Metallurgical American Institute of Mining.

Transactions of the American Institute of Mining, Metallurgical, and Petroleum Engineers, Incorporated online

. (page 47 of 60)
Online LibraryMetallurgical American Institute of MiningTransactions of the American Institute of Mining, Metallurgical, and Petroleum Engineers, Incorporated → online text (page 47 of 60)
Font size
QR-code for this ebook


No. 3 is a composite of the samples obtained from the experiment shown in
Table 1. In these experiments, the material was heated from 6 to 8 hr.
to obtain all the water. The residue is composed of all the oil boiling
over 105° C, fine particles of shale, sand, and limestone, and a few
ciystals of salt and pjrrite, flakes of mica, etc.

Tablb 2. — Distillation Results



Number of Sami>Ie


OU Distimng below
105«»C..» Per Cent.


Water. Per Cent.


Reeidue and Air,
Per Cent.


1
2
3


10.5

10.0

6.3


66.0
66.5
66.0


24.5
24.5
21.0
7.0 air and loss.



* Distillation runs under 105"* C. to detennine the per cent, of water in B. S.

Field Investigation on Emulsified Oil

Oil accumulated at the bottom of a well enters the mechanical parts
through the perforated tubing. It is drawn, through the standing valve,
into the working barrel, when the ball is raised from its seat, by the
vacuum create in the up-stroke. On the downnstroke, the working
valve opens and allows the fluid to pass into the tubing. It is then lifted
about 3 ft. by each successive up-etroke until it reaches the surface and
flows from the well through a lead line to a receiving tank (see Fig. 11).



Digitized by VjOOQIC



442



INVESTIGATIONS CONCEBNINQ OIL-WATER EMULSION




A,djan«r
Adjuncr Board



WalklBf Beam



Pollth Bod



filoedcr for Sampling /



8tafftn|-box



£rCnn=il7



Load Lfnoto Tusk



a=^'



7^3 Bl| Floor






Caalof



Toblng



Backer Bod
Working Barrel

Working Talfa
Bunding Valr*
Perforation



Fig. 11. — Diagram of working pabts of pboducing oil wbll.



Digitized by VjOOQIC



ALEX. W. McCOT, H. R. SfflDBL AND B. A. TRAQER 443

When a sample is to be taken, the check valve on the tank lead line is
closed so that the flmd coming from the tubing is not affected by the back
pressure from the tank. The valve on the short lead line is opened and
the sample is caught in a bucket (see Fig. 12) and allowed to settle
for about 30 sec. The vertical cylinder is then placed in the guides of the
bucket, separating a representative column measuring about 90 or 100
c.c. of flmd, which is drawn off through the small valve in the bottom of the
bucket. The sample is then centrifuged, the oil, B. S., and water separat-
ing out. By plotting the results of such samples, taken every 10 min.,



Fig. 12. — Diagram op sample bucket.

many irregular conditions have been noted. All producing oil wells do
not perform in the same manner; some of the different conditions are
shown by the accompanying charts.

Kg. 13 shows the graph of a well producing a high percentage of water;
there is no emulsion in the fluid. The oil and water remains in about the
same ratio over a period of 8 hr. The well was pumping about 200 bbl.
per day when this test was made. Fig. 14 also shows a graph of a well
making a large percentage of water; there is, however, a small percentage
of emulsion plotted through the entire time of the test. Fig. 15 represents
a well producing a large percentage of water and a comparatively large
percentage of B. S.

Fig. 16 shows the graph of a weU producing a low percentage of water,
a high percentage of oil, and a comparatively high percentage of emulsion.



Digitized by VjOOQIC



444



nnrSSTIOATIONS CONCSBNIMa Onr-WATBB BMUL8I0N



Fig. 17 illustrates a well that has just pumped-ofF a head of water, after
which the percentage of oil suddenly rose and remained about the same for
some time then gradually decreased. At the time of the increase in the



100
90

80

TO

00

|50

80
20





/


V a




A


k>.M<y-e-a^


jCWr-q






\/


^j


V




^


\


.. \


,K-»-0-«"»-«


o


v































B,8.






































