Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

SOCIETY OF PETROLEUM ENGINEERS OF AIMES

6200 North Central Expressway &!& SPE 3691


Dallas, Texas v5206 -

THIS IS A PREPRINT --- SUBJECT To Correction

The Effect of Viscosity on Drag Reducer


Performance in Oil Pipelines

By

Jaime A. Lescarboura and Harry H. Wahl, Members AIME, ContinentalOil Co.

@ Copyright 1971
American Institute of Mining, Metallurgical, and Petroleum Engineers, Inc.

This paper was prepared for the k2ndAnnual CaliforniaRegional Meeting of the Society of
-----
Petroieum Engineers or Am, to be ‘neldin Los Angele~3 Ca~if., ~~~~.~-?~ 19V. D.WM:.“4-..
EGAIUA”DAUU+n
uu
copy is restricted to an abstract of not more than 300 words. Illustrationsmay not be copied.
The abstract should contain conspicuousacknowledgmentof where and by whom the paper is pre-
sented. Publication elsewhere after publication in the JOURNAL OF PETROLEUM TECHNOLOGY or the
s~c~~ OF DWIIROT.WTTM nnvzmwm~
.U..L”UU”...
-A.v .TfXTRNAT, is
------ .-“.--— -- II=IImIV
------V g~~~~~~ UpQn request to the Editor of the
appropriate journal provided agreement to give proper credit is made.

Discussion of this paper is invited. Three copies of any discussion should be sent to the
Society of Petroleum Engineers office. Such discussion may be presented at the above meeting and,
with the paper, may be considered for publication in one of the two SPE magazines.

------------
-..
ABSTRACT lN”l’KUIJUtil’lUN

The polymeric drag reducer CDR has been Drag reduction is the increase in
tested in the laboratory in oils with viscos- pumpability of a fluid caused by the addi-
ities ranging from 3 to 48 cp and in product tion of small amounte of additive to the
and crude oil pipelines at 3, 5, and 9 cp. fluid. It occurs only in turbulent flow.
Three regions have been identified in the At a given flow rate, the percent drag
drag reduction spectrum--(1) an extension of reduction is given by
leminar flow where drag reduction increases
AP - AP
with increasing flow velocity, (2) a trsnsi-
% D.R. .—A+x 100 (1)
tion region where drag reduction first de-
creases and then increases again with increas-
ing flow velocity, and (3) a region of maximum where AP is the pressure drop in the pipe
constant drag reduction. Above a certain without drag reducer, the base
critical concentration of polymer, the transi- pressure drop, and
tion from the laminar flow extension to the
maximum constant drag reduction occurs with- APP is the pressure drop in the line
out any inflection points. Below the critical when drag reducer is added to the
concentration the extension of laminar flow fluid,
increases with increasing oil viscosity. In
the region of extension of laminar flow, Because large pressure drop reductions can
significant drag reductions (up to 50 percent) be obtained with small amounts of drag
were obtained. reducer, drag reduction has a wide poten-
tial application within the oil industry.
--
lue
1-L-....4-A--
&ixJuLuLuLy
~...
uaLu
-L...+&.+
aLLuw Lb&al.
rmn
uu&\
4=
AD ail
effective drag reducer in all the oils tested We have reported recentlyl on a drag
regardless of their viscosity. The results reducing polymer, CDR, that gave substantial
of pipeline field tests in products and oils drag reduction in crude oil pipelines. How-
from 3 to 9 cp viscosity show that in the full ever, these tests were limited to relatively
turbulence region the performance of CDR is low viscosity fluids. Possible applications
independent of viscosity. of drag reduction in viscous solvents have

References and illustrations at end of paper.


I been suggested. TIIisinterest initiated a The fluids used in the field tests were
study of the effect of solvent viscosity on crudes of 3 and 5 cp viscosity at test tem-
drag reduction. peratures and a creosote-like material of
about 9 cp viscosity at pipeline temperatures,
I In this paper we present the results of
field and laboratory tests of CDR drag Laboratory tests were run with die~e~
reducer in oils ranging from 3 to 48 cp in and crude oils of 3 cp viscosity, crude oil
viscosity. of 5 cp viscosity, wood treating fluid of 8
cp viscosity, and pale oil-diesel blends of

I EXPERIMENliL 12, 19, 40, and 48 CP viscosity.

