( 3.
United 6WDWHV
Department of
Agriculture
Forest Service
Pacific Southwest
Forest and Range
Experiment Station
Microwave 2YHQV for Drying Live :LOGODQG
Fuels: an assessment P.O. Box 245
Berkeley
California 94701
Research Note
PSW-349
Richard W. McCreight
J'
March 1981
),il)'; ,,
T
McCreight, Richard W.
1981. Microwave ovens for drying live wild·
land fuels: an assessment. Res. Note
PSW-349, 5 p., ill us. Pacific Southwest
Forest and Range Exp. Stn., Forest Serv.,
U.S. Dep. Agric. , Berkeley, Cal if.
Using commercially available microwave
ovens to determine the moisture content of live
forest fuels is likely to be unsatisfactory. The
conditions affecting drying temperature and
handling techniques can vary considerably.
And procedures for using microwave ovens
have not yet been standardized. Three brands of
microwave ovens used in the field during the
1978 fire season in southern California were
tested in the laboratory. The indicated moisture
content of the fuels tested varied from 3 to 40
percent, depending on temperature. handlin g.
and charring effects. Consequently , it is not
possible to compare fuel moisture meas­
urements obtained by microwave ovens at dif·
ferent locations for the purpose of assessing
wildland fire behavior .
Retrieval Terms: forest fuels. fuel moisture
content, moisture measurement. microwave
ovens. southern California. reliability
he moisture content of living forest
fuels affects the behavior of wildland
fires. It can be measured in a number of
ways, including the reflux solvent distil­
lation method, ' conventional drying
ovens, and kitchen-type microwave
ovens. Drying fuels in conventional dry­
ing ovens appears to be best for monitor­
ing levels of and trends in living chapar­
ral moisture content. 2 Microwave ovens
have been suggested as a faster , more
convenient technique for drying fuel
samples under field conditions.-' They
have been tested on dead, woody fuels in
Montana, and found to yield the same
results as standard oven drying .•
The use of microwave oven drying for
living fuel samples in California has
raised questions as to whether it can pro­
duce accurate enough results for assess­
ing regional fuel moisture trends. Such
questions have risen, in part , because
procedures for using microwave ovens
are not yet standardized .
In microwave oven drying, fuel sam­
ples are subjected to a high-intensity
electromagnetic field. When the micro­
wave energy comes into contact with an
object, it is either reflected, absorbed, or
transmitted-depending on the nature of
the material, angle of wave incidence,
and frequency. In a moisture cellulose
sample , microwaves quickly penetrate
significant distances into the material . In
the process, the water molecule-dipole
oscillations are increased and the sample
dries from the interior outward.
The effect of temperature on the fuel
sample and water-heat relationships must
be considered in determining moisture
content. When fuel samples are sub­
jected to varying temperatures, the dry­
ing curves tend to diverge , resulting in
varying fuel moisture values 5 ifiR· /) .
Although the effect of temperature on
moisture content is well known, the
magnitude of this influence has yet to be
determined . 6
To answer some questions raised
about microwave oven drying, three
commercially available brands were
tested in the laboratory to assess (a) the
temperature variations among ovens, (b)
the reproducibility and variation of the
moisture data , and (c) the effect of char­
ring on the results. The indicated
· moisture content of the living forest fuels
110
Mo1sture content
(percent)
+ 72
u
70
78 5
93 9
-
+-+-+-+-+3
4
5
6
7
8
Drying time (hours)
the indicated moisture content of manzanita
leaf samples dried in a mechanical convec­
tion oven.
tested varied from 3 to 40 percent, de­
pending on temperature, handling, and
charring effects. The results suggest that
moisture content measurements within a
region obtained by microwave oven dry­
ing are not comparable for the purposes
of fire management and prescribed burn­
ing . Such measurements may, however,
be useful in showing local trends in
moisture content.
Only the first experiment tested for
differences between all three ovens while
the remaining experiments served to test
for effects on the fuel moisture determi­
nation process. It is assumed that effects
will vary if ovens are significantly differ­
ent.
