Computer Simulation of the

Reverse-Circulation Well-Control
Procedure for Gas Kicks
Stefan Miska, * SPE, New Mexico Inst. of Mining & Technology; F.E. Beck, SPE, Arco Alaska Inc.;
and B.S. Murugappan, SPE, Dowell Schlumberger
A. computer simulator was and used to evaluate reverse circulation as a potential well-control procedure during
dnlhng. PotentIal advantages of reverse cIrculatIOn verified by the simulator include lower casing pressures and smaller cumulative
pit gains. In addition, kick fluids are removed from the well much faster than with conventional kill procedures. Several equipment
modifications will be required to implement the reverse-circulation kill procedure.
Introduction
Such conventional well-control methods as the driller's method or
the wait-and-weight method have gained widespread acceptance in
the drilling industry owing to their effectiveness and flexibility.
These methods circulate kick fluids out of and kill fluids into the
well following predetermined and controlled actions that require
keeping bottomhole pressure (BHP) constant to prevent additional
kick fluids from entering the well while maintaining annular pres-
sures below a maximum allowable pressure limit. Both of these
methods have proved effective for hydrostatically stabilizing a well,
and each has relative advantages. Recognized advantages of the
driller's method are that it is simple and can be applied immediate-
ly to begin well-control operations shortly after taking a kick. The
wait-and-weight method improves on the driller's method by short-
ening total circulation time and minimizing maximum annular pres-
sures. But this method results in additional calculations and a delay
in the implementation of the well-control procedure, during which
well pressures will rise because of an increase in kick fluid (gas).
The success of each method relies on a well-trained crew that fol-
lows specific procedures. These methods, however, were devel-
oped with the basic assumptions that the well can be safely shut
in and the kick fluids can be circulated out of the well without ex-
ceeding formation fracture or casing burst limits.
During the drilling of almost any well, a well-control situation
potentially may occur in which either of these assumed conditions
does not exist, precluding the use of a conventional well-control
procedure. One such situation commonly occurs when drilling be-
low drive or conductor casing; formation strength is insufficient
to shut in the well if gas is encountered (shallow gas problem), and
the well must flow through a diverter system. Another situation
occurs when a very large kick is taken during drilling of an inter-
mediate hole. While it may be possible to shut in the well initially,
the high casing-shoe pressures caused by gas expansion in the an-
nulus may be such that the shoe will fail under these pressures and
using a conventional procedure through full well circulation would
be unsafe. Typically, one would be forced to operate the choke to
avoid exceeding the maximum allowable casing pressure, which,
in turn, would allow additional kick fluids into the well.
Under these conditions, well-control operations generally become
much longer and the chance of failure increases greatly. Develop-
ment of a well-control method that can be applied in a controlled
and predetermined manner to allow circulation of large kicks out
of the well without breaking down the casing shoe or allowing ad-
ditional kick fluids to enter the well would be desirable.
One method that may prove useful in such situations is to reverse
circulate the kick out of the well by pumping down the annulus,
up the drillpipe, and out through a choke. Reverse circulation has
been used in workover operations because of the reduced fluid
volume required to kill the well. In a workover environment, fluid
losses through the perforations often can be tolerated or perfora-
'Now with The U. of Tulsa.
Copyright 1992 Society of Petroleum Engineers
SPE Drilling Engineering, December 1992
tions are already squeezed off and the workstring is open-ended
so plugging is not a concern. Objections to reverse circulation as
a well-control method during drilling are that it places excessive
backpressure on the formation, causing lost circulation, or that the
bit jets would become plugged soon after reverse circulation begins.
Goins,l Rhem,2 and Adams
3
addressed well-control procedures
under impending blowout conditions. Several authors
4
-
6
proposed
mathematical models to predict casing and drillpipe pressures while
the kick fluid is circulated out of the well. Miska et al. 7 recently
developed a technique for determining stabilized shut-in drillpipe
and casing pressures. We cannot, however, locate literature related
to the modeling of a reverse-circulation well-control procedure and
can find only limited references to reverse circulation as a well-
control procedure during drilling. Thus, upon review of the avail-
able literature, the validity of some of the objections to the use of
reverse circulation during drilling is questionable.
Mills 8 stated that the potential benefits of a reverse-circulation
well-control method would be that casing pressures would rapidly
diminish, higher pressures would be trapped in the drillstring, the
controlled gas expansion would be less critical than with a conven-
tional method, and long choke lines would no longer fill with gas
or gas-cut mud. Mills also stated that large kicks with weak for-
mations exposed may be handled more readily by reverse circula-
tion. Mills' analysis of the potential of reverse circulation as a
well-control method during drilling was qualitative only. The pur-
pose of our study was to develop a simple analytical tool, in the
form of a computer program, to analyze the fluid mechanics in-
volved in reverse circulating a gas kick out of a well during drill-
ing, to develop the basic procedures required to implement a reverse-
circulation kill, and to determine the equipment modifications or
additions required to perform reverse circulation.
9
The purpose
of this paper is to present the results of this study and to compare
a reverse-circulation well-control procedure to the driller's and wait-
and-weight methods under a variety of drilling conditions.
Theory
For the computer program, we assumed (1) steady-state fluid flow
in the well, (2) that the kick fluid was gas and occurred as a con-
tinuous slug, (3) that the real gas equation of state governs changes
in the kick volume, (4) that manipulation of the choke instantly con-
trolled (maintained constant) BHP, and (5) that the gas bubble had
a slip velocity of zero.
The computer program was written to track the position, geom-
etry, and pressure in the kick and in the well as it circulates or
reverse circulates upward. The basic premise of the simulator is
to operate the choke to maintain constant BHP. Well pressures,
p, then can be calculated from the known BHP through the follow-
ing relationships. In the annulus,
Pws=Pc+Phs-Paj' ................................ (1)
andpws=pch+Phsm+Pjm+Phso+Pjo ................... (2)
in the drillstring.
247