A


















A.


^


A^




V


..../


^j




9




V


' '6




^




>»W







9.-00



10:00 11.-00 . 12M 1HX> 2:00 8:00 4.-00 6.H)0

AJyf. Time PJtf.

Fio. 13. — Gbaph of wbll pboduoing moH pbbcbntaqb or watbb.



6:00



100
90

80
70
00
S 50
^40
80
20
10























AvvA


o^'V^N


^'^^


f^\vA .


^Owv^ JO-^


v^"-.^


^^






a« 1


f xr \


V*^ ^


^ n-v^/


a


\j*«^


s/


































„.. ^'


.GEND




B 8 —




















1






















a ,




An y


^ n.




^


A/*






NW


w


'A^


W^


vW


i^vs/"


'v^






o-oo-a-iH


i»*s2:9.«-fli


KQ,-,<^.0-ft-<


■Q-O-O-O-Q^


hftia<ft-fi-ri


rfl-g^Q-g^







8.-00



9K)0



4«0



6:00



lOHN) 11.-00 12KN) 1:00 8H)0 8.1)0

Ajyi. Time PJH

Fig. 14. — Graph op wsll PRODOt^iNa labgs pbbcbmtagb op watbb and small

PBBCINTAGB OP BMUUIION.

percentage of oil| the percentage of B. S. rose and continued to increase
at about the same rate as the oil dropped later. The fluid came from
the tubing much more slowly as pumping progressed* Fig. 18 shows a



Digitized by VjOOQIC



AIiBZ. W. IfoCOT, H. B. 8HID8L AKD B. ▲. TBAGBR



446



greater increase in the percentage of B. S. The amount of fluid produced
during the last hour was 25 per cent, of the amount pumped during the
first hour of the test.




8:00



9H)0



10:00



11.00



4K)0



6:00



12:00 1.-00 2:00 8.-00

Time PJi.

Fig. 15. — Graph op wsll pboducing labgs psbobntaqb op watbb and compara-
tivblt labob psbcbntagb op b. &




8:30



9:30



10:30



11:30



12.-00 1:30

Time



2:30



6:80



8:30 4:80

PJd.

Fig. 16. — Gbaph op wbll pboditcino low pbbcbntaqb op watbb, moH pbbcbntagb

OP OIL, AND COlfPABATIVBLT HIGH PBBCBNTAQB OP BMULSION.

Fig. 19 represents the behavior of a well pumping all water for several
hours; after this was exhausted the percentage of oil increased rapidly.
The water pumped during the early hours of the test was the water that



Digitized by VjOOQIC



446



INYlBSTiaATIONS CONCEBNIKO OIL-WATBR EMULSION



separated out of the fluid behind the tubing. As the head of fluid was
reduced, the oil finally came into the tubing. The percentage of B. S.
was practically nil.




8.-00



4:00



9KK) 10:00 IIKX) 12.-00 1:00 2.-00 8.-00

AM. Time PJd.

Fig. 17. — Graph of wbll that has just pumped off head of water.



5:00




8:00



1:00



iM



6:00



lOHW U.*00 12KK) 1:00 2.-00 8.*00

AM. Time PM,

Fio. 18. — Similar to Fig. 17 but showing greater increase in percentage of b. s.

Fig. 20 is the graph of a well that pumped only 6 hr. during the day.
The fluid in the tubing was composed entirely^of oil and B. S., which did
not separate out. The high percentage of water following was probably



Digitized by VjOOQIC



ALEX. W. McCOY, H. R. SHIDEL AND E. A. TRAGER



447



the accumulation of water behind the tubing, which passed into the hole
during the shutdown. When the water was about exhausted, the per-
centages of oil and B. S. were about the same as when the well was started.



100












\




















1/.




<


.. '■E«END


80












\




J




B.8.




.70














A f.






.00












i








60
































J




















/








W












r ,








10













/









8:00



9:00



10:00



11:00



4:00



12K)0 IKW 2H)0 8.-00

Time PJtf.

Fio. 10. — ^BaBAvioB or wbll pimpiNa all watui vob bevebal houbs.