I Flow Systems

AH laboratory runs were performed in a


A description of the 27 test runs is
given in Table I. The testing and data
analysis procedures have been described at
l-inch pipe viscometer at 800F. The instru- length in reference 1. Drag reduction is
ment has been descrtbed in an earlier paper. 1 calculated from Equation 1.
The low viscosity fluids tested were diesel,
crude oils, and a creosote-like material
used to treat wood. The intermediate vis-
cosity fluids were blends of 80 pale oil
I RESULTS AND DISCUSSION

The results of the pipeline tests are


(80 Saybolt Sec. refined oil) and diesel, and presented in Figure 1 as drag reduction per-
the high viscosity fluids were blends of 400 cent versus Reynolds number. It is known
pale oil and diesel. Testing of the high that, for pipe diameters from 8 to 12
viscosity fluids strained the range of opera- inches, diameter has a linrLtedeffect on
tion of the viscometer and, therefore, only drag reduction.1-3 Thus, in Figure 1, the
relatively low Reynolds numbers were reached. data points for the 5 cp and 9 cp fluids
In fact, the highest viscosity fiuids couid
barely be pushed into the turbulent transi- lines obtained with 3 CP crude. As can be
tion zone (2100 < NW < 4000). Because of seen, they do not. The 250 Vppm (5 cp) data
the nature of the transition zone, the data point lies well below the 300 Vppm (3 cp)
scatter in the area was substantial. line, while the 1470 Vppm (9 cp) data point
is far above the 1000 Vppm (3 CP) line.
In all cases actual test samples were Figure 2 plots the same data in terms of
used to detertine t“hevi~cositY aiidik~L~~~j dr~g red~~~ion versus flOW velocity. vis-
at 800F. Each run consisted of a base run cosity differences are not taken into
of untreated solvent and a drag reduction run account. The solid lines in the figure are
wfth the appropriate concentration of CDR. the performances predicted for the different
concentrations of CDR by the predictive
The field tests with crude oils were run equation developed in reference 1. This
in two Oklahoma pipelines, which have been equation, which is given below, assumes that
described in detail elsewhere. 1 A field test viscosity has no effect on drag reduction.
with the creosote-like fluid was run in the
Yellowstone pipeline in Montana. This (v - 1.5)C
%D.R. = (2)
product pipeline, of about 10 inches ID, runs 23 i- 2.2v+0.093C (V-1.5)*312
from Billings, Montana to Spokane, Washington.
The test segment was 59.3 miles long and
extended between Springdaie and Manhattan
I where v is the flow velocity in ft/sec

pump stations. The line was tapped for pres-


sure measurements at both pump stations and
I C is CDR concentration in volume ppm.

at two intermediate points. The flow rate AS the figure shows, the agreement between
was measured by positive displacement meter. predicted and actual perform&ce is excel-
The pipeline flow rate was essentially con- lent. ~is indicates that in high turbulence
stsnt since the output of the pipeline pumps flow, viscosity has little 6r riosffect c=
was fixed. drag reduction.

I Experimental Materials The raw laboratory data of Figures 3 and


i 4 skt~$Fle%*tk,e‘~>~=iv:o~
------ from bminar to
....-.
The drag reducer used is a high turbulent flow occurs in the untreated fluid
molecular weight hydrocarbon polymer named and in the fluid treated with CDR. This
CDR. It is in liquid form and has a density transition was not observed previously because
I of 5.9 ill/gai.anda V~SCGSit~ of 100 cp at with low viscosity fluids it occurs below the
750F. flow capabilities of the test equipment.
Percent drag reduction versus Reynolds number
E 3691 JAIME A.LESCARBOUP AND HARRY A. WAHL 3

