Riley B.

Needham is Manager of Drilling and Production Research and Serv-

ices in the E&P Group of Phillips Petroleum Co., with responsibility for research
and service work in oil recovery processes, well completions, reservoir fundamen-
tals, reservoir simulation, well simulation practices, and drilling fluids. He holds
BS and PhD degrees in petroleum engineering from the U. of Oklahoma.
Needham served as a 1985-86 SPE Distinguished Lecturer and a member of
the 1985-86 Engineering Manpower and 1983-84 Reprint Series committees.
Peter H. Doe is currently Staff Director, Drilling and Production in the Research
and Services Div. of Phillips Petroleum Co., with responsibility for research in
mobility control, drilling fluids, formation damage, and stimulation. He holds BS
and PhD degrees in chemistry from the U. of Bristol.
Needham Doe

Polymer Flooding Review

Riley B. Needham, SPE, Phillips Petroleum Co.
Peter H. Doe, SPE, Phillips Petroleum Co.

Summary. This paper reviews published results of the use of polymers to improve oil recovery. A discussion of the capabilities
of the available types of polymers and where they have been successful is coupled with the principles of the mechanisms of poly-
mer flooding to serve as a guide for future applications. The scope of this review is limited to case histories where full-scale poly-
mer floods were applied, as opposed to near-well treatments.

Introduction
The purpose of this paper is to describe briefly the principles Fractional Flow. The way in which a section of reservoir
involved in polymer flooding and to review field experience. approaches its ultimate ROS is a function of the relative
Earlier reviews by Jewett and Schurz 1 and Chang 2 have permeability relationships and of the viscosities of the oil and
covered much of this same ground. Chang, in particular, water phases. These are combined in the concept of fractional
presents an extensive review of the polymer flooding literature. flow. By applying Darcy's law to the oil and water phases
Therefore, we have updated the list of literature rather than flowing simultaneously through a segment of a porous
repeating those included in these previous papers. We have medium, the fractional flow of oil, fo' can be derived as
tried to summarize the major points, particularly in relation to
the most recent field case histories. The scope of this review is
limited to what we refer to as "full-scale" polymer floods.
fo= (1)
This includes those cases where crosslinking agents have been 1+ J.Lok w/J.Lwko
used to produce an in-depth permeability contrast correction,
but excludes near-well, low-volume polymer gel treatments.
Consequently, all results of treatments of producing wells have Any change that reduces the ratio J.Lok w/J.Lwko will improve the
been excluded from this review. rate of oil recovery by increasing the fractional flow of oil.
Polymers can do this by increasing the viscosity of the water,
J.Lw' Once they have flooded a zone, some polymers also
Definition and Mechanisms of Polymer Flooding reduce the relative permeability to water, kW'
Oil and water are immiscible fluids. As a result, neither can This effect applies to any part of the reservoir where there is
completely displace the other from an oil reservoir. This is a mobile oil saturation-Le., anywhere that the relative
reflected in the irreducible water and residual oil saturations permeability to oil is greater than zero. However, if k o is
(ROS's) on a relative-permeability curve. Regardless of the already small because the mobile oil saturation is low, then fo
amount of water cycled through the system, the oil saturation will remain small at any achievable k w or J.Lw' The fractional
will not be reduced below the ROS. In polymer flooding, a flow effect therefore is more significant for polymer floods
water-soluble polymer is added to the flood water. This conducted early in the life of a waterflood while the mobile oil
increases the viscosity of the water. Depending on the type of saturation is high.
polymer used, the effective permeability to water can be An additional consideration is the oil viscosity, J.Lo' All else
reduced in the swept zones. Polymer flooding does not reduce being equal, the fractional flow of water will be greater in
the ROS, but is rather a way to reach the ROS more quickly reservoirs where the oil viscosity is high. This leads to early
or to allow it to be reached economically. water breakthrough and relatively high water production when
There are three potential ways in which a polymer flood can there is still a significant mobile oil saturation. Fractional flow
make the oil recovery process more efficient: (1) through the effects are thus likely to be more significant in viscous oil
effects of polymers on fractional flow, (2) by decreasing the reservoirs.
water/oil mobility ratio, and (3) by diverting injected water
from zones that have been swept. Mobility Ratio. Real reservoirs cannot be swept uniformly.
Even a homogeneous reservoir suffers from less than 100%
areal sweep at water breakthrough and at economically
Copyright 1987 Society of Petroleum Engineers achievable water/oil ratios (WOR's). The primary determinant

