William H. Frandsen,Intermounta
n F re Sc encesLaboratoryForestSefvce U S Departrnenl
of Agr cu ture
lvlssoula.N4onlana.
59807
Burning Rate of Smoldering Peat'
Abstract
S|nl)dl'ringgnr|nd1les1LeaddorLlrl!l)out:.|lL
\;th p.ek t(.nrp.rrrtufe!
g t u r d j h . h r s I ) ,)r , , r l i r
Canadianspllaglullpealn|l)ss!s|l.P|l's.n1uti\.]|ue]'ThetlnetobuIna|no\!atl)Li|to|
1erentorganicbulLdel'iliesa|r]n|ois|ul.rnlljnl)rguni(.ontentS.0rgaDic
]].om90fL80kgDr.'\Ioistureanrliiorglljlcon1.n|s$l.r..erpressedasnasi|atiosre]atilettr
lhc|ungco1.nlstoinedrInolclerllgtonrbustnll'\'ll)isl!f.r!ljos.
rate is irlependent oI rhe orgaric hull rlo,sitr. { !,rir,.rsalburninF rate. the unit area bunr ratc lt BRl. r rs obtuin(l br normaliz
i n g t h e b u r n i n g r a t e t o t h e ! f t ' a , ) l t h , l ) , , n , i , i 1 .s.i , r l a ( c . I t i s e x p r e s s e da s f o l l o r ; :
LBR = 0.2i
0.097 R.v -
o.olllllRr-Dl g rn']h'.
irhereD = R, if l,
Introduction
Srnoldcring.although not as risualh dranralic as
flarningcombustion.is an inlportantcomponentof
iblest lires. TL is cornnon in the duff ol the
coniierousi;rest thaLis deriveclfrorn thc accumulation of dctlitus. The dull inclrrdesthe fefmerrtation and hrrmrrs laver-sor thc O. and 0., soil
holizons that lie betleen the surface littel and
'lhe
the mineral soil (Bradv 198,1).
rate of
snloldering- mass loss orer tinre-rnd its dcpeD'
dence on olp(anicbulL clensitv.moisturc. and inorganic conterlt of the duli is the topit, of Lhis
rnanLls(,fipL.
Srnolderingis a lirrm of fife sl]read thal gencrallv octurs irr hrcl afravs thrt alc rnole tightll packed
than those that -quslainflarning. { lransilionfr_om
flaming to srnoltleringi-. erpected *hen frrel particles are Lhin or rerv small in diarnctcr and occup! morc than appro\inlateh lO Percentof Lhe
lolume. Rolhefmel(1972),.hols c dccreasing
rriiction intcnsitr. the rate ol he:rt release of
spleading lire per unit efea of the fcrreslfloor. that
e\trapolillcs lo vcr' lo\\, values r{hen the pacling
ratio ol lfacli(rnal \olumr occupicd b_r lirel np
plorcircs l0 percent.Fole-qtclLrlLthich Ias packing ratios gfealef lhan l0 pclcent. erihibits.mh
smolcJcling.lfhile litter. rrith pacldng ratios less
than 10 percent. e-.,hibits
onlv llarning.(A folm of
smolderingdoes oc lr-in larllc fuels independent
'll&rd
1t16
on Ph.D. di-.sertition {rf,lrilte.l t. tlr. t.n;\fbit',;
Uo,r
\orthr,r:st -Qcicnce.Vol. 65. \o. ,1. 1991
<
1.0. dtl
1l :
1.0 il R,>
l.l).
of packing cluling thc glowing phase in large fuel
pilc buming, but is not addle-qseclhere.i
Thc glountl fire. as the smolcleringfire is
named b! fife (r)nlfol spcciiilists.presentsno thrcat
. , , s . r - " . i t , . r np e r i r rt e r n r ,- ' l i t - r ' a l e , ,- 1p r e a r lH
sisl f.rr davs or $erks and latcr can initiate a Ilaming surfacc ilrc if it is fanned into llarnes bv a
chance gust of riind (Wein l9811. This potential
has blanded the gnrunrl firc m thc 'hold-over lire."
,Smolclering
duil temperaturesin e\.ess of 300oC
hare been obser,;edfcrr-upto 12 hour-sunder a
mired olerstorv ol rrester-nlarL:h lLurch occidenldllJ Nutt.l and lodgcpolcpine (Pinuscontotla\ar.
kttifolict Engelm.l (Hartford anri Frandscn, in
pl e-rs).