for the various fluids and CDR concentrations ultimately, at high Reynolds numbers, the
is given in Figures 5 to 7 for the laboratory higher concentration gives the greatest drag
tests. These figures include a theoretical reduction. This shows that if turbulence is
maximum drag reduction curve. This maximum high enough viscosity effects are un-
drag reduction is defined as the value ob- important. The field tests verify this con-
tained, from Equation 1, when the drag clusion.
reducer extends the laminar regime beyond
Reynolds numbers of 2100. Figures 3 and 4 Figure 7 shows data run at various.
show that this theoretical maximum is usually concentrations in three low viscosity oils.
not reached since the additive increases the The effects observed previously are verified
viscosity of the solution. Thus, the laminar here. The effect of concentration at con-
pressure loss of the solution is higher than stant viscosity is clearly shown by the 300
that of the solvent alone. Another important Vppm, 2300 Vppm, and 5000 Vppm lines for the
fact about this theoretical maximum drag 3 cp oil; the 300 Vppm, 920 Vppm, and 1500
reduction is that it is not zero at 2100 Vppm lines for the 5 cp oil; the 1500 Vppm
Reyn0id5 number. Admittedly, there can be no and 3(KXI Vppiii Iirlesfcr +!-.a
Q nm
~..=. -p m-ii
u+- . .Agai~.~
drag reduction until the flow turns turbulent. at constant viscosity, the lower the CDR
However, as the untreated fluid goes into concentration the sooner transition begins.
turbulence, the pressure drop increases Comparing the 1500 Vppm lines for 5 and 8 CP
suddenly (see Figures 3 and 4). If a drag oils, it is again evident that at higher
reducer extends laminar flow (the triangles Reynolds numbers viscosity becomes un-
in the figures), the jump does not occur; important and equal concentrations of CDR
thus, at a Reynolds number a little higher give equal drag reduction regardless of the
than 2100, the theoretical drag reduction is solvent. This effect is also demonstrated
around 35 percent. by the 3000 Vppm (8 CP) and 5000 Vppm (3 CP)
data. At hw ReY~i~~d~-..”.L--” L.++.* ~~-g
LLUULU=L. “.....

reduction is obtained with a lower concen-


Figure 5 shows the effect of viscosity
tration of CDR in the 8 cp oil than with a
on percent drag reduction at constant CDR
higher CDR concentration in the 3 cp oil.
concentration. Below a Reynolds number of
However, as Reynolds number increases, the
2100 the flow is laminar and there is no drag
5000 Vppm line crosses the 3000 Vppm line.
reduction. Above 2100, drag reduction for
At higher &ynolds numbers, high turbulence,
all viscosities falls on a curve slightly
5000 Vppm gives better drag reduction than
below the theoretical maximum drag reduction.
3000 Vppm, regardless of the viscosity of the
As Reynolds number increases, drag reduction
oils.
begins to deviate downward from the maximum
drag reduction line. TIKLS deviation, or
The results of this work show that there
transition, occurs at lower Reynolds numbers
iP.~m: 7:i~co~it3Tf~~id~, tb.~m.
i~.b.i~h.
x?is- are three regions in the drag reduction
spectrum. These regions are an ~Xt~iLSiO~I Gf
cosity fluids. Aa the figure shows, the drag
the laminar regime, a transition region, and
reduction transition begins at around 3000
a regitm of constant drag reduction. In the
for the 3 cp oil, at around 4000 for the 8
first region, total turbulence suppression
to 9 cp oils, and has not yet begun at 5000
occurs, while in the other regions there is
for the 40 cp oil. The dashed lines in the
an apparent turbulence attenuation. ‘he
figures indicate the probable location of
total turbulence suppression is clearly
trmsition . Since they are drawn through the *-—-—- h,, thm
aemonstrausu
-..->

LU
z- ma-----

rAguL=
‘1

d
.

