EVALUATING CANDIDATE LOST CIRCULATIOEI HATEREUS
FOR GEOTHERMAL DRIUIUG
Glen Loeppke
Sandia National Laboratories
Division 6241
Albuquerque, New Mexico 87185
ABSTRACT
Sandia Potion 1 Laboratorie Geothermal
Technology Development Division is working to
advance the state of the art of lost
circulation prevention and control. For this
purpose, a large-scale Lost Circulation Test
Facility was designed and built. This paper
addresses the evaluation of candidate lost
circulation materials using this facility and
also using the reconnuended practice of API UP
131. Test results from these facilities are
compared and discussed for the materials tested.
1yTELowcT10N
The problems associated with lost
circulation in the well drilling industry are
well known. Uuch has been written and many
solutions have been applied, but lost
circulation remains as one of the primary
problems in drilling a geothe-1 well (Pye,
Cukey 1985). Conventional materials and
techniques used in oil and gas drilling have
been tried in geothermal wells but have been
mostly unsuccessful because of the hostile
geothenual wellbore environment where the
temperature can far exceed the tolerance level
of conmonly used cellulosic materials (Coodarsn,
1981). A Lost Circulation Technology workshop
held in October 1984 pointed to the need for
research and development to provide lost
circulation behavior and control theory and, to
provide improved testing standards for
ooktials evaluation and devetopment (Caskey.
1984). The workshop report includes a biblio-
graphy on lost circulation.
Sandia Pational Laboratories, which manages
the U. S. Department of Energy's Ceothenual
Technology Development Program is actively
pursuing a program to advance the technology of
lost circulation prevention and control
(Caskey, Loeppke, and Satrape, 1985). Undar-
standing and modeling the behavior of fluid/-
Particle flows in the wellbore loss zone
address one facet of this work (Civler, 1985).
*This work was supported by the U. S. Dept. of
Energy at Sandia National Laboratories under
Contract DK-llco4-76DPOO789.
I
DE86 010999
This paper addresses another part of this work;
the evaluation of lost circulation materials
(LCl4.s) that appear to be good candidate6 for
geothermal applications. Tests were performed
both with a modified version of the M I 131
test cell and with a large scale facility
designed and built at Sandia that evaluates the
performance of LCWs in a simulated wellbore
environment (Loeppke and Coskey, 1983).
TESTING DESCRIBED
There are several types of loss zones h a r e
the loss of drilling fluid can be encountered
but those most cOmmOn in g e o t h e m l drilling
are wgular or fractured zones and fragile
underpressured fonnations where fractures can
be induced by the drilling operation. For this
reason, the tests described here were conducted
using a "slot"1 that represents the
fracture. This slot used in the test cell and
the cell are the same as that described in the
rocoarmended practice of MI 131 except that the
depth of the slot has been increased from 0.25
inches to 6.0 inches. This design is used for
all slot sires and is also used in the
large-scale Lost Circulation Test Facility
(LCTF). The purpose for the large facility can
be appreciated by upmining the differences
between it and the modified API test cell which
is shown in Figure 1. The LCTF is designed to
simulate the mud rheologies, flow patterns and
temperatures o f the wellbore envirormmnt
whereas the API test is essentially a static
test. The LCTF uees a 10 foot long section of
8 inch pipe with a smaller 4.5 inch pipe
(plugged on the ends) inside it. The "drilling
mud with LCH" is puntped through the annular
space between these pipes through a center
section where the slot described above is
mounted at the outer surface o f the annular
space. This test vessel io shown in Figure 2.
A seven barrel aud/Lcw. batch is circulated in a
closed loop through the test vessel at rates up
to 200 feet per minute. A heat exchanger can
be included in the pumping loop SO that the LQI
can be tested at temperatures up to 400.F. A
1 Slot (dimensions used in this work: The
"depth" represents the distance into a fracture
and the "sire" represento the fracture width.
ASTER
DISTRIBUTION OF THIS DUCUlENT IS UNUMITED
Loeppke
F i q u r e 1. ?lociified A n 1 131 Test Cell
test is conducted by imposing an increasing
pressure differential across the slot in much
the same way as the API test is perfonned. The
API test cell is not designed to test heated
materials nor does it allow crossflow circula-
tion over the face of the slot. Another
notable contrast that no doubt accounts for
most of the difference in the results that are
presented is the difference in the filtrate
capacities. The useable capacity of the API
test cell is about 3000 ml, vhereas the
capacity of the LCTF is almost 100 times as
much. If the same slot dimensions are used in
the LCTF. the opportunity for a plug to form is
greatly enhanced. The dynamic test and greater
filtrate capacity provide the rationale for the
more repeatable results attained with the
LCTF. Hinkebein et aL concluded after conduct-
ing experiments to isolate the cause of the
wide data scatter from the API test cell that
the probable cause was the plus formation
mechanics in the test cell (Hinkebein 1982).
By using the same slot design, data has been
gathered to show the effect of these different
test methods.
The objectives of these tests were:
1. To evaluate candidate L a ' s at elevated
temperatures,
2. To compare performance of L C W s tested in
the API test cell and in the large scale
facility. and
3. To experiment with combinations of
different Lcn types to determine if
perfomance could be improved.
A particle size distribution analysis was
performed on each material. This is essential
foe comparing the performance of different
La's. Since each material will test
i'i[*urc' 2 . Lost Circulation Test Facility
Test Vessel showinq a cut-awav
of the center section and t h e
simulated fracture.
differently according to its attributes (e.g.,
maximum particle size and size distribution) a
c o m n basis is needed to make an equitable
comparison of LCH's of different sire ranges.
This is accomplished by plotting the ratio of
the slot (fracture) size to the maxhum
particle size of a material. Test results can
then be compared o r plotted together for
different LCH's. The method used in this work
is to plot this ratio verses the material
concentration tested. The plot is a summary of
all tests performed for an LCI and shows what
concentrations can be expected to form a seal
in a given fracture size and hold a pressure
differential of 1000 psi. The maximum particle
size used in the ratio is fixed at 95% (95% of
the LCI is smaller than that size).
In order to minimize the effect of the mud
used in the tests, a standard formulation was
used for all tests. The formula was a 50%
bentonite/50% sepiolite water-based mud with
additives used to control viscosity and gel
strength for elevated temperature tests.
Haterials Tested
From a number of materials screened for
temperature stability, three were evaluated in
the tests discussed here. They were thermoset
rubber (processed from salvaged auto battery
cases), coal, and mineral fiber. The suppliers
(and trade names) for these materials are o h m
in Table 1.
'