6K)0



100








*yO.






on

B.8.
W«tcr


' 1

LEGEND












l\






1


80

70




o-«N


SA-o^




?




p*


-^






Ui














c


























.




UU

in














1






«A






.^


%












QA


1 o-o^




r^^\


.-o-V^


















10







X


J^


/ * V


W^xv.


^t><>^







8:00 9H)0 10.*00 11.-00 12:00 1:00 &00 S.-OO

AJyf. Time PM,

Fio. 20. — Graph or well ptticpbd 6 hr. a dat.



4:00



6.-00



A well of this order, when pumping continually, would not show such an
erratic condition, as there would be no chance for a large head of water to
collect in the hole.



Digitized by VjOOQIC



448



INYBSTIGATIONS CONCERNING OIL-WATBR BMULSION



Fig. 21 shows the action of a well producing an extremely high per^
centage of B. S. with small percentages of oil and water. The well was
shut down for 3^ lu*- When pumping was started again, it first pro-




s'^



10:90



8:a0



4:90



6:80



6:80



11:90 12:90 1:30 2:90

AJtf. Time PJ^

Fig. 21. — Graph op wsll pbodxtgino bztbbmblt moH PSBCBNTAoa op b. s. Ain>

SMALL PBRCBNTAQS8 OP OIL AND WA1VB.




7:00



8.-00



9:00 lOKK) ll.*00 12.*00 1.-00 2.*00 8:00

A.M. Time PJd.

Fig. 22. — Graph op wbll pumping largh amount op PLum at start.



4M



duced the same percentage of B. S. as before. Within 40 min., though,
the high percentage of B. S. dropped to nothing with a great increase of
water. At this time, the percentage of oil increased some. In 2 hr.



Digitized by VjOOQIC



ALEK. W. IfoCOY, H. B. SHIDEL AND B. A. TBAGER



449



the water was exhausted and the well started to operate with a more i^^-
lar flow. The influx of water was due to the accumulation of water
behind the tubing, during the shutdown.







Q




> 1

lu&oa


ni8 1


^


>




1 e 1










t*




<:






\








>






c






>




t^


C


>>








<


>


©-

«


Of


\













o






"V










"r




<




9


>










«l


^


6
6




<


>












c.


\




a


••-3>


' 1


S






<


>>


J
















— 5




S2?g -




















!


'




1 i^l\






-


i-






«» f 8






O
Z


1


'




«n 2^ ' o






Ul








^ OC 8 -1 "J






CJ








5 ^ i-^5






uJ


1


'




5 r »S::


















o«^


(












'






et to fill bi
of bleeder -
ectlon
pumping -
action
uted this
iding (all










'




stroh
Sire
Conn
Time
Prod
per s
e Stat










,


...^




1 u












-t>~


_^












+— *^








J


-r^










N


^ ?




J














I.


<5


f^






»a]


dmnj


u«i9








% i



6



^-S



8 S S ^ 8 S 9



O OS



Fig. 22 shows a well that was pimiping a large amount of fluid at the
start, the greater percentage of which was oil. As pumping continued,
the percentage of oil decreased and that of the B. S. increased. There was
a perceptible rise in the percentage of water too. At the end of the test.



TOL. VKS, — ^29.



Digitized by VjOOQIC



450 INVBSnOATIONS CONCBBNING OIIr-WATBR BMTJL8I0N

the weU was producing about one-third the amount ci fluid that it did at
first.

Fig. 23 illustrates the performance of a well from the time the pump
was started until it was stopped. At first, the well produced a good per-
centage of oil, which rapidly decreased to nothing. This percentage
represents the oil that had settled in the tubing during the 12-hr. shut-
down previous to starting. After this was pumped out, the Well produced
nothing but water for several hours; oil was then noted. By this time,
the water in the tubing, behind the tubing, and some that had probably
been backed up in the oil sand had been pumped off. The oil that had
separated out to the top of the water outside of the tubing was being
pumped. Less fluid was pumped at this time than when only water was
being produced. During the next 2 hr., the well was producing at about
the same rate. Undoubtedly, some of this fluid was partly accumulated
during the shutdown and partly coming into the well during the pumping.
After this, a decided drop in the amount of production was noted, with a
decrease in percentage of wat^r and an increase in the percentage of oil.
The fluid from this time on was probably coming direct from the sand.