EOC

50 100 150 250 300 350
TIME (MINUTES)
Fig. 1-Casing pressure vs. time, reverse-circulation deep
well, 30-bbl kick.
The initial pressure in the gas bubble is the BHP, which can be
determined from a shut-in drillpipe pressure or from an assumed
formation pressure value for simulation purposes. The initial gas
bubble volume is the initial pit gain observed at the surface. The
program then repositions the gas in the well according to a known
volume pumped over a timestep (volume change equals rate times
change in time). The pressure in the bubble then becomes the pres-
sure in the drilling fluid at the lower gas/mud interface.
We then assumed that the pressure in the gas, p, is constant and
that the gas bubble volume, V, can be calculated from the change
in gas pressure by
V
new
=Pold Vo1d/Pnew' ............................... (3)
The subscripts "old" and "new" correspond to the respective
timesteps. The new volume and length of the kick then is deter-
mined for the new timestep with this method. Once the new bub-
ble length is established, the friction and hydrostatic pressure drop
across the bubble can be calculated by the Cullender-Smith method.
The change in gas bubble volume also will be seen as a flow rate
increase in the mud above the bubble. The frictional pressure drop
in the drilling fluid is calculated by Bourgoyne et al. 's 10 method.
The drillpipe and casing pressures are calculated to maintain con-
stant BHP.
The above calculations can be carried out successively to inves-
tigate either normal or reverse-circulation well pressure behavior.
With this tool, a procedure for reverse circulating a gas kick was
developed.
Reverse-Circulation Procedure
After many simulations, we concluded that a well-control proce-
dure using only reverse circulation is infeasible if the direction of
circulation remains constant throughout the procedure. The increase
248
TABLE 1-DEEP WELL DATA
Well data
Total vertical length, ft
Open hole length, ft
Hole size, in.
Casing
11314 x 10.88-in. size, ft
8
5
/s x 7.725-in. size, ft
Drillstring
4 x 3.47-in. drillpipe, ft
5 x 2V2-in. drill collars, ft
5V2 x 2V2-in. drill collars, ft
Bit nozzles, in.
Original mud weight, Ibm/gal
Kick size, bbl
Stabilized SIDPP, psi
Stabilized SICP, psi
Kill rate, gal/min
15,030
2,010
7.625
from top to 9,630
from 9,630 to 13,020
14,400
270
360
3x9/32
14.6
30
400
950
200
3500
i-" KILLMU ISCIRCUL DRILL PIPE
F
/
3000
"
i D
G ....
i'
cA