'Ihc
duif plesents an eflertire harrier to the
t r J n - l - . , , Il . r t t , ,t l r eI n i ' - " . r s1 u i l, l r r r i r rp3r - - a g "
of a surface fire. Hor,'eler. ii the duif is ignited.
the resullant srnoldcring lire is likelv to he brought
into tlirect contactrrith ihe rnineralsoil raisingits
lemperaLureaborc 300"C for severalhours.Flora
and fauna of the duff are con-rlLrnedalong with
r{)olsand see{ls.Organic matcrial in the upper p()rtion oi the mineralsoilis oxidizecl.and roots.seeds.
i r n J- o o " g c n i - r nr-r e , e - - a r tr i , " r . , r , l i n t sI t t i , I .
are killed and pc,ssiblvconsurned. Srnoldcring dufi
ran aiso contribute more than ;0 peraenLof the
air pollutants produced iiom a iire (-sancllierg
-t98ll). II. on the other hand. thc moistule and inorgrnic contcnt of the dulI exceedsthe lirrritsfor
r ! r r l r r r - t, , n t l - ,n r l , , , l i l i - l i l . l ' t " r f l n a i r , u n burrred and nol acL as d sour-ccof hcat (l'randsen
I9B7). l,nder these conditions the smoldeling
e o n b u - l i n n 2 , , n . i . I ' r ' e u r r t r , ll r o m , u m i n g i n t o
dircct contacLwith the mineral -"oil.Ho$,ever.the
moisture limit may be overcorne.A florr of heaL
frorr an abLrnclance
ol surlace lirels {slash)bnrning at thc sur-facc can {lrv Lhe fuel ancl bring it
within the combustion limits.
rrnge of organic bulk densitiesvaried 1iom 90 to
180 kg rn 3. This correspondslo lhc range that
is norrnallr' irund in duII (['oodarcl and Nlartin
are related to
-[980: Harrington l986l. ResLrlt-q
slnolderingin the abscncc of rtincl anrl slope.
-Smolcierirrg
can Le initiatedin the duff after a
lire spre:rdsthrough thc forcst littcl ii the duff is
capablcof sustainedsrnoldering(Frandsenl98i).
Larger litter iirels that sustdn burning. such as
tt'igs and concs. are igniLecland acl as cerltelsto
propagate the firc clorn into thc rluff. Orrce Lhefir'e
;s established. it plopapiates holizontallv rnd r.crticallv until it reachestrrnditionsthat \Lill noi suslain slnoldering.
The rnoistureand inorganic conlenLs\rere eric.r--.
p r , - - , d . r . m r - - r u l i ' Fr - h l i r ' I n l l r e , r ' e . r n m
The rnoistLrre
ratio. R,n. ran:icclfrorn 0 to 0.8 at
0.2 interrals and inorgarricraL;o.Rr. liom near'0
(nalrrralpeat inorganic ratio) to l rl intcrvals of
1 All conrbinatiorrsuere sampled e\cept those that
were outsidcthc snrolderirrglirnit giren bl Frandsen (19tt7) (Figule I). as
Ru + (R14,)< ] t
lf heat from the srnoJdoingcornbusL;onwa\e
is noLsufficientto olercolne the heat of vaporizatiorr requirecl b1'moist ftrcl. snroliering rrust cease.
FurLhermole. inolganic materials within thc fuel
rnalrix can absorb heat bul not o\idize to produce
rnore heat. fhus. thc arnounl of hcal produced peruniL rolurne is reduced. Both rnoi-.tureand in.rrganic content should Ieducc thc cflcctilcness ol
thc alailable heat b propagate the smoldering fire.
Srnoldering is erpected to procccclrdth the greatest
rate rlhen these inhibitinpl variables havc thcir
minirnum valucs,
The lactors that mav alTcctthc ratc of burning
ar.cthc lnoistureand inorganiccontent ol the dufl
ancl its organic buik den-ritl. Both the moistulc and
the inorganic contcnt hindcr the smoldering processand thefeioreare lierred as slowingdo$n thc
mav
rate of blu-ning. lno-casingthe bulk densit-,-slowdoln the supplv ofoxrgcn lo lhe cornbustion
interface and also slor{ the late of bul-ning.