Fig. 24 shows the performance of an oil well over a period of 33 hr.
During that time several experiments were tried and the results noted.
At 9:30 A.M., the pump was started, showing a high percentage of oil,
which dropped materially within 10 min. ; this was the oil that had settled
out during the shutdown. The well produced a large percentage of B. S.
for 1^ hr. From 11:10 a.m. to 11:30 a.m., there was a large increase
in the percentage of oil followed by a sudden and greater drop than before.
This was the oil that had risen to the top of the fluid around the tubing
during the period of shutdown. From 12 : 20 p.m. until 6 : 20 p.m., the well
was pumping about as fast as oil and water were coming from sand; gradu-
ally a big increase of B. S. was noted. From8:30p.M.imtil 11:30p.m., the
well was shut down. When it started to pump again, the first test reported
a high percentage of oil, which immediately dropped and the well produced
fluid in about the same percentages as that before the shutdown until
1:50 A.M. At this time an increase in the percentage of oil was noted,
followed immediately by the correspondingly large increase of water.

The fluctuations following are the results of quantities of oil and water
getting into the working barrel in separate bodies. From 4 until 8
o'clock, the well was operating with about an average percentage of oil,
B. S. and water. The well was shut down from 8:40 am. until 9:20
A.M. and from then until 11 :20 a.m., it operated about the same as it did
from 4 a.m. until 8 a.m. At this time there was a noted increase of water.
The fluctuation of oil and water was due to the separation around the
tubing during the shutdown. From 12:10 until 5:30, the well operated
with a regular, or normal, flow.



Digitized by VjOOQIC



ALEX. W. licCOT, H. R. 8HIDBL AND B. A. TRAGER



451




Digitized by VjOOQIC



452 INVESTIGATIONS CONCEBNING OIL-WATBR BICULSION

An interesting feature of the curve is the fact that after the well had
been standing idle for a while it produced with a regular and then an
irregular flow. The average production of the well was about 40 per
cent, oil, 50 per cent. B. S., and 10 per cent, water. This mixture in the
tubing did not settle out readily during the shutdown so that when the
well started producing again it pumped approximately this ratio of fluid
for the first 2 hr., clearing the tubing of the fluid left there during the idle
period. While the well was not operating, oil and water filled up behind
the tubing from the sand, which was pumped largely as clear oil and water
with comparatively little B. S. after the fluid in the tubing was exhausted.
When the accumulated head behind the tubing was reduced, the normal
production returned. This series of conditions followed each shutdown.
The B. S. content is only important when there is no large head of fluid
behind the tubing. As shown by the graph, the B. S. content increased
materially from 6 imtil 8 p.m., after the normal flow had gone on for 5 hr.
Pumping was no doubt going on at a faster rate than the production
from the sand so that gas, air, or voids in the fluid colimm were admitted
and the oil emulsified to a greater amoimt.

Referring again to Fig. 23, the following computation was made to
determine the amoimt of fluid at different times during the test. During
the first hour (10: 10 to 11 : 10 a.m.) of pumping the average of oil content
was 83 per cent. The beam was making 15 strokes per minute and it
took two strokes to fill a bucket of 7 qt. (6.61.) capacity. Consequently,
this is equal to 786 gal. (2975 1.) per hour or about 18.7 bbl. per hour. If
the oil content is 83 per cent, of this fluid, the amoimt of oil pumped dur-
ing this period is 15.62 bbl. For the next Q^i hr., the weU produced prac-
tically no oil.

From 5 : 20 to 7 : 20 p.m., the weU was producing about 20 per cent. oil.
The b eam wasmaking 15 strokes per minute and it took four strokes to
fill a bucket of 7 qt. capacity. This is equal to 393.6 gal. (1490 1.) or
9.37 bbl. of fluid per hour. If the oil content is 20 per cent, of this fluid,
the amount of oil pumped during this period is 3.75 barrels.