KILL IIUDENTER ANNULUS
I
KlL MUDREAC ESTHEBIl EOC
CHANGE N DlRECTIO OFMUDC RCULATION
500
A
B
E
/
J
Aa.c-D GAS IN DRILL PIPE
l
[).E GAS LEAVES DRILL PIPE
o
o 50 100 150
TIME (MINUTES)
250 300 350
Fig. 2-Drlllplpe pressure vs. time, reverse-circulation deep
well, 30-bbl kick.
in circulation pressure caused by drillstring friction is applied direct-
ly to the casing annulus and invariably exceeds the maximum al-
lowable casing pressure. This is true for both a reverse driller's-
or a reverse wait-and-weight-type procedure. Consequently, a proce-
dure was developed by which the kick is reverse circulated from
the well and kill-weight mud was circulated normally into the well.
We assumed in the following discussion that the well can be reverse
circulated or normally circulated through the choke manifold. The
method follows.
I. Reverse circulate the well by pumping fluid down the annu-
lus, up the drillstring, and through the choke.
2. Pump pressure initially will be shut-in casing pressure (SICP),
but reduce it to shut-in drillpipe pressure (SIDPP) once the kick
is displaced from the annulus. A pump pressure schedule can be
developed to accomplish this.
3. Once the kick is in the drillstring (adjust the surface choke
to maintain constant Pws)' hold pump pressure constant until the
kick is completely displaced from the drillstring.
4. The pump should be shut down and the well routed back to
a normal circulation path.
5. Kill-weight mud can be circulated into the well in a normal
direction following procedures of the driller's method.
The simplicity of this procedure can be appreciated by analyzing
Eq. I. BHP will remain constant as long as the right side of the
equation is constant. Pump pressure is controlled by choke manipu-
lation, hydrostatic pressure is constant as long as a single-density
fluid occupies the annulus, and annular friction will be constant
as long as flow rate and density are constant. The choke is useful
for regulating pressure as long as gas remains in the drillstring.
Thus, drillstring friction is not applied to the annulus until gas is
out of the well and the choke is no longer effective.
TABLE 2-MEDIUM-DEPTH WELL DATA
Well data
Total vertical length, ft
Open hole length, ft
Hole size, in.
Casing
13
3
/s x 12.515-in. size, ft
Drillstring
5 x 4.276-in. drillpipe, ft
5 x 3-in. heavyweight, ft
8 x 2s-in. drill collars, ft
Bit nozzles, in.
Original mud weight, Ibm/gal
Kick size, bbl
Stabilized SIDPP, psi
Stabilized SICP, psi
Circulating rate, gal/min
8,500
5,800
12%
top to 2,700
7,360
900
240
3x 12132
8.5
60
90
280
200
SPE Drilling Engineering, December 1992
TABLE 3-SHALLOW WELL DATA
Well data
Total vertical length, ft
Open hole length, ft
Hole size, in.
Casing
20x 18.73-in. size, ft
Drillstring
5 x 4.276-in. drillpipe, ft
6 x 3.5-in. drill collars, ft
8 x 3.5-in. drill collars, ft
10 x 3.5-in. drill collars, ft
Bit nozzles, in.
mud weight, Ibm/gal
Kick size, bbl
Stabilized SIDPP, psi
Stabilized SICP, psi
Kill rate, gal/min
3,105
2,095
17%
top to 1,010
2,445
120
180
360
3 x 15/32
9.5
15
150
600
300
Figs. 1 and 2 give the expected response of the well to the proce-
dure in terms of casing pressure and drillpipe pressure as functions
of time. Table 1 gives data used to generate these graphs.
Fig. 1" demonstrates the expected casing pressure response dur-
ing reverse circulation. Note that SICP is the maximum casing pres-
sure seen during the procedure and that the choke is manipulated
to maintain constant BHP. As Fig. 1 shows, pressure drops be-
tween Points 1 and 2 because hydrostatic pressure increases as the
kick circulates into the drillstring. The kick is entirely in the drill-
string at Point 2, and the casing pressure is constant from Points
2 to 3 as the kick circulates up the drillstring. The kick is entirely
out of the well at Point 3, and the pressure rapidly decreases from
Points 3 to 4 as the circulation direction becomes normal. The differ-
ence in pressure from Points 3 to 4 equals twice the friction pres-
sure of the annulus. From Points 4 to 5, kill-weight mud is pumped
down the drillstring. This mud reaches the bit at Point 5, and casing
pressure drops from Points 5 to 6 as a result of the increase in hydro-
static pressure caused by the kill-weight mud. At Point 6, the sum
of hydrostatic pressure and annular friction equals BHP, and casing
pressure is 0 psi before circulation is complete. This is an unstable
equilibrium, and kill-weight mud still must be circulated completely
to kill the well in a static condition.
Fig. 2 plots expected drillpipe pressure during the reverse-
circulation kill procedure. Pressure falls from Points A to B as cir-
culation begins as a result of drillstring friction. Pressure increases
from Points B toC as the gas bubble cifculates into the drillstring
owing to drilling fluid hydrostatic pressure loss. Gas is entirely in
the drill string at Point C and expands from Points C to D. Gas
reaches the surface at Point D, and drillpipe pressure falls as drill-
ing fluid replaces the kick fluid. The kick is circulated completely
1200
1000
t"1

