You are on page 1of 9

.

6iik6 m

SocietyofPetroleumEnginears

SPE 37084

Horizontal Well Length: Drill Short or Long Wells?


Samir E1-Sayed and Bill Beckie, Wascana Energy Inc.

Copyright 16SS, Sm?mlyof Petroleum Engmwr8, Inc


face together with the oil water interfaee being close,
This paper was prepared for prwentawx al Ihe 1996 SPE Inlernattonal Conference cm premature conning takes place close to the heal of the well.
Hoozontal W*II Ta&moldgy held In CalgaV. C@nnda, 1&20 Novemba 19%
Once the water table establishes communication to the
This ~ wa$ soticd for prwentabon by ●n SPE Program Cc+-n”tIW following rWI.W of
Inlofmawan cantamad m ● ●bstmct submted by IIW Duthor(:) Contwws of the paper, w wellbore, say, 300m. from the heal the well is reduced to
prw.nlea nw~ no! Own rWt@w@d by the SWIety or Petroleum EngirNws and IN. subject to
GYrrocl!on by ffm ●uttwr(s) Tlw malwul, w prwbntod, d.ms not newswr!ly rofkct any
300m long well as far as oil production is ecmeemed. Most of
POSIIIIXIof the Sc&ty of PUroleum EWnwsrs, Its ofknrs, or memhrs Papers presented at the drawdown will be satisfied by the less viscous water
SPE m..tngs wc subjul 10 publtcsoan r.vmw by Edtorial Commdwes of the SocMy of
Petroleum Ertgnwrs P.nmsston to copy IS reslncted to ●n ●bstrscl of not mom than 303 coming from the infinite acting aquifer, and oil production,
words Illustrabws may not M copmd Th. .bstracl should conlmn Conspicuous
aCKnOwlOdgMWM0[ where and by whom tfw pspar was prasenled ‘$+YIleLlbrarlsn, SPE P O
from the pump intake at the heel to the water entry point, will
Box B33836 Richardson TX 75063 -363S U S A fax 01.214-952-9435 quickty vanish.
The above diagnosis led to two corrective approaches: 1) try
to shut-off the water cone by injeeting water shut-off material,
Abstract and 2) try to isolate the suspected area of the water cone
In a previous paper Ref. 1 it was mentioned that the reserves mechanically.
of horizontal wells of heavy oil over water is not linearly
related to the well length. A series of workovers were
performed to verifj’ this ecmclusion. New wells were Water Shut-off Treatments
especially designed to account for the non-linear relationship For a long time the oil industry has been looking for the ideal,
between reserves and length for wells longer than the cheap, reversible, non-damaging anti-conning or water shut-
optimum for reserves per meter. This paper is an update with off material.
regard to the technique of segmenting the well and producing Now if we can find a smart fluid that when hot it behaves just
different segments separately. The main criteria for the like water, preferentially displaces water and as it cools off it
success of the technique is to increase the reserve of the well becomes less mobile ( a solid ); it would be ideal for use as a
while maintaining a relatively high oil rate to maximize water shut-off material. If with such material the well inflow
revenue. is redueed below economic values it is easy to restore some
Details of a new water shut-off method specifically suitable to inflow. Hot water injection into the well will heat up the well
water conning is described in detail. bore and sand-face rendering the materiaJ mobile again. Once
the well is pumped some of the materiat will be produced and
the productivity of the well is improved.
Introduction Wax, yes normal problem causing paraflh wax, was the
The Winter field ( Fig. 1) has been developed utilizing material of choice. A refinery supplied two samples of
horizontal wells to produce heavy oil ( 13.7 API oil of 2800 paraftln wax, one of them has a melting point of 38-40
ep. viscosity) in a 12 m ( 37 ft.) thick Cummings sand degrees C (100-102 F’)was considered perkt for the Winter
underlain by a very strong aquifer. Ref. 1 showed that the Field of reservoir temperature of 26 degree C (79 F). When
reserve per well for the first six phases ( each year program is the wax is heated to 60-90 degree C (140-195 F) it is very
called a phase) is the same despite the fact that the average fluid with a viscosity much closer to water than to the 3000
well length for the different phases varied from 574m. (1883 cp. oil.
ft.) to 1111 m. ( 3645 ft.). It also showed that for over 250 When a well waters out it is very diffkult to induce a draw
horizontal wells, producing oil over water reservoirs in down even with a high rate of production. If we inject water
Saskatchewan, the reserves per well is not linearly related to into the well, it takes it on vacuum. In other words wellbore
the well length once the well length is longer than 500 m. flow resistance is reduced dramatically. This was another
(1640 ft.). Both of the above lead to assume that for a long reason to choose hot wax
well ( longer than a critical value around the 500 m ) the Here is the concept, if one cannot induce a pressure by
nature of the pressure distribution in the wellbore and sand injecting water into a watered-out well, this is because the