From 7 : 20 to 10 : 00 p.m., the well was producing about 25 per cent. oil.
The beam was making 15 strokes per minute and it took six strokes to
fill a bucket of 7 qt. capacity. This amoimt is equal to 262.2 gal. (992 1.)
or 6.24 bbl. of fluid per hour. If the oil content is 25 per cent, of this
fluid, the amoimt of oil pumped during this period is 4.16 barrels.

The total amount of oil produced is: 15.52 + 3.75 + 4.16 = 23.43
bbl. The total amount of B. S. and water produced is: 115.38 + 14.99
+ 12.48 = 146.04 bbl. The total amount of fluid is: 23.43 + 146.04
= 169.47 bbl. During the 12-hr. shutdown, the oil and water were
allowed to accumulate in the well. The tubing remained full of fluid
as it was when pumping was started. The fluid entering the weU filled
up behind the tubing.



Digitized by VjOOQIC



ALESJL W. IfcCOTi H. R. SHIDBL AND E. A. TBAGEB 453

The following computation shows the amount of time required for the
raising of the fluid at the bottom of the well to the top: Area of 34n.
tubing is 7.06 sq. in. Area of ^-in. rods is 0.44 sq. in. ; difference, 6.62 sq.
in. The number of cubic inches in 1 ft. of 3-in. (76-mm.) tubing is equal
to 6.62 by 12 in. or 79.44; then 2.91 ft. of 3-in. tubing contains 1 gal. of
fluid. The depth of the well, 2425 ft. (739 m.), divided by 2.91 is equal to
the number of gallons of fluid in the tubing, which is 833.3 gal. or 19.84 bbl.
Pumping at the rate of 18.71 bbl. per hour, the time that it would require

19 84

to &oapty the tubing would be loVr which is equal to 1.06 hr. or 1 hr.

4 min. It will be noted by the graph that the big influx of water came
in 1 hr. 10 min. after the well start^ to pump.

Referring to Fig. 24, the following computation has been made to show
the amoimt of time required in this well fbr the raising of the fluid from
the bottom to the top. The depth of the well (2475 ft.) divided by 2.91
is equal to the number of gallons of fluid in the tubing, which is 850 gal.,
or 20.2 bbl. Pumping at the rate of 8.71 bbl. per hour, the time required

20 2

to empty the tubing would be ^- which is equal to 2.32 hr. or 2 hr.

19 min. The well was pumped at the above rate beginning at 11 :30 p.m.
and the time required for the first big fluctuation to occur was 2 hr.

20 min. These figures give an idea of the time required to pump the
fluid from the tubing and the variation of the same in different wells.

The question naturally arises as to whether or not the separation of oil
and water while passing through the tubing is sufficient to cause a dis-
crepancy in the ratios of oil and water in each unit volume as it flows from
the bleeder, and the ratio of the fluids as they enter the perforations. In
other words, after the well has pimiped the full column of fluid in and be-
hind the tubing, are the proportions of oil and water at the bleeder the
same as they are in entering from the sand ? The rate of separation of oil
globules in a water colimm depends on the difference in the specific
gravity of the two liquids, the temperatiu^ of the same, and the size
of the globules. If the full column of fluid is lifted 2500 ft. (762 m.) in 1 or
2 hr., certainly that time is siifficient for considerable separation if the
fluid remains quiet. However, it has been noted by experiment that a
slight stirring will prevent any separation of the fluids, and since the
rods are constantly flapping through the fluid column, it seems that any
tendency to separate while pumping would be greatly if not altogether
reduced. Moreover, if the water from each unit volume should be con-
stantly descending to the next unit volume etc. all the way down the
colimm, the bleeder would still receive something of the same ratio,
only apparently at later time.