I'
200
o
o
___ DRILLERS REVERSE
, _ DRILLERS NORMAL
1_ WAIT & WEIGHT
r-
V
f'
../
-
1 __ ,

"

....
....
100 200 300 400 500 600
TIME (MINUTES)
Fig. 3-Casing pressure vs. time, deep well, 30-bbl kick.
SPE Drilling Engineering, December 1992
TABLE 4-0FFSHORE WELL DATA
Well data
Water depth, ft
Rig derrick floor, ft above sea level
Total vertical length (RKB), ft
Openhole length, ft
Hole diameter, in.
Casing
3-in.-ID chokeline, ft
3-in.-ID kill line, ft
13318 x 12.615-in. casing, ft
Drillstring
5 x 4.276-in. drillpipe, ft
6% x 2-in. drill collars, ft
Bit nozzles, in.
Original mud weight, Ibm/gal
Kick size, bbl
Stabilized SIDPP, psi
Stabilized SICP, psi
Kill rate, gal/min
500
30
10,200
2,200
12%
530
530
set at 8,000
9,400
600
3x 12132
12.5
20
200
500
200
out of the well at Point E, and the circulation direction changes
back to normal. The pressure increase from Points E to F is the
circulation pressure of the well. Then kill-weight mud is pumped,
and it reaches the bit at Point G. Drillpipe pressure is held con-
stant at this point by choke manipulation. Once hydrostatic pres-
sure and annular friction exceed formation pressure, the drillpipe
pressure will increase because a choke increase will no longer con-
trol well pressures. This pressure increase also is seen in conven-
tional kill procedures.
Comparison With Conventional Procedures
Four drilling scenarios weredeve10ped to evaluate the potential
benefits of a reverse-circulation well-control procedure. Tables 1
through 4 summarize the data for each scenario. Table 1 is for
a relatively deep well, Table 2 is for an intermediate.-depth well,
Table 3 is for a relatively shallow well, and Table 4 is for an off-
shore well. Simulations were run for each scenario for kills with
the driller's method, wait-and-weight method, and reverse-'
circulation procedure described above. Figs. 3 through 14 pres-
ent casing pressure, drillpipe pressures, and pit gain as functions
of pumping time.
Tables 5 through 8 were generated with data from Tables 1
through 4 with initial pit gains of 30, 60, 15, and 20 bbl to com-
pare the reverse-circulation method with the driller's and wait-and-
weight methods in terms of maximum casing-shoe, maximum
casing, and maximum drillpipe pressures; total pit gain; time to
remove gas from the well; and total kill time. Reverse circulation
thus can be evaluated for each drilling scenario in Tables 1 through
4. We present simulation results so that individuals may draw their
own conclusions about the advantages of reverse circulation.
3500
3000

w
I!:

w
I!:
...
r
'500
II:
;;1'000
I!:
c
500
0

I
I
j;
lH
I II
III
I H
,
- DRILLER'S NORMAL
.-+!'--+----+---+--1,--- DRILLER'S REVERSE
- WAIT & WEIGHT
I ii
I II
I II
;:
1
=

0 100 200 300 400 500 600
TIME (MINUTES)
Fig. 4-Drillplpe pressure vs. time, deep well, 30-bbl kick.
249


::r J- DRILLER'S NOIIIIAL
;. 80+_----_+------",-----+-1 --- DRILLER'S REVERIE
Z 1_ WMr. WEIGHT

i
80
p--- --J. ------ --
, . .--"""
-

TIME (MINUTES)
Fig. 5-Plt gain VS, time, deep well, 30-bbl kick,
-+-----+-----+---+---+------1
- 11 1-DRILLER'S NORMAL
t 8OO+_--+--+---+--f-__.---+--1I--- DRILLER'S REVERSE
II! 1 - WAIT. WEIGHT
=
f
!i

rr-

i+1--+_--t---t---t-ft-t---t---t---t---+___i
,-- l

o 50 100 150 200 250 350 450
TIME (MINUTES)
Fig. 6-Caslrig pressure vs. time, medlum-depth well, 60-bbl
kick.