ENERGY VARIATION
Experiment I
PROCEDURES
The three brands of microwave ovens
tested were Kenmore, Tappan, and
Sharp carousel. 7 All three operated on
120 volts , a.c., at a frequency of 2450
MHz. Voltage output, in watts, and oven
capacity volume of the ovens varied. The
Sharp oven featured four power settings,
but only the continuous 600-watt power
level was used. The Tappan oven pro­
duced 600 watts of power, and the Ken­
more 400 watts . Oven cavity size varied
from 0.6 fe (0 .02 m') for the Kenmore,
to !.3ft' (0 .04 m' ) for the Sharp, and 1.4
ft' (0.04 m') for the Tappan.
In all three ovens, temperature was
unknown . Control was achieved by vary­
ing the exposure time (min/sec). The use
of metal objects , including temperature
sensors , could have damaged the oven by
reflecting microwave energy into the
power element. Indirect techniques for
measuring energy-temperature charac­
teristics, therefore, were required for this
study.
These experiments were done to ascer­
tain those characteristics of microwave
ovens necessary in determining the
moisture content of living fuels :
• Experiment I (Kenmore, Sharp, Tap­
pan ovens): test differences in energy
between ovens and times by measuring
moisture loss at specific times.
• Experiment II (Sharp): test for internal
energy variations by multi-level
isothermal enhancement of energy pat­
terns .
• Experiment III (Sharp, Tappan): test
for differences between treatments by
following the drying procedures de­
scribed by Palmer and Pace, ' used with
and without modification.
• Experiment IV (Sharp): test for effects
of pyrolysis by inducing and measuring
it under various influences.
Differences in energy between ovens
and times were determined by using three
polypropylene jars, each containing 100
g distilled water, which were placed on
marked locations in the oven cavity and
exposed for 1-, 3-, and 5-min periods .
Moisture loss was assessed by weighing
samples after each exposure period on a
190-g capacity analytical balance to the
nearest 0.01 mg. Five moisture loss
measurements used fresh samples and
were made for each time and oven . A
two-factor analysis of variance was used
to evaluate the data .
Experiment II
To provide a clearer picture of energy
variation within a microwave oven , 15­
by 15-inch (38- by 38-cm) blotting paper
was soaked for I min in distilled water
and mounted on a plastic table. The table
was placed in the Sharp oven with the
carousel unit removed, at one of three
levels, 1, 4.5, and 7.5 inches (2.5, 11.4,
and I 9 em) above the oven floor, and
exposed to maximum power for 2-min
intervals. At the end of each 2-min
period, the table was removed and dry
areas delineated and labeled with the
total drying time. The procedure was re­
peated three times for each level. A sec­
ond aspect of this test involved placing
one to three water samples at various
locations on the l-inch (2.5-cm) high
table and repeating the exposure delinea­
tion process .
0 .32 em) diameter old-growth chamise
(Adenostoma fasciculatum). Eighty
samples weighing about 75 g each (fresh
weight) were dried in pairs in the Tappan
oven with a 30()-g pumice block used as
an energy absorber. Samples were
heated, removed, and stirred for ventila­
tion, weighed, and reheated. Initial oven
time was 3.5 min and was linearly de­
creased to 1.5 min. The 1.5 min expo­
sures continued from three to seven times
or more until the change in sample
weight was equal to, or less than, 0 . 1 g
between successive weighings.
Ninety additional chamise samples di­
vided into three batches, were subjected
to different treatments by changing the
energy absorber, oven type, or both . No
changes were made in the handling pro­
cedure. Initially, the pumice energy ab­
sorber was replaced with a glass beaker
containing 100 g distilled water. Then
the Tappan was replaced with the Sharp
oven and the energy absorbers again al­
ternated. Two subsamples of each sam­
ple run were distilled in m-xylene at
282°F ( 139°C) for 90 min and used as a
control . The five methods : Tappan with
pumice, Tappan with water , Sharp with
pumice, Sharp with water , and
m-xylene, were evaluated by a one-way
analysis of variance , and at-test of paired
differences was used to evaluate the per­
cent difference between oven dryings
and the m-xylene method .
Experiment IV
Forty-five 0-to 0.125-inch (0 to 0.32
em) diameter chamise samples were used
in measuring sample pyrolysis . About
two-thirds of the samples were dried in
the Sharp oven to the point of charring .
The remaining one-third was removed
and stirred during the drying process to
avoid pyrolysis. Subsamples from each
group were distilled in m-xylene and
compared .