_.
.Loeppke
TABLE 1. L M * s Svaluated Ground Coal
. Two brand names for the omma product
Lclr mBe pmulier Trade Yame(s1 (Kolite and Kol-Seal) were used for these tests
'Thermoset Westera Company Hi-Seal, now lhrd because of availability. There was a-
Rubber Save (Fine) difference &a particle size distribution of the
(particles) samples testad so the slot/particle size ratio
plot has been used here where comparisons are
Thermoset Poly-Cycle Indus. Woo-Bridge made. The ground coal demonstrated good
Rubber Super temperature stability in both the API tests
(particles C Ues-Bridge (Figure 7) and the LCTF tests (Figure 8 ) .
flakes1 S U I Flake However, as can be seen in Figure 7 and 9, the
material did not plug a slot size comparable to
its largest particle sire. This is bslfeved
Ground Coal Dowell Kolite due to the brittle nature of the material which
appeared to break up in the LCTF test. Potice
Ground Coal UcCabe-Uoody Kol-Sea1 that the slot/particle ratio for an -always
plugs" condition at 20 lbs. concentration for
Uineral Fiber Rockwol Indus. StrataUool the API test (Figure 7) was 0.8, whereas for
the LCTF test (Figure 9) the ratio is about
TEST RESULTS 0.1. Further evidence of this effect in the
LCTF tests was the significant viscosity
Results of the tests conducted both with increase in the nrud/LCU mix that turned a
the modified API Test Cell and the Lost charcoal grey color. Because of this
Circulation Test Facility (LCTF) are shown in phenomenon. a new md/LCn batch was lniwed for
the accompanying graphs. The test facility and each concentration and tentperature tasted.
conditions are noted for each figure. Since the nrud/LCU is continuously pumped in a
closed loop for several minutes during an LCTF
DISCUSSIOY test, it can be argued that the test is more
realistic or perhaps too severe. Regardless of
Thermoset Rubber the point of view, the brittle material does
Earlier tests of thermoset rubber LCn not perform as well in the slot test.
usins the modified API test cell showed this
material to be superior to several cellulosic Mineral Wool
materials that were also tested (Hinkebein, It is well known that a wide particle size
1982). Tests conducted in this series using distribution and a variety of shapes makes the
the same test cell showed the perforrmnce best LCH (Uessen6er. 1981). While a fiber
actually improved somewhat after the aud/LCU material alone is effective for stopping loss
mix was exposed to 400.F for 4 hours in a in peramable fomations, it is much less
roller oven (see Figure 3). However, when the effective in fractures. When a fiber LCn is
material was tested at elevated temperature in tested alone in the API test cell a degree of
the large-scale me its perfomance suffered control over fluid loss can be achieved but a
noticeably as the temperature was increased high pressure seal cannot. A series of tests
(see Figure 4 ) . Although there was very little was run using the API test cell to evaluate
evidence of melting when the bulk material was different blends of mineral wool (Stratawool)
heated to 500'F in the screening tests, these and a particle LCU (Hi-Seal/-Thermoret Rubber)
tests show that the material does soften which to determine the perfo-ce of the combined
is manifest by its reduced ability to hold a materials. The results, plotted in Figure 10,
seal at high temperatures and high pressure show a significant improvement in filtrate loss
differentials. The plastic nature of the before a seal was achieved by adding the
material is actually one of its best attributes StrataUool mineral fiber to the Hi-Seal
at temperatures below 200.F. since it will particle LCLI at a ratio of 1 to 2 by weight.
deform under pressure and temperature without Further tests are planned for the LCTF.
fracturing. This was evident by examining the