Honcvcr'. Ohlcmiller-eL
al. \1979) found with flerible poh urethane-"that "the smolder processis Ilexible enough to adapt to rr $iclc r[ngc of oxvgen
low leveLs.'
supph lerels. inclLrdingextraordin:rrilv
t h e l ; l J o r . i . rfpr o , c J ur , u , r l l ; r ,- J l r \ t r F r i r n F ntli '
' n F a - u r et l p i l c f c r d " r r . . " i t l r , r n r - - ' u r . , r n r l t i u n
ratc of srroldering peat on ihe abo|e I;lctols.
(1)
a
t
t
t
a
t
a
a
t
-
t
t
I
F
IJJ
g 0.4
a
I
a
t
a
t
t
^ q ^
(J
-
--t
tl.,
D
I
F
L
o
=
a
a
t
t
o
_
a
t
I
I
o t {
,-',1
tt,). a
a
Methods
Canadian sphagnunrpeal rnoss\\as seleded as a
repfesentatire fuel ioI eramining the Irte of
sirnulatedfield
smoldcringin the lalioratorl Lrncler
conditions. lt has similar palticle sizes and bulk
" n d p a r l i c l e d e r r - i t i e -r - , 1 r r h .i - . r n i n p " r t r n t
ground coref in the tet boreal forestthat supports
smoldering combu-ition (Vein l9B3). and is a
uniform fuei that is cornmcrci:rllvarailable. The
0.o
INORGANIC
RATIO
Figure l
Samphrg design. Each circle designates a sunplnrg
cornbiDation of moisrue and inorganic ratios {or each
o , p . r , i . . o r r , r r . r ..oFr" l '
nl.rrti.r "'p
h-ated three tnnes. Ferr obserations irere attempted
berond the erpected smolde ng limit (Br.
B . r n i r g R r t - , , i S m , i ' l . " i n gP - a t
lo7
Peat r,as moistened or clried to leach the
sclectedmoislure ratios.Inorganic mateial in the
lbrm of 22tr0ncsh (56 4 partide sizc) powdered
silica $as added to moist pcat to obtain inorganic
r r t i o s h i g h e rt h . r r rt h - i n h " r e n t i n o r S a n i i, r , r l . n l
ofpeat. -Silicalas uscd to ensure that thcrc lould
be thc lcast arnollnt of chcmical interaction betrreen the addcd inorganir' material dnd the combustion process.Tt is speculatedthat natural in
organic rniltcrialocculring in pcat rnav react in a
mannerto alter lhe combustionprocess.llorc dctail on sarnple prcparation can be founcl in l'rands e n( 1 9 8 9 ) .
The cxperimental sample rr.ascontained tilhin
a noncombLlsliblcinsulated box with an open top.
The box tas placed in a sealcclglass c,vlincler.The
cvlinder }as scaled except for an inlct and exit to
allrl for a florr of air. The flol las held constant
at 3.5 g min-r (rrear3.5 | min 1). an arnountthat
t'as determinrd to be suificient to maintain
- m " l , l , r i n g r ' i t h n u t, , r r c , n { t u ' \ l l t i . n . . r - ' \ p ' r i enced in the field. The sample volurnc lithin the
in-rrlrtedcomr,rr.tiunI,u',s.r.5 cm r i on in rrr,'.sect;onand,1 cm deep. The walLst'ere 2.5 crn
thjck. The insulatingmatcrial r,;asceramir.board
t'ith therrnal properlie-r similar Lo peat (lrandsen
l9fl7). Conscquentl:-. the heal transierred to the
wails wa-s not greally dilTerent iiom the hcat
transferrcdlithin the peat. This allolcd the burning peat to approximate part oI an expanded
\ olume. Thcle r{as no e:tcessivchcat loss or cooling at the u'allsof the container.as eridenccd bv
th-.rlrslnn
' o l r r r r L u n r Fl J, r J t J J j r r ' . n tt o t h . r , . J l - .
The upper cxposedsuriice of the pcdt n'as if(,
nited bv ftrur rol,".sof lesistance coils heated to a
red glorv. The gioling area approximated thc area
of the pcat slrrlace and was held t'ithin 5 nrn of
l h e . r r r f . r , . l , , r 2 m i n u t e - .P r i , , r- r l , e r i m n n t n t i o n
i n d i ,J l , d l h J l 2 m i n L r t . \-\ a - - u f l l ,i p n tl n i e n i t cI ' c a l
at all comhinationsup to the highestmoisturcand
inorganic ratios. lhe arnount of organic material
consumedunder enhancedburning during the ignition periocl is srnall cornpared to thc total mas-r
consumed.(ionseqLrerrtlv.
thcre is little error in thc
total burn time ofthe samplc that ranged iiom B0
to 25i rninutes.