RESULTS
Experiment III
Experiment I
Guidelines for microwave oven drying
developed by Palmer and Pace' served as
a baseline drying procedure for determin­
ing the reproducibility and variation of
moisture results foro-to 0. 125-inch (0 to
Results of the energy test revealed dif­
ferences , significant at the 5 and I per­
cent levels, between ovens , times, and
oven-time interactions (table /) . Simply
stated, the indicated moisture values
2
Table 2-Effects of changing the type of oven or
energy absorber on the residual moisture in
chamise fuel samples (the average percent dif­
ferences between microwave oven-dried and
xylene-distilled samples)
Table 1-Analysi.r of variance for ovens and times,
hy source; SS =sum of squares, d.f. = deKrees of
freedom, MS = mean square , F = F-ratio
ss
Source
191.07
15.27
158 .57
16.99
0.23
Total
Oven
Time
Oven-time
Error
F
44
2
2
4
36
7.63 1203.94
79.29 12503.75
4.25
670.03
0.01
Oven
Absorber
X residual
moisture
(xylene­
microwave)
S.D.
Percent
Tappan
Tappan
Sharp
Sharp
were different between treatments. Av­
erage percentage of moisture loss for
5-min exposure periods serves to contrast
each oven (jig. 2). The 1- and 3-min
exposure periods showed the same con­
trast, but at smaller scales.
90 c
80 70 60 Sharp
50
40
30
Tappan
Kenmore
10
Figure 2-The three ovens were
compared for average moisture
loss from 100-g water samples for
5-min exposure periods .
1.1
1.6
1.1
1.7
Experiment IV
Experiment II
The isothermal enhancement, or de­
lineation ofhot-cold patterns of the Sharp
oven interior in operation (jig. 3), re­
vealed that microwave energy emitted
from the top of the cavity (col. A, top to
bottom) is directed down and outward,
ensuring maximum wave energy mixing
in the midlevel region of the oven, and
then is refocused onto the oven floor as a
series of well-defined hot and cold spots.
Isothermal patterns were fairly consis­
tent between successive tests for any one
level (note 2.5-cm level, bottom of col. A
and B), but changed when water samples
were introduced. The samples dissipated
the energy field, thereby increasing the
paper drying time. The drying range for
blotting paper without samples was 4 to
10 min; that with three samples was 8 to
16 min.
3.2
7.2
4.6
7.8
Pumice
Water
Pumice
Water
Hot
Cold
Figure 3-lsothermal enhancement of Sharp
microwave oven interior during operation. Col­
umn A displays microwave energy distribution
and intensity for three descending horizontal
levels; Column B shows the dissipating effect
of moist samples on the energy field.
Experiment III
When chamise samples were dried in
the Tappan oven, repeating the Palmer
and Pace 3 experiment, the average re­
sidual moisture (the difference between
oven and m-xylene results) remaining in
the oven-dried samples was 3.2 percent
and the correlation coefficient was 0.99.
These values were identical to those of
the Palmer and Pace 3 study; however, the
standard deviation was I. 9 in the Palmer
and Pace' study and I . I in this study.
Alterations made in the treatments,
that is, changing the energy absorber, or
the oven itself, resulted in variations in
the indicated moisture contents (table 2).
Statistically, this variation is significant
at the 5 and I percent levels (table 3).
Furthermore, the oven-dried and
m-xylene samples differed by more than
5 percent on the basis of a two-tailed
t-test. Pickford also found a fairly high
correlation coefficient between micro­
wave drying and xylene distillation re­
sults, but reported an average 6. 7 percent
residual moisture remaining in cedar
(Cedrus) needles dried in a microwave
oven. 8
3
Results of pyrolysis tests indicated that
when power continued without frequent
interruption to ventilate samples,
"runaway" heating was inevitable. The
hot area within the sample absorbed more
energy than the cooler portions and was
increased further in temperature. This
temperature increase continued until the
sample charred. Amount of charring, by
volume, ranged from 5 to more than 50
percent of the sample. By comparing any
three samples in a test set, a similar range
in amount of charring, with the center
sample having the most damage, was ob­
served. Pyrolysis times varied from 6.5
min for samples with a moisture content
of 57 percent, to 8 min for those with a
moisture content of 70 percent.