Journal of Petroleum Technology, December 1987 1503

of areal swee~ for a given well pattern and spacing is the the water used to dissolve the polymer. Similarly, the residual
mobility ratio of the flood, defined for waterflooding as resistance factor is the ratio of the mobility of water measured
before the injection of the polymer solution to the mobility of
water after polymer injection.
kw/Lo With these definitions, the benefits of fluid diversion are
M=-- , , (2)
achieved by high, long-lasting residual resistance factors.
/Lwko
Crosslinking of the polymer in situ can result in higher,
longer-lasting residual resistance factors compared with the use
For a given reservoir situation, the oil recovered before of polymer solutions alone.
water breaks through to the producing wells decreases as M
increases. Recovery at any later stage of the flood is also less
for a given volume of water injection. Polymers may improve Polymer Types and Properties
the mobility ratio of a flood by the same effects described Polymers that have been used in actual polymer floods
above-that is, by reducing k w or increasing /Lw' Again, there comprise the two general types of synthetic polymers and
is more potential for improvement when the mobile oil biopolymers.
saturation, and hence ko ' is relatively high. This again favors a A synthetic polymer almost always means polyacrylamides.
secondary over a tertiary application for a polymer flood. A variety of these are available from several manufacturers. In
general, the performance of a polyacrylamide in a flooding
Fluid Diversion Effects, We have identified two beneficial situation will depend on its molecular weight and its degree of
effects of polymer flooding: a more rapid oil displacement hydrolysis. In a partially hydrolyzed polyacrylamide, some of
through improved fractional flow characteristics and improved the acrylamide is replaced by, or converted into, acrylic acid.
areal sweep efficiency through improved mobility ratio. Both This tends to increase viscosity in fresh water, but to reduce
of these could be effective in homogeneous reservoirs, but both viscosity in hard waters. All else being equal, a high-
act only on a mobile oil saturation in the polymer-flooded molecular-weight polymer will produce higher viscosities and
zone. Perfectly homogeneous reservoirs do not exist. Most resistance factors than a low-molecular-weight polymer for a
reservoirs contain significant heterogeneities in the areal and given concentration. These potential advantages may be offset
particularly in the vertical sense. This leads to preferential by a greater tendency for shear degradation, which reduces
water entry into the more permeable zones and to a more rapid molecular weight, and by a reduced injectivity, which can be
sweepout of these zones. In the areas of the reservoir that are significant in low-permeability formations. For large-scale
contacted by the flood water, oil recovery may be very applications, polyacrylamides are available in powder form
efficient. The problem is that much of the reservoir remains (90% + active), in the form of a pumpable inverse emulsion
untouched. (33 to 55% active), or can be manufactured on site in a
A polymer injected into the flooded-out zone of such a concentrated solution form.
reservoir may recover very little oil from that zone. The effect Biopolymers are derived from a fermentation process, rather
of the polymer can be very beneficial, however, because of the than by direct synthesis from their monomers in a chemical
fluid diversion it produces. The polymer will build up flow reactor. The most commonly encountered biopolymer is
resistance in the portions of the reservoir it penetrates through xanthan gum, which is produced by the bacterium
the permeability reduction or viscosity increase discussed Xanthomonas campestris. The bacteria are cultured in a
earlier. This increased resistance to flow will divert suitable fermentation medium and produce the biopolymer as a
subsequently injected water into unswept or poorly swept byproduct of their metabolic processes. The polymer is
areas. separated from the rest of the cell material and sold for use in
In most floods that are initiated at high WOR's, fluid the form of a cQncentrated broth, typically containing 3 to
diversion will be much more significant than fractional flow or 13 % active polymer. In terms of molecular weight,
mobility ratio effects. Maximum benefits will likely be biopolymers fall toward the low end of the range encountered
achieved if the effects of the polymer can be sustained over a with polyacrylamides. Their molecular structure gives the
long time. This tends to place a premium on permeability molecule great stiffness. This characteristic gives biopolymers
reduction as opposed to straight viscosity improvement because excellent viscosifying power in high-salinity waters and makes
permeability reduction can be very long-lasting. Optimized them very resistant to shear degradation. In very fresh waters,
permeability reduction may make crosslinking of the polymer however, they have less viscosifying power than
desirable. polyacrylamides.
Methods of crosslinking polymers in situ have been known Each polymer type has advantages and disadvantages.
and practiced in the field for more than a clecade. The Polyacrylamides have a relatively low price, develop good
crosslinking can be achieved in a number of ways, including viscosities in fresh waters, and adsorb on the rock surface to
the use of multivalent cations and organic compounds. The produce a long-lasting permeability reduction (the residual
crosslinking causes the polymer to be linked into a network resistance effect). Their primary disadvantages are a tendency
that results in greater reductions in water permeability as well to shear degradation at high flow rates and poor performance
as longer-lasting permeability reductions in the part of the in high-salinity water (low viscosity and frequently excessive
reservoir where the crosslinking occurs. The resultant retention) .
permeability reduction causes subsequently injected water to be The primary advantages of biopolymers are their excellent
diverted into zones that have not been completely flooded. viscosifying power in high-salinity waters and their resistance
These fluid diversion effects would be expected to be most to shear degradation. Biopolymers are not retained on rock
important for floods initiated at high WOR's, where it is surfaces and thus propagate more readily into a formation than
already too late for fractional flow and mobility ratio polyacrylamides. This can reduce the amount required for a
improvements to be of much significance because of the flood but also means that there is no residual resistance effect.
prevailing low values of k o in the swept zones. It is therefore a questionable advantage.
In the literature on the behavior of polymer solutions in Both polymer types are restricted in the range of reservoir
porous media, "resistance factor" and "residual resistance conditions where they can be effective. Biopolymers thermally
factor" are frequently used as measures of the effectiveness of degrade too fast at temperatures above 200°F [93°C]. At
polymer solutions compared with that of water. By definition, temperatures above 170°F [n°C], polyacrylamides may
the resistance factor is the ratio of the mobility of water to the precipitate in waters containing too much calcium. In principle,
mobility of a polymer solution. Such a property depends on this does not prevent their being used successfully in fresh
the poro\.lS medium, the particular polymer used, the water, but it makes control of the salinity of the flood water
concentration of the polymer, and the salinity and hardness of much more critical.

