radioactive chains w ith half- li ves on the

order of at least 1 billion yea rs. Elements

w ith shorter half-li ves have long di sap-
peared, and the new rad ioactivity created
by cosmic bombardment is minimal. More
than half the radiogenic hea t comes from
Significant Heat-Generating Isotopes
potassium-40, which occu rs primari ly in the
mantl e, along w ith the radioactive seri es Isotope Half-life location
originated by thorium-232 (see "Signi ficant
Potassium-40 1.3 billion years Mantle
Heat-Generating Isotopes," right). Uranium,
the only other significant radioelement, is Thorium-232 14 billion years Mantle
found mainly in th e cru st. Although a few Uranium-238 Crust
4.5 billion years
add iti ona l ve r y l ong -li ved isotopes
exist-for example, rubidium-87, lutecium-
176 and rhenium-187- their concentrations
and heat generati on are insignifica nt.
Since the core is assumed to contain no Minerals
• Quartz
heat sou rces, it makes no contribution to
heat flow at th e Earth's surface. The fraction
co ntributed by rad i oac ti v it y i s rising ,
""-· "' Pyroxenes and " Olivines, Garnet,
Epidote
although the absolute rate of heat genera-
tion by radioactivity halves about eve ry bil-
~Feldspars Mtcas Amphiboles

lion years. Th e current rate of rad iogen ic • Water

• Ice
hea t flow is about 5 perce nt of its va l ue
when the Earth was formed. Rocks
/'Rock Salt
Temperature readings at many points at or ------
near the Earth 's surface yie ld an average - - - - - - - - - - - - - Limestone ""sandstone & Quartz
heat flux of abou t 1.2 X 10' 6 ca l ori es per
square cent im eter (em) per second, - - - - - - - - Schist '-.....Granite
although indi vidual values may vary by up - - - - - - Gabbro
- - - - - Tuff
to one order of magnitude. Hea t flux from
the sun is much larger, but makes no sign ifi-
0 0.05 0.1 0.1 5 0.2
ca nt contribution to the geothermal grad i-
Thermal Conductivity, k cal;cm s. oc
ent- most solar radiation is either refl ected
or rad iated back into space. Furthermore, 0 Ranges of thermal conductivity for minerals and rocks at standard temperature and pressure. The
diurnal temperature variations penetrate the range of thermal conductivity of silicates (feldspars, micas, and pyroxenes), for example, is controlled
by silicon-oxygen bonding: the more shared oxygen atoms, the higher the conductivity.
surface only about 1 meter [3.3 feet]; annu-
al vari ations penetrate a few tens of meters
on the continents and a few hundred meters and VT is temperature grad ient. Since co n- compos ition and porosity (above). Water,
in the oceans. Climati c variations, however, ductivity and temperature gradient can vary w hich responds on ly sluggishl y to th ermal
at the scale of a few thousand years, can individually with depth, each must be speci- changes in its envi ronmen t, has a therma l
influence measu rements hundreds of meters fied for a given location. Near the surface of d iffusivity an order of magnitude less th an
below the surface, and variations of 10,000- the Earth, the gradient is dT/dz (z increasing that of typica l rocks and mineral s.
year period ici ty ca n be sensed down to 2 w ith depth). Thus, the rate at w hich heat is Although there is no consensus on a mix-
kilometers [1 .2 milesJ.2 co nd ucted ou tward through th e Eart h ing rule for thermal conductivity, a reason-
depends in a simple way on how tempera-
How Heat Reaches the Surface ture and thermal co nductivity vary w ith
1. Howell BF )r: Introduction to Geophysics. New York:
Conduc ti on and convection are th e two depth. However, to pred ict th e variat ion of McGraw-Hill Book Co .• Inc., 1959.
common modes of heat tran sfer in the Earth. temperature w ith depth requ ires knowing Anderson RN: Marine Geology. New York: John
Conduct i on a ri ses from in t e ra ct i ons th e ve rtica l d ist ribu ti o n of heat sources Wiley & Sons, 1986.
2. Garland GD: lntrodvcliontn GPnphy.< ics-Mantle,
between neighboring atom s, w hereas con- (rad ioact iv it y) and therm al diffusivity-a Core, and Crust. Philadelphia: W.IJ. Saunders Co.,
vection transports heat by fluid movement mea surement of the r elat ive ability of a 1971.
mate ri al to co ndu c t and to retai n heat, 3. Hearst )Rand Nelson PH : Well Logging for Physical
resulting from the interaction between gravi-
Properties. New York: McGraw-H ill Book Co., 1985.
ty and thermally induced density cha nges. k/pC. in wh ich p is density of the rock- fluid Horai K-1: "Thermal Conductivity of Rock-Forming
First, consider cond ucti on. sys tem and C i s heat capac ity per unit Minerals," journal of Geophysical Research 76, no. 5
In an isotropic medium, conducti ve hea t ma ss.J Because th e verti ca l distribution of (February 10, 19711: 1278 - 1308.