The ignitel coil,<rlere liltcd ar\,avliom the pedt
sur{acewithin l5 sccondsof s$,itchingofi thc igniter'. The dur-ation of srnoldering rras taken from
th(] time the igniter !,,ass$itched off until -qmoldering stopped as evidenced bv the abscnceof carbon monoride in thc air llot'ing fiom the cvlinder
168
Frandsen
containing the smoldering peat. Carlron monoxide
las considered absent ll.hen the volurnerrrc con. e r , r r r r i L r $n r . l e - - r h a n ( r . 0 0 0 5 \ \ .
Results
Datr lere obtained lrom J72 sample ohsen'ations.
This inclucles aclditions Lo the original design for
cornparison sarnpJingsand losscsdue to those conclitionsthat rrould not sustaincombustion.
'l
he average srnoldcring burn rate lvas
calculatetJ by dividing the organic rnass (sample
rnass inorganic rnass)bv the duration olsmoldering. lhe organic mass l,as cquivalent to the total
mass lost because al1of thc organic mass was consurned during smoldeling.
'lhere
is a general trcnd torrard lorler burn
ratcs lrith iDcreasinginorganic ratios above 1.0 fcrr
e a ,h r n o i - t u r 'r-: r t i or l i g r r e 2 r . B . r r nr a t , r e r i . r l i o n l ithin replictrtionsof the organic bulk densitl' rvere
as greatol-lileaterthan !arialions bctweetrorganic
brrlk densities.Hcnce. there is no dependenceon
the organic bulk densit,v.Because of this independence.one mav viern-each combinationof the
rnoisture and inorganic ratios as replicated by as
much as 12 times if all burns l.ere successful.
Tt'o commerciai rarietics of peat Iere used.
The ruiginal choice was used for all p ot expenmcnl- dnl iritirl oh-elrrti"n- Jr m(illJrc r.rtios
of 0 and 0.2. S'hen additional peat wa,<necded.
no more ol lhaLrarictv was alailable. A rcplacemcnt \ras used at rnoisturcratios of 0.4 through
0.8. Thc search lbr the replacemcnt included those
peatsharing similar bulk phvsical pr-opeticsand
thc lotest inorganic contcnt apptoaching that oi
the original pcat. Con4)arisonsof the burn rates
sho$,edthe l\\o pcats to be equi,,alentat 0 moisture
ratio. but the burn rate ofthe originalpcat tlas 20
percent grealer at thc lar€iel rnoisture ratios (0.,1
to 0.8). ObsenaLions$'ith the replacementpeat
rere adjustcd up\rard bl- 20 perccnt so that they
could be rrsedto complementthe oliginal obserlations (Frandser I989).
Noting thal lhc burn rates al inorganic ratios
n e J r f t a r ' l r t I - h o r ' L i t t l ed i l h r F r , . J t m o i - t u r e
ratios 0 throrrgh0.4 (Figure 2). *e can linearize
the burn rate bv cornbining thc burn rates at I *'ith
those neal zero and gir'ing the combined inorgani<:
ratio a dummv valuc of 0 and reducing all the
highcr inorganic ratios bv l. (This is not the case
for mo;sturc ratios 0.6 ancl 0.8 but the data at
moislure ratio 0.6 account for onlv 14 percent of
the data and onlv 2 percent at 0.8 moistureratio.)
Moisture
Rotio
I
_c
tr
T
c)
Y
t
M
tr
T
v
1
^v -- \rJ\. \ J r l
ot -- lr' \
C
t :O.2
o:0.4
o :0.8
. u
A
T
T
A
I
o
1
T
l.
A
v
c
t:l
rn
z
0
J
lnnrrtnnin
!igrre
5
Prr{in
2. tsunr rut. dr:pc,!l, n(r'on lhc i,n)rgoni,i ratb at moisture ratios of0 to tl.il. Note the decline ofthe burn rate irith increasirg niorga . ratn hegh.iig ar 1 Eror bars rilaL. t,) th{,sknda
d { ' v i d r i o n .n : 6 t o
1 2 o b s e r l a t i o n s .S i n c e o b s e r y a
tions hare no error hars elcepl for moisture ratio 0.,1 at inorgaDic rrtjos o10.09 ard 1 {hcre Lhr: rrror is lcss than th.
s r m b o l s i z e .{ l g h ' = 2 . i 8 a l 0 ' k g s ' r
Thc modi{icd data hale the appearance of being linear in both the inorganic and rnorsturerallos.