1504 Journal of Petroleum Technology, December 1987

 TABLE 1-POLYMER FLOOD CASE HISTORIES-KEY PARAMETERS

Amount of
Starting PolymerlWater Polymer Used Recovery
Project Reference Flood Type WOR Lithology Type (Ibm/acre-tt) (%OOIP)
Northeast Hallsville 2 Secondary <1 Carbonate Polyacrylamide/fresh 12 13
Vernon 2 Secondary <1 Sandstone Polyacrylamide/salt' 136 30
Huntington Beach 2 Secondary 1 Sandstone Polyacrylamide/fresh 26 4
Brea Olinda 2 Secondary 1.2 Sandstone Polyacrylamidelfresh
Taber Manville South 2 Secondary 1 Sandstone Polyacrylamide/' 20 2
Skull Creek South 2 Secondary 0 Sandstone Polyacrylamide/fresh' 32 8
Brelum 2 Secondary <1 Sandstone Polyacrylamide/fresh' 70 9'
Wilmington 2 Secondary 0
North Burbank 2,8 Tertiary 100 Sandstone Polyacrylamide' '/fresh 63 2.S
North Alma Penn 2 Secondary 3 Sandstone Biopolymer/salt 18
Pembina 2 Secondary 9 Sandstone Polyacrylamide/fresh 22 0
West Semlek 2,4 Secondary 0 Sandstone Polyacrylamide/fresh 6 S
North Stanley 2,S Tertiary 70 Sandstone Polyacrylamide/fresh 2S 1.1
West Yellow Creek 2,6 Secondary 3 Sandstone Polyacrylamide/fresh
East Coalinga 2,7 Secondary 9 Sandstone Biopolymer/' 2 0
Skull Creek Newcastle 9 Secondary 3 Sandstone Polyacrylamide/fresh 28 10
Sage Spring Creek Unit A 10 Secondary <1 Sandstone Polyacrylamide' '/fresh S 1.2
Eliasville Caddo 11 Tertiary 32 Carbonate Polyacrylamide/fresh S6 1.8
Chateaurenard 12 Secondary 9 Sandstone Polyacrylamide/fresh 188' 2.S'
Storms Pool 13 Tertiary Sandstone' Biopolymer/' 100' 0'
Oerrel 14 Secondary 4 Sandstone Polyacrylamide/fresh 46 23
Hankensbuettel 14 Secondary S Sandstone Polyacrylamide/fresh 13
Owasco 1S Secondary <1 Sandstone Polyacrylamide/fresh 21 7
Stewart Ranch 4 Secondary 1 Sandstone Polyacrylamide/fresh 2S 8
OK 4 Secondary Sl:lndstone Polyacrylamide '/fresh' 16 3
Hamm 4 Secondary Sandstone Polyacrylam ide '/fresh ' 21 9
Kummerfeld 4 Secondary Sandstone Polyacrylamide '/fresh' 7 6