To avoid sample pyrolysis, exposure
times were reduced by as much as 63
percent and samples removed and stirred
after each exposure to allow them to
cool. This increased sample handling and
extended the total drying time from about
45 min to I h 25 min for samples ranging
in moisture content from 57 to 73 per­
cent. The indicated moisture contents,
however, were more in agreement with
Table 3-Ana/ysis of variance for treatments, by
source, SS = sum of squares, dj. = degrees of
freedom, F = F-ratio
Source
Among
treatments
Within
treatments
Total
ss
F
232
3
216
57
448
60
20.38
xylene results when this was done. Sam­
ples that were exposed to the point of
slight charring showed a 10.5 percent
higher indicated moisture content than
the xylene control group and a standard
deviation of 3.6 as against 1.2 for the
xylene samples. When the exposure
times were reduced, the microwave indi­
cated moisture content was about 5. 9
percent lower than the xylene group. The
standard deviation also improved, with
1.6 recorded for microwave and 1.2 re­
corded for xylene samples.
On the basis of the entire series of
tests, I 0 variables were identified as hav­
ing differing degrees of influence on
sample drying temperature. They were:
oven type, energy output, distribution,
and absorption by sample, additional
energy absorbers, container type, fuel
type, fuel moisture content, bulk density,
and handling.
Each of these variables exhibited suf­
ficient individual influence to alter tem­
perature and moisture results. By com­
paring residual moisture fluctuations, of
the 10 variables identified, energy distri­
bution, additional energy absorbers, fuel
moisture content, sample bulk density,
and handling technique were the most
critical to temperature control. Nonethe­
less, to assure consistent moisture data,
most variables required nearly identical
duplication between tests.
DISCUSSION
The basis for deriving accurate, repro­
ducible fuel moisture data from any dry­
ing method lies in the accuracy and
reproducibility of the determination pro­
cedure itself. Procedure is actually a con­
sortium of component phases including
sample collection, preparation, drying,
and calculation of moisture values. In­
consistency in any phase can lead to in­
accurate and unusable data . In the drying
phase, temperature is one of the most
vital variables affecting moisture data.
When the drying temperature is in­
creased to 2!2°F (I 00°C), water within a
fuel sample vaporizes and accounts for
most of the weight loss. Some volatiles
also will start to evaporate at this temper­
ature, although more oils and crude fats
are distilled off as the temperature is in­
creased. As the temperature reaches
280°F ( 173°C), pyrolysis starts and in­
creases in rate depending on the organic
composition and inorganic salt content
and percentage. If the fuel temperature
goes above 212°F (I 00°C) for any length
of time, moisture contents will be overes­
timated because of volatiles lost.''
The inherent problem of determining
live fuel moisture by using the three
ovens tested is that microwave energy is
controlled by time exposures, not tem­
perature selection. Temperature is un­
known and easily influenced. Samples
require frequent handling to control tem­
perature, avoid pyrolysis, and compen­
sate for strong temperature gradients
within the oven and samples themselves.
This handling increases the probability of
human error. Most importantly, indi­
cated moisture values reflect, often to a
high degree, the influences of these tem­
perature variations and handling errors.
Although handling guidelines can help
control drying temperatures, they do not
provide the means for standardizing
them. To do this requires standardizing
most of the 10 variables listed in the
Results section. The irony of this is that
fuel moisture, one of the major variables
influencing temperature, is the unknown
being sought. Consistent, live fuel
moisture data is possible when the same
oven type, energy absorber, sample type,
sample bulk density, and handling tech­
nique are carefully and routinely re­
peated . Oven data may vary consistently
by 7 percent or more from xylene results,
however, depending only on the influ­
ence of the energy absorber used
(table 3).
Many of the newer microwave ovens
feature a temperature probe used as a
thermostat to control temperature. Al­
though this represents a step forward, it
does not necessarily solve the problems
relating to temperature variation. Figure
3 shows variations of temperatures in a
relatively small area. Monitoring tem­
perature for only one point, or one sam­
ple will not assure a uniform drying tem­
perature for the surrounding area or other
samples, unless the probe is continually
moving through the energy field, thereby
representing an average temperature.