A lineal mociel derelopecl b_r appl_-ving
a lcast
squar-esfit to Lhe modified data is shol,n bclor'.
It also cmbraccs thc shift in the inorganic axis
disursscd abovc.
B R : 7 . 5 8 - 2 . 7 2 R r / 0 . 9 1 8 ( R J- D l s h '
(l s h' = 2.78 r 10-' kg -"-r) \2)
lhere BR : bLLlnrale
andD: R' ifRr < ].0
D:
1 . 0i f R I > 1 . 0
The burn r-ttc \.clsus moistlrre ratio is shor'n
in I'igure 3. along tith the linearized rnodel- The
percentage error (predicted- ar'Lual)/acLual.
is
.Lrlsrr ur-r th- r.rns, ol inurtrril rrlio- Jt n
rnoistureratio oi 0 in Figurc 4. Estirnatcsof the
burn rate frorn the linear noclcl indicatc that 91,
89. and 90 percent of the prcdictcd lalues lie
rvithin t 10 percent of the actual valuc for'
rroisture r-atiosof 0, 0.2, and 0.i[. Err-ors ale
greater at moisturc ratios o{ 0.6 and 0.8. Onlr
about 70 percent ol the predicted values !,,ere
within + l0 percent ol the actual valLres.The lower
predictive capabiJitv is not surprising considering
thc dcputurc of these data from the form of the
data at the lol'er-moislureratiosin Figlre 2. Furthennore. these data. as mentioned above. accounl
for onh 14 and 2 pcrccnt of all the obserrations.
Discussion
\lany people think in terms of the linear spread
latc of smoldering t'hen reference is made to burn
fale. rather than to the milssloss burn rate presented
here. S'e gain insight inb the -(preadrate ;f1r'econvert thc burn ratc to a unit area burn rate by normalizing thc burn late to the area of the burning
srrrlaceof thc smolderingl avc. Thc resultingunit
area burn fate rclales to lhc smoldedn€t$ave and
is independent of the combustion bc,^.
'fhe
cross-sectionalarea. A. of the burning
volume of the combusL;onbor (Figurc 5) approximates the burning surface a-qit mores dolvn*ard.
Thc lincar sprcad rate can be obtained by equating
the mass loss burn ratc, Am/At. to the product ol
the organic bulk dcnsit_-v.p,. and the volume
Burning Rate of Smoldering Peat
169
I n o r g o n i cR o t i o
o:0.04 -0.09
v =3
o:4
a :1
n:2
I
_c
a)
t
c
lf
0.0
0.2
0.4
0.6
0.8
1.0
Moisture Rqtio
Figurc 3- l.in($ least squaFs lit to olscnatioD-r. Enor burs rclare ro the stardarrl rlcriation. | = 6 Lo l2 observations. Silglr
ohserrlri,nrshalenoerrorLars.\ot.singlelinear{lrro;n.reanicrariosfron0.0,1tol itIh'=2.78rl0'k8s'l
25
20
M o i s t u r eR o t i o - O
t 3
t-
L!
c
o
10
5
0
o
0_ - 1 0
- t 3
-24
0.0
1.0
2.O
J.0
I n o r g o n i cR c t i o
figure-1. lnors
t 70
i,r rnalietiDg rhe lrrn
Frandsen
ral,. \
4.0
5.0
rndD=&iIRr<1.0
I ) = 1 . 0i i R , > 1 . 0
-t.
I BR rangesflorn alroLrt0.15 to 0.i30 g crn-']h
L.singa cornmonr oluc for thc olganic bulk dcnsitv. 0.1 g r,nr-r in Eq. (1). rre arrire at a range of
1.:) kr ll.0 .nr hrthat is in the right c,r'derc,1
rnagnitudefor smoldering i\\ ein 1983).