'Information uncertain or unavailable.

"Crosslinked flood.

More advanced polymers, both synthetic and natural, that

remove many of these temperature and salinity limitations have
DYKSTRA-PARSONS METHOD
> 1.0
been developed. Their high cost generally makes them of :::.:::
questionable utility in today's economic situation, and they are
considered outside the scope of this review. Z
0 0.6
r-
Field Experience -0::<l: 0.6
A number of papers have been published on field applications <l:
>
of polymer flooding. Ref. 2 lists 15 field applications. Updates
have subsequently appeared on four of these projects 4-7 and >-
r- 0.4
one has been expanded from pilot to commercial scale. 8 -.l
Twelve case histories were added to this list,4,9-15 giving a -CO
total of 27 floods to consider. Table 1 summarizes results of <l: 0.2
these applications. In some cases, the values appearing in the w
~ 100
table were not explicitly presented in the referenced paper. In 0::
W
those instances, we have estimated the parameters shown using
the relevant data presented in the appropriate paper. An
asterisk has been used to indicate that a value is considered
particularly uncertain, or that a value cannot be calculated.
0... 0
0 10 20 30
OIL RECOVERY, % OOIP
40
"'"
50 60

Of the 27 floods, 23 were essentially secondary operations, Fig. 1-From Fig. 3 of Ref. 16 with Sw =O.35-the economic
initiated at WOR < 10. The average polymer flood recovery limit is for WOR 25. =
from these case histories is approximately 8 % of the original
oil in place (OOIP) (counting apparent failures as zero
recovery). The average amount of polymer injected is around 2. A technically successful tertiary polymer flood requires
30 Ibm/acre-ft [0.011 kg/m 3 ] of reservoir. more polymer per barrel of oil recovered. The amount of
Four of the floods were tertiary applications, initiated at polymer used to recover a barrel of oil appears to have been
WOR's of 30 to 100. One was a failure. 7 The other three about six times greater in tertiary than in secondary
produced an average of 1.8% OOIP for a polymer usage of applications (2 Ibm/bbl compared with 0.3 lbm/bbl [5.7 kg/m 3
approximately 50 Ibm/acre-ft [0.018 kg/m 3 ] of reservoir. compared with 0.1 kg/m 3 ]).
Averages like these should be treated with caution. Only two of the 27 field projects were in limestones. Both
However, they do illustrate some important aspects of polymer were successful, however, and they include one secondary and
flood applications. one tertiary application. The conclusion must be that polymer
1. Polymer flooding has much greater potential as a flooding is applicable to both sandstones and carbonates, but
secondary process than in postwaterflood applications. The has been less widely applied in carbonates.
averages presented above indicate roughly four times the The case histories include only three biopolymer floods. The
potential recovery for a secondary compared to a tertiary remainder were all polyacrylamide floods in relatively fresh
flood. injection water. The most saline water used appears to have