=bu
...! ?74

.A~U..=
,-,,,va ,4

-f .1 . ..-
lowest data point measured for a particular
extension of the laminar flow line to
run, the lines are an upper bound to the
begi~.~ki~.g
of t-t-me+ t+~” ~.~~e ~-~IJ~& ~~~w
Reynolds numbers above 2100. The behavior
..-.”- .-”.. .
is shown in Figures 5 Lo 7 by the Iiileiif
that during transition drag reduction de-
maximum drag reduction located slightly below
creases, reaches a minimum , and then begins
the line of theoretical maximum drag reduc-
to .
.------- ag~iil.
~ucIe.ds=
tion. The theoretical maximum is not reached
because CDR makes the solution slightly more
The data in Figure 6 shows the effects
viscous than the solvent alone. Aa the
“of both concentration and viscosity. The
Reynolds number increases beyond 2100, the
3 cp data demonstrates that the fall-off from
drag reduction increases until turbulence
maximum drag reduction is concentration
appears. There is then a transition region
dependent aa well as viscosity dependent.
in which drag reduction first decreases and
Transition begins earlier for the 230 Vppm
then increases again as Reynolds number
solution than for the 690 Vppm concentration.
increases. Finally, a region is reached
Another important effect is shown by Figure 6.
where drag reduction remains constant as the
Ait’houg’hduring transition a ktigher clrzg
flow rate increases. This behavior is
reduction is obtained with 230 Vppm in 12 and
demonstrated in Figure 7 by the 300 and 900
19 cp oils than with 690 Vppm in 3 cp oil,
Vppm curves.
—..--- ~~~ ~~~;
4 THE EFFECT OF VISCOSITY ON DRAG REDUCER PERFORMANCE IN UIL PIFEIZNES

It might be argued that the maintaining N% = 2950 for 2300 ppm, but at N% = 4400 for
of laminar flow to Reynolds numbers above 5000 ppm. However, at a constant 1500 ppm
2100 is a non-Newtonian effect similar to CDR, the transition zone begins at NRS = 2950
that observed by Dodge and Metzner. q lhey for 5 cp oil, but at NW = 4400 for 8 CP oil.
found that as the fluid becomes more non-
Newtonian, as indicated by their non-Newtonhn These tests also show that the full
index n becoming less than unity, it required turbulence performance of CDR is independent
a higher Reynolds number to generate turbulent of the solvent, being dependent only on the
flow. However, a non-Newtonian index of 0.4 concentration. How soon this maximum per-
is needed to extend laminar flow to NRS = formance is reached is a function of solvent
~o~~● Me ha.~= fnt~ndthat
.—- ---— a solution contain- viscosity.
ing 1150 Vppm of CDR is still Newtonian
.. - -~ -L..- -.=---=1fiaq-l~,~+nns could
(n = 1.0), but extends the laminar regime to ~~~ OL L11=5c SGUGA-A tivLw-...-..--
Reynolds numbers over 5000 (cf. Figure 5). be deduced from Figure 5 and Figure 6. How-
Thus, the effect observed in this study is ever, the fact that most of these data lie
due to the specific action of a drag reducer in the turbulent transition region makes the
and is not a purely non-Newtonian effect. results harder to interpret. Moreover, the
measurements in the transition region
The appearance of the three regions in generally show more scatter than either
the drag reduction spectrum is dependent on laminar or fully turbulent data.
the viscosity of the solvent aa well aa on
drag reducer concentration. The field data, Figure 5 shows transition beginning below
which is in the transition region, indiea?es NRe = 2950 fo~ 3 w dl, at NRS ~ 29!20 fo~
that the appearance of the drag reduction 8-19 cp oil, and above N& = 5000 for 40 CP
regions is also diameter dependent. The flow oil. Figure 6 shows that for 230 ppm CDR
velocity in the field tests was not high the transition region begins below NW = 2900
, .4nm nf (-~p.~~~n.t drag
enougn to reaeklEk ree~w.... for 12 arrd19 cp oil. The fact that differ-
reduction. ences in solvent viscosity mask the effect
of concentration is shown by a comparison of
The discovery of the concentration and the 690 ppm, 3 CP run with the 460 ppm, 40-48
diameter dependent of the drag reduction cp runs. Transition begins later in the
spectrum verifies the work of Liaw5 who viscous oils than in diesel even though the
classified drag reducing poiymer soiutioiias concentration -= bun . . .,.6,,-.
UL
nnn 4. h40h-Y i~- ~~-~ ~ie~el

dilute or concentrated, depending on whether than in the viscous oils.