plugs formed in thesa tests. A more rigid Combined LCU Shapes (Thermoset Rubber)
aaterial will fall away from the face of the A range of sires of irregular shape flakes
slot (fracture) when the differential pressure were gleaned during the processing of the
is reloived whereas the thermoset rubber salva6ed battery cases and were used ia an
remains wedged in place. Also, because of ita oxperbent to d e t e d n e what performonce
good particle stability, there was no dramatic improvement could be mado by adding flake
increase in lnrd viscosity vfth t.rpperature as material to the particle LQI, Hi-Seal. A plot
can be expected with materials that decompose summorizing the results of thesa toots is shown
(0.g. cellulose) at high tem?eraturo. Figure in Figure 11. It shows a 50% improvamant was
S<LpI UP 131 test) and Figure 6 (LCTP Test) attained by using a flake size distribution
show the test results for this material plotted twice the size of the particle size distribu-
in the slot/particla size ratio fonnat tion at a 4 to 1 (particle to flake) weight
described earlier. ratio. The combined matorial sir.
distribution, plotted in Fiiure 12, was then
tested in the LCTF. Those results, plotted in
Loeppke
F i g u r e 13, a h w a 2.4 slot/particle ratio at REFERENCES
"20 lbs concentration/alw~ysplugs" compared to
about 1.1 (rigure 6) for Hi-Seal alone. Caskey. 8. C.. "Lost Circulation Technolosy
However, the temperature sensitivity of this Workshop", SMDBS-0109. March 1985
matorial above 200-F is mom pronounced in the
larger fracture/slot sizes than is Hi-Seal Caskey. 8 . C., Loeppke. C. E. and Satrape. J . V.
8101182 because of the flexible nature of the 1985. "Lost Circulation in Csothenual
flakes. A blend of particles and flakes called Wells: Research and Development Status at
Super Wes-Bridge which has a material Sandia. Geothermal Resources Council.
distribution of larger particle and flake sizes TWSACTIOIS, Vol. 9, August 1985.
(Figure 14) was mixed 1 to 1 (by weight) with
Hi-Seal and tested in the API test cell. Those Civler, Rick C.. 1985, Plugging Xechanlsms in a
results. plotted in Figure 15, show a slot/- Lost Circulation Model," Ceothemal
particle ratio (at 20 lbs concentration/always Resources Council. TBWSACTIONS, Vol. 9.
plugs) of 0.9 which can be compared to Hi-Seal August 1985.
alone (Figure 5 ) of less than 0.8.
Goodman, H. A . , 1981, "Lost Circulation i n Ceo-
TEST BeSULTS COWPARED thermal Wells: Survey and Kvaluotion o f
Industry Experience."
Figures 16, I t . and 18 show a comparison of
reaults for the different materials tested and Hinkebein, T. E.. Behr. V. L., and Wilde. S. ,.I
for the two test methods used. The ability of 1982, "Static Slot Testing of Conventional
a given material to plug almost a 50% larger Lost Circulation Haterialo," Sandia
slot when tested in the LCTF is shown in Figure National Laboratories Report. SAIiD82-1080,
16. Pigures 17 and 18 are comparisons of the January 1982.
material combinations tested in the two
facilities (API test cell and the LCTP). The Loeppke, G. E. and Caskey. B. C., 1983, "A Full-
important findings shown here are: Scale Facility for the Evaluation of Lost
1) If a particle L a is used, it is of little Circulation Xaterials and Techniques,"
benefit to use concentrations of more than 20 Geothermal Resources Council. TRAMSACTIOW
lbdbbl. In fact, the same size fracture can Vol. I , October 1983.
oe plugged with 5 lbs/bbl but the filtrate loss
and the time to plug is much greater. Uessenger, J. U.. "Lost Circulation" Penn Well
2) If an LCM with a combination of shapes is Publishing Co., 1981.
used. concentrations as high as 30 or 40
lbslbblmay be practical. Pye. Steve, 1985, personal comnication to
3) Larger fracture sizes are best sealed using 8 . C. Caskey. February 1985.
larger particle La's or by c e i n i n g particles
and flakes or fibers for more effective seals.