423
2

water immediately finds the water saturated cone and invades Segmenting the Well by ECP
it, the flow capacity of the cone to water is very high. Now if It was mentioned in Ref. 1 that the use of ECP and scab line]
we inject hot wax ( closer to water than oil in viscous proved effective in isolating the part of the production linel
properties) it will preferentially invade the cone; if we let it that is producing nothing but water.
cool-cIffthe cone is sealed and no more water can access the It was suggested that good engineering practice would sugges
well bore from the wax treated zone. to divide the well into at least two segments utilizing ECI
The idea was acceptable ( theoretically) but it has to be tested and scab-liner and produce only a portion of the well at tht
in the field to prove it practically. toe; thus the rest of the well is at Virgin condition. Thi
Winter well 2C6-32 was drilled in 1994, by April 1995 the oil procedure eliminates the need of costly sand cleaning prior t(
cut was less than 80/0 after producing only 3000m3. 7m3 of the ability to move or pull out the liner. As that portion i!
hot wax were injeded into the well after the well was close to being watered out the scab liner could be pulled ou
preheated with the injection of hot water. The well was shut to the next ECP or out of the well. Having no production frorr
in for two days then put back on production. The oil cut the portion of the well blanked by the scab liner means m
increased to more than so~o and the calendar day oil rate sand entry to the liner and ease of pulling the scab liner
increased form less than 9m3/d to over 20m3/d as shown in Another operating practice would be to wait until the firs
Fig. 2, The effect of the treatment on the reserves is more part waters out, pulling the scab liner to the next ECP am
astounding as can be seen in Fig. 3 the oil cut oil cum plot placing a bridge plug at the original ECP to block any irrflov
indicates that before the treatment the expected reserves at a from the watered out portion of the well. Fig. 4 shows :
2% economic oil cut would be about 10,000 m3; after the wellbore schematic of well 3D16-6, it has 2 ECP’S; initia
treatment the well cumulative production exceeded 11,000 completion included a scab liner set at the deepest ECP; i.e
m3 and the estimated reserves would be close to 20,000 m3. the liner was blanked with scab liner from the heel to tht
This is at least double the reserve without the treatment. deepest ECP. Only the toe, one third of the well wa!
The cost of this treatment in April 1995 was $7,000.00 , produced. The well produced in this completion from Mrq
currentty such treatment could cost as littte as $3,000. The 1995 until April 1996 at which time the well has produ=
success of this well proved that wax treatment concept is 14,000 m3 of oil and the oil cut was around 20V0.The punq
technically valid and yields a very profitable outcome. needed to be changed with a bigger pump , it was decided a
Other wells were treated with different degrees of success; that time to pull out the scab liner from the well to access tht
without a single deterioration of well performance. In other virgin part of the well and increase oil rate. Over th{
words, treatments performed to date would either improve the following three month the well maintained its oil cut at abou
well performance or does not change it at all. This could 20% while producing progressively higher oil rates as th(
imply that the volumes of wax in the treatments were on the pump speed is gradually increased. 3600 m3 were produce[
small side. This could be very true for the following reasons: for a cum. total of 17600 m3 in 14 month, that’s an averag(
1- One does not want to shut the well inflow with a material rate of over 40m3/CD. The last month production o
known in the oil industry as trouble. 2- It is thought that there 72.3m3/CD at an oil cut of 20’XO speaks very favorably of tht
could be multi cones in the wellbore, with different inftow proposed completion practice. Fig. 6 shows the reserw
capacities, therefore, multi small treatments would be more estimate of that well. It is clear thatremoving the scab line]
successfid than a single large treatment. Unfortunately, no arrested the decline of the oil cut, actually it increased the oi.
more than two treatments were given to the same well. cut. It is too early to quanti& the effect of producing the entir(
The treatment proved itself again during the 1996 drilling wellbore on the reserves; the next few month will show tht
program. One of the newly drilled wells produced at a 100%’o rest trend of the oil cut. The reserves of this well will ( veq
water. A tracer survey was run and proved communication conservatively) exceed 30,000 m3 and could reach 50,000 m:
with an old well. A treatment of 2m3 of wax did not improve which is much higher than a typical well in such reservoir. I
the performance. A second treatment of 4m3 wax resulted in would have been more conclusive that the propose(
the oil cut increasing to 80% and the oil rate to be higher than completion is better, if the old portion of the well was isolat~
20 m3/d; this was a remarkable improvement for a total cost and only the virgin part was producd, however, the results St
of less than $6,000. far are veV encouraging.
The concept of the wax treatment is presented here together
with the proof of its validity to encourage the industry to t~ it Considering that the difference in cost of drilling a 600m
and to challenge everyone to work hard on perfecting the long horizontal or a 1600m horizontal ( both with cementw
practice. intermediate) is relatively small most companies opted fol
Other materials, such as asphalt, sulfur, .etc., that could be drilling longer wells. Unfortunately, longer wells did no)
cheaper or more suitable could be used. always produce more oil. Now with very little added cost ~
less than $40,000) for ECP and scab liner these longer well!
could be insured to yield much more reserves at higher rate!