a critical concentration, Cc, is exceeded.
Above Cc pressure loss-flow rate plots show CONCLUSIONS
a gradual deviation from an extension of the
laminar line with no transition region. The major conclusions of this study are
This behavior is supported in our results by that CDR performs better in viscous oils
the 3000 Vppm solution of CDR in wood treat- than in low viscosity oils at low Reynolds
ing fluid (cf. Figure 7). Liaw found that numbers (2100 to 10,000) and that it performs
Cc is dependent on diameter but seems to be equally well in all viscosity levels at
independent of the solvent for a given Reynolds numbers above 10,000. For a given
polymer. Critical concentration increases pipe diameter, viscous fluids require higher
with pipe diameter approximately as D2. It flow velocities in order to reach a given
is possible that the drag reduction spectrum Reynolds number. 7%us, the velocities
is independent of the type of solvent for a required to achieve turbulent flow with
given polymer, However, the solvent WLS- viscous fluids may be outside the practical
cosity does have an effect on the appearance pipeline limits.
of the three regions of the drag reduction
spectruoo. most clearly sh~
~.i~ is in Three regions have been identified in
Figure 7 but can also be seen in Figure 5 the drag reduction spectrum: i) a region of
and Figure 6. total turbulence suppression where drag
reduction increases with increasing Reynolds
Figure 7 shows that CC for 8 CP oil is number or flow velocity, 2) a transition
below 3000 Vppm since the data shows no region where drag reduction first decreases
transition zone but only a gradual deviation and then increases again with increasing
from laminar flow. Similarly the Cc for 3 cp Reynolds number, and 3) a region of maximum
oil is above 5000 ppm since a transition zone constant drag reduction where the drag reduc-
still exists. It is apparent that the tion is independent of Reynolds number or
beginning of the transition, when there is flow velocity. Above a certain critical
one, is dependent on the solvent as well as concentration, the transition zone does not
on drag reducer concentration. Tiiu.9,fOZ appear mu ~ -1.-~lnT.*
~~b=.&VW .=e_..- full
mnuima g~e~ f~orn

*3 CP Gil. the
begins at
+Fsn~i~iQp-
..-.
ZO~e
turbulence suppression to maximum drag
PE 3691 JAIME A. LESCARBOU
reduction gradually without any inflection v- mean flow velocity, ftlsec
points. The critical concentration is a
function of solvent viscosity, It decreases AP - pressure drop in the line without
as solvent viscosity increases. Below the CDR, psi
critical concentration, where there is a
transition region, its beginning and end have APP - pressure drop in the line with polymer
been shown to be dependent on both concen- in crude, psi
tration and solvent viscosity.
v- fluid kinematic viscosity, consistent
For ease of operation it is desirable to units
operate above the critical concentration. ...
From the economic standpoint, however, it is Vppm - voiume parts per million; voiumes of
preferable to operate below Cc. The use of CDR per million volumes of solvent
CDR in viscous solvents should be most eco-
nomical in the region from NW 2100 to the
beginning of the drag reduction transition REFERENCES
since an early maximum drag reduction is
obtained. When operating below Cc, however, 1. Lescarboura, J. A., Culter, J. D., and
care must be taken not to overshoot into that Wahl, H. A., llDragReduction with a
portion of the transition region where drag Polymeric Additive in Crude Oil Pipe-
reduction is poor. lines,” ~r. —. J., 11 (1971)

ACKNOWLEDGMENT 2. Patterson, G. K., Zakin, J. L., Snd


Rodriguez, J. M., “Drag Reduction,”
The authors thank the Management of I&EC, Q, 22-30 (1969)
Continental Oil Company for permission to
publish this paper and acknowledge the in- 3. Rodriguez, J. M., Zakin, J. L., and
valuable comments and suggestions of co- Patterson, G. K., “Correlation of Drag
workers. Special thanks are due the Reduction with Modified Deborah Number
Management and personnel of Continental Pipe- for Dilute Polymer Solutions,” Trans.
line Company for their assistance in running Sot. Pet. Engr., 240, 325-332 (1967)
the pipeline tests.
L
T. u“..&b,
T) ~. , ----
Tk-trlcma “.
and Mntmiar
--..”---- , .A.
. . R..
-. ,
NOMENCLATURE “Turbulent Flow of Non-Newtonian Systems,”
A.I.Ch.E. J., ~, 189-204 (1959)
c - CDR concentration in volume ppm
5. Liaw, Gin Chain, “The Effect of Polymer
D - internal pipe diameter, inches Structure on Drag Reduction in Non-
Polar Solvents,” Ph.D. Thesis, University
D.R. - drag reduction of Missouri at Rolls (1968)