co~cLusIoPs 3000 I
- AMBIENTTEMPERATURE
I

----
I I
AMBIENTTEMPERATURE
AFTER
Thermoset cubber LCM (ground battery AGING HRS AT 400.F
4
casings) was tested at elevated temperatures in
8 large-scale facility. The plastic nature of
the material makes it an excellent plussing
material at temperatures below 200.P but at
higher temperatures the material softens and
ita ability to seal at high prersure differ-
entials is reduced. Ground coal perform
essentially the E- at all temperatures; but.
bec8use it is a brittle material it does not
p h g fractures as well as thermoset rubber.
Stratauool. a mineral fiber. added to the
particle tcLI (Hi-Seal) made the greatest
I
perforrnnnce improvement of the materials tested
but is yet to be tested at high temperature in
the large facility. Tests using the la-e-
scale Lost Circulation Test Facility show there
is little benefit in using particle LCH
concentrations above 20 lbslbbl; but, if an Lcw 200 400 600 800 1OOO
with a combination of shapes is used higher PRESSURE (psi)
concentrations can be beneficial. tcw*s
perfom better in the large-scale facility Figure 3 . API Test Cell results showing
which has a larger filtrate capacity and where the effects of temperature
the mud/Lcn is circulated through the test aging Hi-Seal (Thermoset
ves8.l across the face of the slot or fracture. Rubber). Slot size = 0 . 0 8
inches.
 Loeppke
. . . .- . ..
lo00 25

3 20
-d
\

16

f
W
10
0
z
0
0 5
4
5 400
I 1
u)
0.04 0.06 0.08 0.10 SLOT SUE linl
%
U 0.435 0.652 0.870 1.087 &aLE RATIO

= 200 Figure 7. API Test Cell results showing

the plugging ability of
Kolite (ground coal) at
10000 psi; pressure differ-
" ential and ambient tempera-
100 150 200 250 300 350
ture and, after temperature
TEMPERATURE (F)
aging 4 hours at 400 F.
Figure 4 . LCTF test results showing tile
effect of testing Hi-Seal
(Thermoset Rubber) at elevated KOL-SEAL
temperature. LCTF
TEMPERATURE EFFECTS
0.08 ill SLOT
1000 I

- 26
800 .-

$ 20
-
0 F I
5 16
F
f
W
0
10

-7
0.040
0.63
Figure 5.
0.060
0.80
0.080
1.07
0.100
1.33

API Test Cell results showing

0.120 SLOTSlZE(inl
1.60 &%, RATIO *0°0 0 I 6 10

CONCENTRATION (Ib/bbl)
16 20

the plugging ability of

Hi-Seal (Thermoset Rubber) Figure 8 . LCTF test results showing the
at ambient temperature and performance of Kol-Seal
1000 psi pressure differential. (ground coal) using a 0.080
inch slot at elevated
temperatures.