424
3

than the shorter wells.

Conclusions
1- Further evidence to the non-linear relationship betwwn
reserves and well length were shown, uneven water advance
and breakthrough is the cause of this non linear relationship.

2- Hot wax as a water shut of material is technically,


operationally and economically attractive.

3- The principle of segmenting the well into more than one


portion, by ECP and scab liner, and producing them
separately or in different sequence is valid and yields a higher
reserves and rate per dollar spent.

Acknowledgment
The authors would like to acknowledge the contributions of
many people in Wascana Energy Inc. Field staff for the
successfid implementation of the projeet. Technieal statT for
very stimulating discussions and suggestions. Management
for approval of cmying experimental tasks on a very
profitable field and for having the foresight to allow
publishing this paper.

References
1. Samir E1-Sayed,and Bill Beekie: “ Evolution of Production
Practice and Reservoir Management in the Winter Field”.
47th Annual Technical Meeting of CIM, Calgary, Alberta,
June 10-12, 1996.

425
+

II
FIG. 2 Wl_NTER WELL 2C6-32 PRODUCTION HISTORY
1000 100000

10000

100

1000 2“

10

100

1 10
94.5 95 95,5 96 96,5 97
TIME
0
m

<

v)

<
L >

0
0 0
0

428
FIG. 4 WINTER WELL 3D1 6-6 TRAJECTORY
0 300

@~m !$3Mm39
GOOC

. 0 200

‘$ 4af9 FAIR

100

POOR
0

-40 -1oo
ECP ECP

-50 -200
500 1000 1500 2000 2500
MEASURED LENGTH, M ECP1 1270-1300 ECP2 1670-1700
— TRAJECTORY , PAY — SLOTS
UN w-no

WI

ml

t+

<

+
FIG. 6 WINTER WELL 3D1 6-6 RESERVE ESTIMATE

T \

0
o

\
\

1
0 10 20 30 40 50
Thousands
OCUM, M3