flow is q = k VT, w here q is the heat flu x heat sources i s not routinely determined,
(calories per square cen timeter per second), most emphasis is given to determini ng ther-
k is th erm al conduc ti vity of th e medium, mal diffusivity.
The wide range of thermal diffusivities in
rocks and minerals results from va riation in

Volume 1 Number 1
55
 Descending Descending
Convection Currents Convection Currents

0 The relation bet ween convection and plate tectonic movement, as proposed by A rthur Holmes in 1929 (vertical scale exaggerated). Rising con vection
currents pull continents apart and create new ocean in a rift zone. M o untain ranges and deep sea trenches form where convection currents descend.
(A fter Press F and Siever R: Earth. San Francisco: WH. Freeman and Co., / 978: 459.)

ably succ ess fu l empiri ca l approach is to ri ed by the rising current is Cvt!.T, i n w hich 120 . ------------------------ .
take the volume-weighted geometric mea n C is the fluid's heat capac ity (calori es per
of the conductivities of the constituents: cubic centimeter per degree), vi s its vertica l
velocity (centimeters per second), and LiT is
kmix =D i<f· the temperatu re difference between the top
and bottom of the convecti on cell.
in which kmix is the conductivity of the mix- The mantle itself is thou ght to be suffi-
ture, and k; and ¢1; are the conductivity and ciently fluid to parti c ipate in convec tio n . u..
0
vo lume fra cti o n, respec tively, of th e i-th For example, a rock flow rate as low as 1
co nstituent. This rule applies to both non- em 10.4 inch! per year across a temperature ~
:::J
porous and liquid-filled porous rock.4 d rop of several tens of degrees produces a ~
Conductio n was in itia lly thought to be heat fl ux of about 10'6 ca lories per square
Ql
Q_
•
E
responsible for most heat flow in the Earth, ce ntimeter per second, th e same order of Ql 105
f-
but studies show that convection also trans- magnitude as the total o utward fl ux at the Ql

ports heat efficiently and often plays a mo re surface.2 Not o nly does this con vection in 0
.r:.
important role than conduction. In relatively the mantle influence heat fl ow to the crust, E
0
uncon strained systems, convection occurs but large-scale convection cell s d rive p late g 100
CD
by means o f convection cells, in w hi ch a tectonic movement (above).
rising fl uid carries heat upward and transfers Hydrothermal convection takes p lace at
it to cooler material at the top of the ce ll. As lower temperatures, where w ater can exist
the fluid ri ses, it cools, becomes more dense as a separat e pha se . Hi gh -temp erature
95
and sinks to the botto m, where it is heated hydroth ermal convection is responsible for
again . In some environments the fluid may loca l heat transfer to th e Earth's surface in
not find a return path to the bottom. In this the form o f hot spri ngs and geysers. Low-
case th e convecti on fluid must be rep len- temperature co nvecti on is probabl y an 90 ~--------------------~
ished from another source, for example, impo rtant mechanism of fl uid (inc lud ing 1 2 3
wa ter entry from the surface through an hydrocarbon) flow in sedimentary basins. tk + tlt
aquifer. In either case, the heat flux, q, car- M
Measurement of H eat Flow Rate O Bottomhole temperature recordings in a Paris
Direct measu rement of heat flow rates is Basin well from fo ur comecutive logging rum
difficult to achieve, even in the subsurface made up to 18 hours since mud circulation
portion of th e Earth immediately avai labl e ceased. The measurements are extrapolated on a
Homer-type plot to yield true forma tion tempera-
to us. Co nseque ntl y, heat fl ow is al most
ture. tk is the length of time the mud was circulat-
ed; 6 t is the elapsed time between cessation of
circulation and a given logging run . The extrapo-
lated value of 116°F was confirmed by subse-
quent continuous temperature logging. (After
Dowdle and Cobb, reference 5.)