N* - Reynolds number, dimensionless,


given by $
TABLE 1 - DRAG REDUCTION TESTS RUN ON CDR AT VARIOUS VISCOSITIES

Nominal Pipe Viscosity


Solvent Size, Inch Cp CDR Concentration, VPPM

Diesel Oil 1 3 230, 690, 1150, 2300, 5000

Crude Oil 1 3 300, 1520

Crude Oil 1 5 300, 920, 1520

Wood Treating Fluid 1 8 1520, 3040

80 Pale/Diesel 1 12 230, 1150

80 Pale/Diesel 1 19 230, 1150

400 Pale/Diesel 1 40 460, 1150

400 Pale/Diesel 1 48 460

Crude 8 3 150, 230, 300, 480, 600, 1000

Wood Treating Fluid 10 9 1470

Crude 12 5 250
40 I 1 I I I I I I
CMICJ=NTRATION
1
I
+ 250 VPPM IN 12’’ CRUOE LINE ‘“”’;;~’;;pj-”-

D
+
o
250
1470
vPPM
vPPM
IN 12’’ CRUDE LINE
IN 10 “ PRODUC7LIN 36 01470 VPPM IN IO” PROOUC7 LINE d

/1
25 ■ ● 150 VPPM IN e“CRUOE LINE
● 150 VPPM IN 8“ CRUDE LINE
z A 230 VPPM IN ,, - - A 230 VPPM IN 8“’ ,, ,
o 32 , 480v PPM IN 8“ ‘, ‘- 1000
+ ■ 480 VPPM IN ., .,
0 / 1
20 /0 0 300 VPPM IN .( ,, ,, 0 300VPPM IN 8“ ,. ,, 1

F’
:

A 600 VPPM IN H ,$ I, A 600 VPPM IN e“ I ,,


A z 28
2 A ❑ 1000 vP~ IN ,, ,, ,, 0 Cl 1000 VPPMIN C!’” s ,, //:- .s0
w o +
F
: Is ~ 24
a
/ 0

‘o /
a u
n a

g 10 / w 20
Id 2
o C/ 0
K ~ 16
w
Ls I&l
v
: 12
L 1
I I I I I I I
o
) 40 60 80 100 200 400 600
REYNOLDS NUMBER IN THOUSANDS

Fig. 1 - Performance of CDR vs Reynolds


number for various concentrations in
lines Of 8 to 12 in. in diameter.
I 2345678
FLOW VELOCITY: FT./SEC.

Fig. 2 Orag reduction obtained


in f eld tests compared to
pred cted drag reduct [on.

30
25 TURBULENT
FLOW-l c?
nn E 4
5.0

_ 3.G
u)
~- 10
‘. 2.0 0
a. a
a n 8
u
a
6
(%
(n 5
u 0.0 u
a a
2 0. 4“
WY
: 0.5
a
3
a
2.5
0.3
2
0.2

,.“=
1.5

I !
0. I
4 567890
0.2 0.4 0.6 1.0 23 579 I 1.5 2 2.5 3
MASS FLOWRATE, W,LB/SEC.
MASS FLOWRATE,W, LB/SEC.

Fig. 3 - Raw data for 1,150 vppm CDR in Fig. 4 - Raw data for 1,150 vppm COR
12 cp mix of 80 pale and diesel oils at in 40 cp mix of 400 pale and diesel
800F in l-in. pipe. oils at 80°F in l-in. pipe.
“&
L
s
m.
O(JI
cm
0 .-
0+
.—

E(!I
D3
00
.-
+!-

F’--lT
06
-DY//’- >

L
ac
~ .-

m
l.,ll. ~i!! !- .—
L

0 0 0 0
0 : F1 N
In

NO11911Q3M 9V80 lN33M3d

Nou9na3tJ 9VM01N33M3d

z ‘?. “\:_
N&+,
\ o
u-l
, .-
0

c
.—
.-m
IA

XOOQ+Z

You might also like