation of the
data obtained in this test program are
given in Table 3, which shows species
with discharge severity (DS) greater
than 0.1. In the flue gas stream, NO and
Table 1. Flue Gas Emissions3
Species
flange
Average
Oz. percent dry
COz, percent dry
HzO, percent
C0a, ppm at 0 percent Oz
ng/J
NO. ppm at 0 percent 02
ng/J as NO 2
NOz
TUHC. ppm at 0 percent Oz
ng/J as CsHa
SOz, ppm at 0 percent Oz
ng/J
SOz
Solid paniculate. ng/J
(Method 5)
Condensable paniculate, ng/J
(Method 5)
Smoke. Bacharach
1 4 to 2.4
12.6 to 14.0
2.7 to 3.0
15 to 51
4.5 to 15
68 to 79
33 to 39
0"
0.5 to 9.0
0.2 to 4. 1
—
—
Ob
—
—
0
1 9
12.9
2.9
40
12
76
37
0
3.3
15
56
106
00
13
1.4
0
"Includes peak emissions at the start of burner-on cycle.
^Nitrates and sulfates were absorbed in the tank water.
Table 2. Organic Emissions Summary
Flue Gas
mg/dscm
Tank Water Discharge
mg/l
Total Chromatographable
Organics (TCO}
2.6
Gravimetric (GRAV)
Total
Naphthalene
Phenanthren e/A nthracene
0.9
3.5
fjg/dscm
36
2
<0.1
<0.1
vg/t
0.4
0.08
Table 3. Discharge Severities Greater Than 0.1 for the Low-Emission Condensing
Furnace System
Pollutant Species
Cu
sot
/VOx
SOz
Fe
Ni
Cr
Se
CO
Mn
S
Zn
Aldehydes
Pb
Carboxylic acids
Na
Flue Gas
Emitted
Concentration
/jg/dscm
59
—
9.9 x 704
1.3 x 10s
15
8.6
3.4
0.25
3. 1 x 70"
1.4
480
7.5
100
2.8
200
220
DS
0.03
11
10
0.015
0.57
3.4
0.001
0.77
<0.001
0.48
0.002
0.40
0019
0.20
0.11
Tank Water
Discharge
Concentration
H9/I
5.0 x 705
5.3 x JO5
7. Ox 70"
7,000
700
700
—
730
7.0 x /O4
—
70
—
—
DS
100
67
67
4.4
2.8
2.0
—
0.76
0.40
—
0.28
—
—
* US.GOVERNMENTPRINTimOFFICE:IN).559-017/0800
-------
SO? emissions were responsible for the
highest DS values, both exceeding unity
by a factor of nearly 10. CO and THCs
(primarily aliphatics) were present in
nonhazardous concentrations (DS
<1 0). Four elements with DS greater
than 0 1 were found m the flue gas Cr,
Ni, Na, and S, with only Cr having a DS
exceeding unity Both Cr and Ni emis-
sions, however, are suspected contam-
inants inherent in sample preparation
for trace element analysis
Two organic categories had potential
DS values greater than 0 1. However,
these organic category DS values were
calculated under the conservative as-
sumption that all the organic content
assignable to the respective category
consisted of the compound with the
lowest discharge multimedia environ-
mental goal (DMEG) potentially present
in the sample
Trace elements in the tank water for
which DS exceeded unity were Cu, Cr,
Fe, Ni, and Se. Cu concentrations, in
fact, exceeded those of any other
element detected in the tank water This
high concentration of Cu is attributed to
leaching of heat exchanger copper coils
immersed in the warm acidic tank
water. In fact, concentrations of most of
the other three trace elements of
potential concern can be attributed to
the leaching of metal surfaces in
contact with the tank water Sulfate (as
sulfunc acid) in the tank water repre-
sents the next greatest potential concern.
Results of these bioassay analyses
showed that the XAD-2 sorbent extract
exhibited moderate mutagenicity in the
Ames bioassay, and low to nondetec-
table toxicity in the CHO assay The tank
water discharge exhibited moderate
cytotoxicity in the CHO assay, non-
detectable toxicity in the rodent whole
animal test, and nondetectable muta-
genicity in the Ames bioassay
C. Castaldini is with Acurex Corp., Mountain View, CA 94042.
Robert E. Hall is the EPA Project Officer (see below).
The complete report consists of two volumes, entitled "Environmental Assess-
ment of a Low-Emission Oil-Fired Residential Hot Water Condensing Heating
System:"
"Volume I. Technical Results," (Order No. PB 82-239 344; Cost: $12.00,
subject to change)
"Volume II. Data Supplement," (Order No. PB 82-239 351; Cost: $16.50.
subject to change)
The above reports will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone. 703-487-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
PS 00003a9
------- |