25 -
=
0

-z2 -
n
20

9 15 -
t NEVER mum
E 10-
f
0
Z 6 -
0
V
0' 1 1 I
0./2 SLOTS12Elh)
0.04
0.63
0.06
0.80
0.08
1.07
0.10
1.33
0.12 LLOTSlZf lid
1.60 RATlO
'0.04
0.34
0.06
0.06
0.08
0.67
0.iO
0.84 1.01'eu RAnO

Figure 6 . LCTF test results showing the Figure 9 . LCTF test results showing the
ability of Hi-Seal (Thermoset plugging ability of Kol-Seal
Rubber) at ambient tempera- (ground coal) at ambient
ture and 1000 psi; pressure temperature and 1000 psi
differential. pressure differential.
 Loeppke ,
---
-
I
S A N l FLAKE

lo-
n HCSEM

SIEVE SUE (in1

Figure 12. Material size distribution

for Hi-Seal (particle LCM)
and Sanl (flake LCM) sized
for best performance with
0 200 400 600 800 lo00 Hi-Seal when combined 1:4
(by weight).
Figure 10. API TEST CELL results showing
the effect on fluid loss by
combining a fiber LCM (SW/
Strata Wool) with a particle 25
LCM (HS/Hi-Seal) , 0 . 0 8 0 inch -
slot and ambient temperature. 20
-
0
16
5E 10
2500
E
V
2 6
0

--
V
2000 01 ' ' I " ' I
0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 SLOTSUElinI
E 0.63 0.60 1.07 1.33 1.60 1.87 2.13 2.40 2.67 SLOT
P A ~ L E
E
d 1500 Figure 13. LCTF test results showing
the plugging ability of a
combined LCM (Sanl flake t
Hi-Seal/l:l by weight) at
6 1000 ambient temperature and
1000 psi pressure differ-
d
Lu ential.
600

_I
PARTICLE
0 200 400 600 800 1000 r--1 TOTAL
I 1
PRESSURE (psi) I I

Figure 11. API Test Cell results showing

the effect on filtrate loss
using a 5 lb/bbl concentration
of different sizes of flakes
combined with the particle LCM
Hi-Seal (1:l by weight) c o m -
pared to ifi-Seal only, 0 . 8 0 8
inch slot and ambient 4; .-,
temperature.
SIEVE SIZE (in1
 PARTICLES PARTICLES 76%PARTICLES 25% FLAKE
KOUTE HI-SEAL HI SEAL + SUPERWES-lIRIOGE

0.080 0.120 0.160 0.200 0.240 SLOTSUElinl "

0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50
0.452 0.678 0.904 1.130 1.366 p#$LE RATIO SLOT
PARTtCLE SUE RAT'o
Figure 15. API Test Cell results show-
ing the plugging ability of Figure 17. API Test Cell results for
Super Wes-Bridge and Hi-Seal different LCM's compared, flake
combined (1:l by weight) at or-fiber size was not considered
ambient temperature and 1000 in the slot/particle size ratio
psi pressure differential. calculation, tests at ambient
temperature and 1000 psi pres-
sure differential.
25 1 I 1

GROUND THERMOSET

(50% HI-SEAL 60%BANI FLAKE)

0.40 0.06 0.08 0.10 0.12 '6.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.60
SLOT SIZE (in) &&LE SIZE RATIO

Figure 16. The test results for Hi-Seal

LCM compared from the API RP
131 Test Cell and the Lost
Circulation Test Facility
test at ambient temperature
and 1000 psi pressure differ-
ential.
F i g u r e 18. LCTF T e s t r e s u l t s for d i f f e r -
ent LCM's compared, flake size
was not considered in the slot/
particle ratio. Tests at
ambient temperature and 1000
psi pressure differential.

DISCLAIMER

This report was prepared as an account of work sponsored by an agency of the United States
Government. Neither the United States Government nor any agency thereof, nor any of their
cmployccs, makes any warranty, express or implied, or assumes any legal liability or responsi-
bility for the accuracy, completeness, or usefulness of any information. apparatus, product, or
pnxxss disclosed, or represents that its use would not infringe privately owned rights. Rcfer-
encx herein to any specific commercial product, process, or seMe by trade name, trademark,
manufacturer, or otherwise does not necessarily .constitute or imply its endorsement, rccom-
mendation, or favoring by the United States Government or any agency thereof. The views
and opinions of authors cxpresscd herein do not ncctSSarily state or reflect those of the
United States Government or any agency thereof.