Figure 5. TLe bunnig roLLunerithin tlie cornbustioD bor sho\r
rrg the burning surface. ,\. as it nores rtorrnnarrl
a r l i s t r n r r ' . J r . s ' r r 1 , i n g r ) , , ta \ o l u t r t r 'r l f r I n t . l l
c.m-eumptiorlrate. Alv/At. as s\rept out b! thc
btrrning sLrliacein Figrrre 5.
lm/41
:
Q,AA!/-\I
g h-'
{3)
Transposing.1\'c halc lhc lincar ratc ol spr-ead
(4)
-\rilt : (lArn/lt)/{)iq.
or h 1
rrirele Arn/At is the burn latc. Blt
antl (\n/At)/-A. is the ulit nca blun latc . LBR
Although Lhe bor tas riesigneclltrr a cro-.sscr,tionalar-eaof 25 t mr. the adual areaincreased
slightll iiorn dete oraiicrr o1 the rr'alls lith cach
burn. l'hree bores rrcrc uscd during thc cxpcriment to enslrr-cthat thc alcr rlicl not dcparl grcalll lrom thc designcd lrr-caof 25 cmr. Thc arcl
renged frdn 26 to 30 cnl'. lhe meclian.28 crn'.
rras used to lppro\imatc thc ar-caof thc bulning
surlace. Relinements in cstinuting thc arca ar-cpl-cclLrdcLlbl unceftainlies irr the shape of Lhe srrr'face.The LrrriL
area burn rate is a uni\ersal l)ararneter Lhatrelatesonl! to the hurning surlaceoi the
srrolderingrrarc arrclis e\l)ressedrs lhe lnassl(rss
r a l e p e f u n i L a r e a o f L h eb L r r n i n gs u r ' { r c e .
LBR : BR/28
R e p l a c i n gB R r r i t h e q u a t i o n( 2 )
t t s R = 0 . 2 i - 0 . 0 9 ?l t . , - 0 . 0 3 3 t t-, l l l
gcm
n
ll
( 1g c m ' h ' = 2 . 7 8 r1 0 ' h g mz s ' )
lhele [-BR = unit areaburn rate
Note that in llq. 1.1)thc burn r-utc is rclatcd
directll to thc lincar splcad lalc through thc
If the densiLr
organicbulk riensiLri.{ is r.orrstant).
is halvecl. the linear spread rate rnust be doublecl
to maintairl thc sanrc nrass loss ratc. This is
feasonable$hen rorr consider thal a combuslion
Jront must pas-sthrough a le-rsdense luel alrav
fastef lhan through a nloLe dense arr:rv to consume
nas-r at the -ranlelate. lt should be clear that the
line:rr spread rate rnlv be thought of as r'olumetic
burn ralc as the burning surfir'e srleelrsoLrtvolunle. but it cannot be thought of in the snrne sense
as the mass burn ratc.
I'lobable errol arralvsiswasernplovedto examine Lhe rariabilitv of the independent rariables
(Flandsen t9B9). Tladitional -statistical
methods are
not appJicable $hen the sample is prepared to
varachicvcprcsclcclcrilalues of thc irrdepenclent
iablesas designedin FigLrre-t. Basedon this analr . i - . t h r m o . t r u ' r . r l i uh . r .. r n , n u r l l r . r il . u E r ,J l , f
than 3.2 percent.Lheerror in the organicbulk densih i,cno gfeaLerlhan 3.5 percent.and the inorganic
ratio has an error no grealer than 2.6 percent.
Afpro)iimatelJ 90 percent oi the predictions
frorn the lincar modcl alc rrithin t 10 pclcent oi
Ihc acttralralLrcat moisLureraLiosof 0. 0.2. and
0.4. \ealll all elrors grertel than il0 percent
are rt the highrr inorganic r-tlios. Thc prcdictivc
" . r l ' . r l , i l i, tl 'r ' , t , t. , , . r r r' ' n , l , t l , r r r ,' i l J l I n u i - l J .
r a t i o so f 0 . 6 a n d 0 . 8 . T h e l ) f e p o n d e f a n coef d a t a
afe at the lo\ref inorganic ratios.at and belou l.
This is a rellection ol lirnitecl sust:rined smolder'ing and uncertaintics lhcn appr-oachingthe coml , . r - t ,i r rl i r r ' . rt F | r r r , l - - r lrq H l ' . r l I n ' ,- l , r r cf i r l ' at and rbor e 0.,1. lt is likelv that rrhltever order
liness there mav lr in the relatiolship lrreaks dotn
as that boundlu'r is rpploachcd.