56 Oilfield Review

J
alw ays determ ined by independent in situ
m easu rement o f the geo th erm al gradien t Temperature
and of the thermal conductivity of rock sam- 150 OF 250 Temperature
OF
p les in the laboratory. Very recently, howev- Tension .c 150 160
er, a w ireline method for thermal conductiv- · o lbs 1,000
a.
<ll
ity determination was successfull y tested in ------ ------------ 0
several we ll s.4 Casing Collar
Locator Temperature Gradient
Th e geoth erm al grad ient is most c om - - 0.5 0.5
---------------- ----------------- ----
monly acquired by w irel ine measurement of
tempera ture in boreho les, alt hough mea-
surements have been made i n ocean sed i-
ments by implan ti ng a spear w ith vertically
,_
spaced temperature sensors (high heat flow
in ocean sediments correlates w ith hyd ro-
ca rbon deposits). M ost w ells are logged and l= i
1- '
11-: I=
the logging tool invariably records a temper- ; h - t-1..-.::-J
I= ·. ~- t--;:_;_
ature maxim um, usuall y bottomho le tem- t-H::.
!:
Jd :: ti- : -~=; = =
perature, that depends on the elapsed ti me
since m ud ci rcu latio n ceased. A Horn er-
HL~-1- ~ ~nf
type plot has been found empirically to con-
vert temperatu res recorded on consecutive
- I= ~ 1- 8-t-= 1
=
l · - - --
1-1-1-1-·

8 i~ -L ~:
logging runs to the true fo rmatio n tempera- - L-1-•
~- CO
~ l
tu re (below left ).5
1--Ef : ~
0

! ·: 1-
·~ ~
Bottomhole readings, unfo rtunately, are 1- ; ~
1- : c-~
r-::t= _. . l . ~~
not continuou s measurements. At best they
yi eld onl y a grad ient average in a sma ll
area, w hen one bottom hole measurement is
1-
combi ned with another at a different depth
in a nearby w ell .
H
Co ntinu o u s l o gs o f t emper atur e a re
recorded w ith platin um resistance th er-
:=== 't, 1-l
mometers or therm istors, usuall y located at I-!==
the bottom of the logging string to min imize 1-
I=
infl uenc e of the sonde body o n the mea-
su r e ment. I f th e dri ll i n g mud h as not 8C\J
reached ther mal equi librium w ith th e fo r- 1=11="~ f- co
1- ; ! ·-
m ati o ns, cor rec ti o ns mu st b e app lied at
each depth . A con ti nuous temperature log
pro vides T(z), from which the local gradient
dT/ dz may be determin ed directly {right;
next page, left, top and bottom ).
Sometimes the gradient at a parti cular D Use of a temperature log to identify the source of water production in
depth in a given w ell does not represent the cased hole, indicated by heating anomalies. Based on production in an off-
regional geo thermal gradient. Local depar- set well, and a sonic log, the operator perforated this well from 8118 to
8133 feet. But the well produced water and the shut-in pressure was consid-
tures can result fro m stratigraphi c conduc- erably higher than expected for a pressure-depleted gas sand. The tempera-
ti v ity c hange s, water flow {convecti ve or ture log shows that very little flow is coming from the perforations, and most
otherw ise), expand ing (coo ling) gas enteri ng is from an interval belo w, around 8136 feet. The heating anomaly from
th e ho le, c ement top s not yet i n th ermal 8 176 feet upward indicates that water production is coming from at least
that deep and channeling up between the formation and casing. The final
equilibrium, and recent fluid injection.J
diagnosis: the perforation job ruptured the casing, producing a vertical gash
Rock thermal conducti vity is determined that permitted water to migrate upward, possibly robbing production from
preferab ly w ith the so-ca lled d ivided-bar the perforated zone.
method.6 A disk cut from a core is held firm-
ly between two materials o f the sa me di a-

4. Dove RE and W illiams C: " Thermal Conductivity Esti-

mated From Elemental Concen tration Logs," Nuclear
Geophysics 3, 1989 (in press).
5. Dowdle WL and Cobb WM: "Static Formation Tem-
perature from Well Logs-An Empirical Method," jour-
nal of Petroleum Technology, 28, no. 35 (November
1975): 98-10 5.
6. Costa in JK and W right PM: "Heat Flow and Precision
Temperature Measurements in Boreholes," Trans-
actions of the SPWLA lOth Annual Logging Sympo-
sium, Houston, May 25-28, 1969, paper ).

Vo lume 1 Number 1
57