Rotating the sample through an energy
field, however, does not solve the prob­
lem of temperature variation within the
samples themselves. Close examination
of burned samples taken from the Sharp
oven revealed a strong temperature gra­
4
dient within them. Samples exhibited
smoldering interiors, although the sur­
rounding outer material remained green
with considerable condensation present.
Only frequent mixing of the material it­
self was found to offset these gradients.
Removing fuel samples and stirring
them can prevent runaway heating and
compensate for temperature gradients;
however, the handling required and the
risk of losing sample contents when
transferring material in and out of a con­
tainer one-half dozen times or more
encourages a higher probability of error .
When chamise is partially dry, for exam­
ple, careful transfer of contents provides
enough disturbance to cause many of the
small needles to snap off the stem as
projectiles , traveling as far as 3 feet (I
m). Unless careful attention is given to
this, and all material recovered, substan­
tial measurement error (from 0.5 to 15
percent) can result. When replacing
sample material in the container, it is also
necessary not to overcompact the sample
so this increases the possibility of sample
pyrolysis even at low exposure times. If
pyrolysis occurs, measurement errors of
10 to 45 percent or more can result.
Because of all the variables involved,
data derived from using the three micro­
wave ovens tested are relative meas­
urements, of some use for establishing
local fuel moisture trends. These data,
however, are limited in other applica­
tions. Of most value is the inability to
compare fuel moisture data between field
stations, because data could reflect a 5
percent or more fluctuation in values re­
sulting from temperature variations
alone. Even larger variations on the order
of 15 percent or more could result if sam­
ples were burned or material was lost
during handling . These effects would act
to nullify or exaggerate naturally occur­
ring moisture variation for the area,
thereby reducing the usefulness and
cost-effectiveness of the overall monitor­
ing system .
Although safety practices in using
microwave ovens were not part of this
study, certain precautions are urged . By
law, visible warnings alerting users to the
danger of exposure to microwaves are
required. No safety monitoring program
now exists, although many microwave
ovens designed for the ''kitchen
environment'' are actually used under
more rugged conditions in the field.
NOTES
'Buck , Charles C., 1939. The solvent distillation
method for GHWHUPLQLQamoisture content offorest
litter . J . For. 37 :645-651.
' Countryman, Clive M ., and William A . Dean .
1979 . 0HDVXULQa moisture content in living
chaparral: afield user' s manual . Gen. Tech. Rep .
PSW-36 , 27 p . Pacific Southwest Forest and Range
Exp . Stn . , Forest Serv . , U .S . Dep . Agric . , Berke­
ley , Calif.
' Palmer, Thomas Y., and George D. Pace . l974 .
Microwave ovens for drying wildland fuels . 1.
Microwave Power 9(4):289-293.
4
Norum, Rodney A. , and William C. Fischer.
1980 . 'HWHUPLQLQa the moisture content of some
dead forest fuels using a microwave oven. Res .
Note INT-277, 7 p., Intermountain Forest and
Range Exp. Stn. , Forest Serv. , U.S . Dep . Agric . ,
Ogden , Utah .
'Personal communication from Clive M.
Countryman, Pacific Southwest Forest and Range
Exp . Stn . , Forest Serv . . U.S. Dep . Agric., October
17, 1978.
•Anderson , Hal E. , R . D. Schuette , and R. W .
Mutch. 1978 . Timelag and equilibrium moisture
content of ponderosa pine needles . Res. Paper
INT-202, 10 p . Intermountain Forest and Range
Exp . Stn. , Forest Serv ., U.S . Dep. Agric., Ogden,
Utah.
' Trade names and commercial enterprises or
products are mentioned solely for necessary infor­
mation. No endorsement by the U.S . Department
of Agriculture is implied .
'Personal communication from Stewart G.
Pickford, College of Forest Resour., University of
Washington, Seattle, Wash . , October 23, 1977 .
•Personal communication from Hal E. Ander­
son , Intermountain Forest and Range Exp . Stn .,
Forest Serv . , U.S. Dep . Agric., August 3, 1978 .
Author:
RICHARD W. McCREIGHT is a forestry technician assigned to the Station ' s research
unit studying the management of chaparral and related ecosystems in southern California,
with headquarters at Riverside, Calif. He attended the University of California, Riverside ,
where he earned his bachelor' s degree in physical geography (1979) .
5