Journal of Petroleum Technology, December 1987 1505

contained 400 ppm calcium, which is well within the tolerance In the field practices reviewed, all the successful applications
limits of polyacrylamides at low temperatures. None of the used polyacrylamides in relatively fresh water. They would
biopolymer floods can be clearly identified as a success. Two therefore have included permeability-reduction effects,
were clearly failures, although it is not certain that the polymer regardless of whether these were part of the design philosophy
should be implicated as the cause of the failure. of the flood. Simulation results indicate that a two- to five-fold
Five of the case histories (four secondary, one tertiary) used reduction in the permeability of the highest-permeability zones
aluminum citrate crosslinking to enhance permeability produces the most significant effects. Noncrosslinked
reduction effects. 4 ,8,10 The reported recoveries from the polyacrylamide floods in fresh water can reach the middle of
secondary floods are about 1.5 times higher than for the this range.
noncrosslinked floods per pound of polymer used. The tertiary An additional parameter that is always significant in
flood recovery appears comparable to noncrosslinked simulation studies is the retention level of the polymer on the
applications. reservoir rock. The lower the retention level, all else being
This survey is by definition selective. It includes only case equal, the better the flood performance. Biopolymers do not
histories that can be documented by publications. A much adsorb and polyacrylamides in fresh water have moderate
larger number of polymer projects have been run and not adsorption, but polyacrylamides in salt water generally show
reported. Not all parameters of interest can be determined, very high retention levels, without any corresponding benefit in
even for the published cases. For instance, some publications terms of improved permeability reduction. This may be the
do not indicate how much polymer was injected or how much reason why no successes have been reported with
oil was recovered. Some major gaps in our knowledge stand polyacrylamides in highly saline brines. There have certainly
out. Are there any successful biopolymer case histories and been some unreported failures.
have polyacrylamides been successfully used in high-hardness,
high-salinity injection waters, for example?
Conclusions
Guidelines for Polymer Application 1. Economic and technical successes have been reported for
polymer floods in both secondary and tertiary applications and
Field case histories are not sufficiently complete or extensive
in both sandstones and carbonates.
to provide solid guidelines for identifying suitable candidates
2. Secondary floods recover substantially more oil for less
for polymer flooding. Filling in the gaps requires some kind of
polymer usage than tertiary floods. Polymer flooding is
model study. The use of a model to design a successful
therefore best applied early in the life of a waterflood. The
polymer flood requires that the objective of the polymer flood
average performance of floods initiated at WOR> 10 appears
be clearly identified and that the critical reservoir parameters
be established. Fig. 1 qualitatively shows the interaction of the to be significantly lower.
important parameters of mobility ratio and reservoir 3. On the basis of published results to date, all successful
applications have used polyacrylamides in relatively fresh
heterogeneity as reflected in the coefficient of permeability
water. These floods would have the combined effects of an
variation, K v. The information 16 is for the performance of a
increase in the water viscosity and a reduction in the
waterflood and can be used to identify the particular parameter
permeability to the injected water.
that is primarily limiting the oil recovery from water injection.
4. Our experience in simulation of the results of polymer
Frequently, it is not just one parameter, and the relative
floods indicates that mobile oil saturation is a key variable that
importance of each parameter needs to be considered to
determines whether a polymer flood can be successful.
determine whether a polymer flood that has a long-lasting
Heterogeneous reservoirs or those containing viscous oils will
residual resistance effect or a large increase in water viscosity
reach high WOR with significant remaining mobile oil.
is more advantageous.
In addition, where available, the waterflood history can be a
very good indicator of the reservoir heterogeneity and its effect Nomenclature
on the oil recovery. fa = volumetric fraction of the total flow that is oil
Simulations of the performance of polymer floods indicate ko = permeability to oil
that the mobile oil saturation present at the initiation of the
kw = permeability to water
polymer flood is a key variable in determining the
Kv = Dykstra-Parsons coefficient of permeability variation
effectiveness of the flood. In general, a high mobile oil
saturation will be encountered in three situations: (1) in any M = mobility ratio (water mobility/oil mobility)
secondary flood (low WOR at flood initiation); (2) in a tertiary !-to = viscosity of oil
flood in a viscous oil reservoir (high WOR caused primarily !-tw = viscosity of water
by adverse mobility ratio and fractional flow characteristics for
waterflood); and (3) in a tertiary flood in any heterogeneous
reservoir (high WOR caused primarily by poor vertical References
conformance of the waterflood). I. Jewett, R.L. and Schurz, G.F.: "Polymer F1ooding-A Current
The first two cases are considerations for mobility-control Appraisal," JPT (June 1970) 675-84.
type of polymer floods and would be expected to place a 2. Chang, H.L.: "Polymer Flooding Technology-Yesterday, Today, and
premium on viscosity development by the polymer used. The Tomorrow," JPT (Aug. 1978) 1113-28.
3. Craig, F.F. Jr.: The Reservoir Engineering Aspects of Waterflooding,
third can be successful only through fluid diversion, which
Monograph Series, SPE, Richardson, TX (1980) 3, 45-47.
tends to emphasize permeability reduction. Cases 1 and 2 4. Mack, J.: "Process Technology Improves Oil Recovery," Oil & Gas
should be candidates for polyacrylamide in freshwater floods J. (Oct. I, 1979) 67-71.
or for biopolymer floods in waters with a wide range of 5. Smith, RV. and Burtch, F.W.: "Study Shows N. Stanley Field Polymer
salinities and hardness. Case 3 is a candidate only for a Flood Economics," Oil & Gas J. (Nov. 24, 1980) 127-34.
polyacrylamide flood, and performance might be enhanced by 6. Gordon, S.P. and Owen, O.K.: "Surveillance and Performance of an
crosslinking to obtain more permeability reduction. Existing Polymer Flood: A Case History of West Yellow Creek," paper
One clear precaution is required when considering SPE 8202 presented at the 1979 SPE Annual Technical Conference and
application of polymer flooding to reservoirs that contain Exhibition, Las Vegas, Sept. 23-26.
7. Peterson, J.E.: "Coalinga Polymer Demonstration Project," DOE/SAN/
viscous oils and that also exhibit a high degree of 1556-5, Final Report, Natl. Technical Information Service, Springfield,
heterogeneity. As shown in Fig. 1, success would require a VA (July 1976-Dec. 1980).
very large decrease in the mobility ratio, starting at a ratio of 8. Zornes, D.R., Cornelius, A.J., and Long, H.Q.: "Overview and
100 and a permeability variation of 0.8 or higher. Such Evaluation of the North Burbank Unit Block A Polymerflood Project,
increases may not be economically achievable using available Osage County, Oklahoma," paper SPE 14113 presented at the 1986
polymers in available injection waters. SPE IntI. Meeting on Petroleum Engineering, Beijing, March 17-20.