From a number of the ctuves, it will be noted that after a well has
been pumped for several hours, the ratios of oil, B. S., and water tend to



Digitized by VjOOQIC



454 INVESTIGATIONS CONCERNINQ OIL-WATER EMULSION

remain nearly constant, without large or rapid fluctuations. This may
continue for a long time, only after the fluid head behind the tubing has
been reduced. For that and the above reasons, we have considered the
ratios at the bleeder when the fluid head is once reduced to be the same
as the ratios of water and oil entering from the sand and have caUed this
the ''normal flow."

Conclusion

Permanent B. S. is an emulsion of very small water bubbles in oil
having a diameter generally less than 0.5 mm. The oil may be relatively
clean or it may contain variable amoimts of suspended matter. There are
generally a few air bubbles present.

The behavior of B. S. on heating may be used as an economically
important basis for division* into two groups. In the first group, the
water separates from the oil rapidly with a smaU amoimt of heating.
In the second group, the water can be removed only by distillation.

To form B. S., it is necessary to have present, in addition to oil and
water,^ either air, a gas, or voids in the continuity of the fluid, i.e., a break
in the fluid.

The percentages of oil, B. S., and water vary in the individual wells;
each well is a problem in itself.

A small steady amount of B. S. is probably due to bad valves and cups.
Percentages of B. S. are increased as the column of fluid aroimd the tubing
is exhausted; such a condition allows air to enter the working barrel or a
break to occur in the column of fluid. This condition is responsible for
large amounts of B. S. The bubbles of the different liquids and gases are
made smaller and consequently more stable by the whipping of the rods.

The maximmn efficiency of a pumping weU, which is producing both
water and oil, is obtained when the fluid level is kept above the perforated
tubing and below the point where the accumulated head of water would
stop the flow of oil into the hole, and when the fluid is pumped at the same
rate that it comes from the sand. Such conditions can only be dete>
mined by a special test of the individual well.

DISCUSSION

A. W. Ambrose, Washington, D. C. — ^Did you make any analysis of
the amount of emulsion at the well and after you flowed it through a
lead line to the storage tank?

E. A. Tracer. — B. S. can be formed in passing through a lead line by
the friction due to the roughness of the pipe and the irregularities at the
joints.

R. W. MooRE. — Did you find the percentage of water to be limited to
the percentage of oil in the emulsion which formed?



Digitized by VjOOQIC



DISCUSSION . 455

E. A. Trager. — Yes, the percentage is about 67 per cent, water and 23
per cent. oil.

R. W. MooBE. — If you added more water would the emulsion be
permanent?

E. A. Trageb. — ^Yes, it would be permanent, but the excess water
would separate out.

The Chairman (C. W. Washburnb, New York, N. Y.>. — ^Did you
use hot or cold water in these experiments?

E. A. Tracer. — It makes no difference which is used. We tried to
determine whether the composition of B. S. formed in the presence of
excess water would differ from that formed in the presence of excess oil.
The percentage composition in each case appears to be the same.

F. G. CJoTTRBLL, Washington, D. C. — In electrical demulsification
experiments in the West a number of years ago, we f oimd no lower limit
to the size of globules in an emulsion that could be dealt with, and I
believe this has been borne out in the operation of the commercial plants
that grew out of this work and are in operation today. I am therefore
surprised at the results that Mr. Trager has secured, and am inclined to
think that he may not have applied the treatment in the same way,
because it was with those very fine emulsions that we were working in our
experiments.

E. A. Trager. — ^The chemical laboratory worked on this same sub-
ject and tried using a high voltage current to break down the emulsion,
but the results were not commercially practical for it was found necessary
to treat fine emulsions several times before they were completely broken
down.

P. G. Cottrell. — Do you know the details of the experiments — the
voltages and conditions?

Chairman Washburne. — I believe you used high voltages in your
experiments did you not, and alternating current?

P. G. Cottrell. — Yes.

Chairman Washbttrne.— It seems to me, since there is no doubt that
every globule must have its charge of static electricity, the smaller the
globule, the easier it would be moved by electrical currents and discharges.
The normal static charges will be of like kind and proportional to the



Online LibraryMetallurgical American Institute of MiningTransactions of the American Institute of Mining, Metallurgical, and Petroleum Engineers, Incorporated → online text (page 47 of 60)