Thc rcsultsof thi,.stutJl clcfincthc clepcndcncc
of Lherate ol burningon lhe nmistureand inorganic
content oi smoldeling cluli'lil<e materiaLand -shou'
a lack ol dcpcnricnccon thc organil bulk dcnsitv.
\ - i r r ' 1-, r r , ' l - l , r 1 , - . l i ,' '- t l r ,- ' , r . n r at,- r ' . t : i- - ,
that infelences can be made about the rate oi
'fhese
snlolcleing.
results contlibute to a better
urlclerstrrrlcling
ofthe combustionof srnolderingdufi.
BulningRate of Smolcleling
Peat
171
Acknowledgements
The author tishcs to thank Plof. Donald F. Potts
ofthc SchoolofForestn, at the Lnivelsit,vofl\{ontana lbr his encouragementLo pursuc this line o1
iDvestigation.
LiteratureCited
Bfrd\. \. C. 19111T
. h e \ e r L u e a r d I r f t r l r i { s , ) l , r o i l .g r h e d . .
Thc \lacnillan Conpam. \er Iorl. ;50 r).
l r u J s t n . \ 1 . H . ! 9 i J ; . T h e n l u e n c e o f m o i s t L u ca n u d n { i r l l
s o j l o r t h . r o r r b r s t n , n l i n i t s o l s m o l d e r n i gf o r e s t d u f t
Car. J. For. Iles. l;:15,10-1511.
F r a n d s e n . t r . H . 1 9 8 9 . - r r $ l d . r i , , s ( 1 , , 1 1l i:m i l s . h e r t e r o l r e d .
and burn rate. \Iissoula-l\lontrnr. t,nirersitr of \Ion
tana. Ph.D. liissertatidi. 1'16 I)HarrnrSon. \{. tll. 1986. C,r!r)dris,)n ol lorcn lloor deprh to
loadlng Lelationshiq,s nonr *r.rrl
pon,l.rroso pine
s t a n d s .R e s e a r c hN o t e R \ ' l . 1 6 3 . F o r t a o l l i n s - C 0 : L . S .
D.rrn m f nr of r\sriculnn e. Forest SeN ice. H o.[\ N ou nLdiii!(,,, st ud Rrng. Iirperiment Station.5 p.
Hatford. Il. {.. urrl \\r. H. Irra.dsen. In press. Cround
lernt)eratLre liistori{\ l;otri 1;r.: An ilhtstration.I'lissor a.
\lT: l-.S. liefartrnent,)J Asri(! lu(. forest Serrice.
Liternrurruin
R(is.u.(ih Stuion. 1t1 p. Editorial draft.
ohlemiller. T. J.. J. Bellc,,. ur
r''. Rogers. 1979. -{ model
of smoldering ronrbustion applled to flexible
polrurethare loads. Q,mb. Flame. 36:197 215.
I t-2
l-randsen
Rothermel, R. (. 1972. .{ mlthenarr al n,odcl lbr predicting
poper INT t15.
lire spread in i,itdtand fir*. R|silrh
Ogden. LT: U.S. Depatmert ol {sri(! tur. Forest i\er
rice. Intelnountain Forest ard Rangc lirpcriment Sta
S a n d b e r g .D a r n l \ . l 9 8 i l . R . s { ! r c h l e a d s r o l e ! ! s m o L e r r o n
tresdibed lir.s. ln: PrrlrdnHs. 1983 \othNest Fire
Council {rnua1\leeLnrg 1983 \oonber
2l 22i Olvn-
pia.tr'A. 2i 1,.
trein.R. tr. 1981.Chrr!.r.filrics
andsuppre-.;ion
of fires
ii orsanic rerhiii h A,,strali!. Alsr..l. fof. ,1,1:162 169.
Sein. R. X. 1983. tjrc brhN\ior and rological effectsni
orsanic rerrain. 1a -qCOI'E 1B: 1hc Role o{ Fire in
\,,frhol Cncunpolar Ecollsten*. R. \Itin. and D.
A . \ ' l n ( j l ( j { n -( a l s . J o h n \ t i l e r & S o n s . \ e r \ o r [ . p .
8t 95.
\! oodard. P.. ard R. E. Nllrrin. I q80. Duffweight and depth
in a high eloation Pirir
J. For. Res. l0:?-9.
rontorra Dougl. forest. Can-