1506 Journal of Petroleum Technology, December 1987

 9. Janeczko, M.A.: "Skull Creek Newcastle Sand Unit-A Successful 15. Weiss, W.W. and Chain, J.: "Owasco Unit Polymer Flood Increases
Polymer Flood," paper SPE 8380 presented at the 1979 SPE Annual Oil Recovery," Oil & Gas 1. (Aug. 7, 1978) 80-82.
Technical Conference and Exhibition, Las Vegas, Sept. 23-26. 16. Johnson, C.E. Jr.: "Prediction of Oil Recovery by Waterflood-A
10. Mack, J.C. and Warren, J.: "Performance and Operation of a Simplified Graphical Treatment of the Dykstra-Parsons Method,"
Crosslinked Polymer Flood at Sage Spring Creek Unit A, Natrona Trans., AIME (1956) 207,345-46.
County, Wyoming," lPT (July 1984) 1145-56.
II. Weiss, W.W. and Baldwin, R.W.: "Planning and Implementing a Large- 51 Metric Conversion Factors
Scale Polymer Flood," lPT (April 1985) 720-30.
12. Labastie, A. and Vio, L.: "The Chateaurenard (France) Polymer Field acre-ft x 1.233 482 E+03 m3
Test," Enhanced Oil Recovery, Elsevier Scientific Publishing Co., New ibm x 4.535 924 E-Ol kg
York City (1981) 213-22. JPT
13. Craig, F.F. III: "Enhanced Oil Recovery By Improved Waterflooding,"
This paper is SPE 17140. Distinguished Author Series articles are general, descriptive
DOE/ET/12065-66, Final Report, Natl. Technical Information Service, presentations that summarize the state of the art in an area of technology by describing
Springfield, VA (Feb. 1984). recent developments for readers who are not specialists in the topics discussed. Written
14. Maitin, B.K. and Volz, H.: "Performance of Deutsche Texaco AG's by individuals recognized as experts in the area, these articles provide key references to
more definitive work and present specific details only to illustrate the technology. Purpose:
Oerrel and Hankensbuettel Polymer Floods, " paper SPE 9794 presented To inform the general readership of recent advances in various areas of petroleum engi-
at the 1981 SPEIDOE Enhanced Oil Recovery Symposium, Tulsa, April neering. A softbound anthology, SPE Distinguished Author Series, Dec. 1981-Dec. 1983,
5-8. is available from SPE's Book Order Dept.

Journal of Petroleum Technology, December 1987 1507