=- Iii
Ie I f I
i ,: J - DRILLER'S NORMAL
___ DRILLER'S REVERSE
::! ", i
III I I - WAIT. WEIGHT
f I I
1000 I !
Ii: I I
i i.


nME (MINUTES)
Fig. 7-Drlllplpe pressure vs. time, medium-depth well, 60-
bbl kick.
Deep WeD Example. A review of Table 5 reveals several interesting
points. The maximum casing pressures predicted for all three methods
are practically equivalent. The only major difference in casing pres-
sures between a conventional method and the reverse-circulation
method is that the reverse-circulation method reduces the time that
the annulus is exposed to high pressure. Drillpipe pressure, however,
is considerably higher during reverse circulation but not to a detri-
mental degree. Distinct differences arise between conventional and
reverse-circulation methods when pit gain and time required to re-
250
250
I
1_ DRILLER'S NORMAL
1 ___ DRILLER'S REVERSE
I
_ WAIT. WEIGHT

.{;

---
.---
--_.
80
o
o 80 350 450
TIME (MINUTES)
Fig. 8-Plt gain vs. time, medium-depth well, 60-bbl kick.


_ DRILLER'S NORMAL
___ DRILLER'S REVERSE
II! - WMr. WEIGHT


f


Fig. 9-Caslng pressure vs. time, shallow well, 15-bbl kick.
100
800
(700
l
1
800
II
1
800
II
;i
i'
Ii
::1
300
II
\
- DRILLER'S NORMAL
-J--!!i-----+-------+----i --- DRILLER'SREVERSE I----

- WAIT. WEIGHT
100
0
0 50 100 150 200 250
TIME (MINUTES)
Fig. 10-Drlllplpe pressure vs. time, shallow well, 15-bbl kick.
move gas from the well are compared. Maximum pit gain during the
reverse-circulation procedure is much less than during either conven-
tional procedure, and the gas is removed from the well in a fraction
of the time required with a conventional procedure. The reverse-cir-
culation method removes the gas from the well in the time required
to circulate the kick volume plus one drillstring volume, whereas
conventional methods require one cbmplete well circulation to remove
the gas. For obvious reasons, the reverse-circulation method re-
quires slightly longer to complete than the wait-and-weight method.
SPp Drilling Engineering, December 1992
::;
III
!!!.
z
<
Cl
>-
!L
120
100
80

I.
60
II
1_ DRILLER'S NORMAL
II
1 ___ DRILLER'S REVERSE
_ WAIT.WEIGHT
40
20
1------

-----
:
;,.,.,...--
o
o 50 100 150
TIME (MINUTES)
200
Fig. 11-Plt gain vs, time, shallow well, 15-bbl kick.
250


1 ___ DRILLER'S REVERSE
600 - DRILLER'S NORMAL t


IE
(:J -i J

__ __ ____ ____ +-____
I

... -.... - .... --
... ... __
o 100 200 300 400 500 600
TIME (MINUTES)
Fig, 12-Caslng pressure vs. time, offshore well, 20-bbl kick.
Medium-Depth Well Example. The results of the medium-depth
well example (Table 6) show essentially the same trends as the deep
well example, except that maximum casing pressure is much low-
er during reverse circulation than during a conventional kill, Max-
imum casing-shoe pressures once again are virtually identical,
Shallow Well Example. Table 7 identifies the same relationships
for a shallow well as Table 6 did for a medium-depth well but does
not provide additional insight into the comparison of reverse-
circulation with other methods.
Offshore Well Example. As with the shallow well example, this
well geometry provided no additional conceptual insight into the
1400
I

1200

;

!
w
a:

U>
U>
w
IE
w
...
!L
....
....
ii:
Q

600
400
200
o
o

=
100
---

.--
1_ DRILLER'S NORMAL
1 ___ DRILLER'S REVERSE
____ WAIT. WEIGHT
200 300 400 500 600
TIME (MINUTES)
Fig. 13-Drlllplpe pressure vs. time, offshore well, 20-bbl kick.
::;
III
!!!.
Z
<
Cl
>-
!L
120
,..--.
100
I
80
60
40
20
.----"
I
./
I
1_ DRILLER'S NORMAL I
I
-
___ DRILLER'S REVERSE
-
1_ WAIT. WEIGHT
I
o
o 100 200 400 500
TIME (MINUTES)
r-
600
Fig. 14-Plt gain vs. time, offshore well, 20-bbl kick.
comparison of the procedures. A very sharp increase in casing pres-
sure observed with the conventional methods is associated with the
chokeline. Note that the chokeline ID is only 3 in. This phenome-
non is even more conspicuous in deepwater drilling.
Cumulative analysis of the four scenarios indicates that a reduc-
tion in pit gain can be expected for a reverse-circulation procedure
and that kick fluids will be removed from the well in a shorter time,
Although lower casing pressure may be realized during a reverse-
circulation kill procedure, specific well geometries will dictate this
response. Note that the reverse-circulation method always would
result in lower casing pressures under these assumptions and thus
would appear to be an attractive alternative to conventional well-
control procedures. The intuitive assumption that reverse circula-
TABLE 5-DEEP WELL
Reverse-
Wait-and-Weight Circulation
Driller's Method Method Method
Initial Pit Gain (bbl)
30 60 30 60 30 60
Maximum casing-shoe 10,835 11,284 10,835 11,284 10,835 11,284
pressure, psi
Maximum casing 1,024 1,400 950 1,400 950 1,400
pressure, psi
Maximum drillpipe 3,120 3,120 3,095 3,100 3,095 4,040
pressure, psi
Maximum pit gain, psi 103 160 126 173 45 74
Time to reverse gas, 222 210 219 207 30 33
minutes
Total kill time, minutes 534 534 282 282 330 336
SPE Drilling Engineering, December 1992 251
TABLE 6-MEDIUM-DEPTH WELL
Reverse-
Wait-and-Weight Circulation
Driller's Method Method Method
Initial Pit Gain (bbl)
30 60 30 60 30 60
Maximum casing-shoe 5,445 5,514 5,445 5,512 5,423 5,556
pressure, psi
Maximum casing 755 1,020 732 800 242 375
pressure, psi
Maximum drillpipe 775 777 772 772 1,600 2,395
pressure, psi
Maximum pit gain, psi 137 200 151 203.5 52.5 73
Time to reverse gas, 196 184 196 182 22 18
minutes
Total kill time, minutes 488 488 256 256 294 300
TABLE 7-SHALLOW WELL
Reverse-
Wait-and-Weight Circulation
Driller's Method Method Method
Initial Pit Gain (bbl)
15 30 15 30 15 30
--
Maximum casing-shoe 1,256 1,856 1,256 1,856 1,256 1,856
pressure, psi
Maximum casing 600 1,200 600 1,200 600 1,200
pressure, psi
Maximum drillpipe 710 710 690 690 890 1,350
pressure, psi
Maximum pit gain, psi 100 154 130 170 26.5 43
Time to reverse gas, 103 96 99 91 5 6
minutes
Total kill time, minutes 243 243 124 124 135 137
TABLE 8-0FFSHORE WELL
Reverse-
Wait-and-Weight Circulation
Driller's Method Method Method
Initial Pit Gain (bbl)
20 40 20 40 20 40
Maximum casing-shoe 5,700 6,000 5,700 6,000 5,700 6,000
pressure, psi
Maximum casing 725 1,000 635 800 500 800
pressure, psi
Maximum drillpipe 950 950 950 950 1,365 1,875
pressure, psi
Maximum pit gain, psi 98 141.5 111 154 44 59
Time to reverse gas, 243
minutes
Total kill time, minutes 573
tion will break down the formation is valid only if kill-weight mud
is reverse circulated into the well.
Equipment Change.
To implement a reverse-circulation kill procedure, several equip-
ment changes or additions are required. First, a system to circu-
late the well quickly and safely either normally or in reverse is
necessary; this system requires pressure testing because it is an in-
tegral part of the well-control equipment. We recognize the exper-
tise of drilling contractors in these matters and make no further
attempt to suggest how to "replumb" the rig circulation system.
Brief conversations with several drilling contractors indicated that
this system either already exists or would be fairly straightforward
to design.
To eliminate the risk of plugging the bit jets, a circulation sub
(valve) should be added to the bottomhole assembly (BRA). We
252
219 237 228 27 27
573 300 300 342 345
believe this paper will encourage equipment designers to develop
the proper valve if such equipment is not yet available. As with
routing the well, we leave the drilling contractors the task of alter-
ing the BRA to allow reverse circulation.
Concluding Remark.
1. A simple simulator was developed to analyze the potential of
using reverse circulation to remove a gas kick from a well during
drilling.
2. The simulator confirmed that reverse circulation offers poten-
tial advantages over conventional procedures in terms of lower
casing pressures and reduced pit gains.
3. When large kicks or long chokelines cause excessive casing
pressures during the kill procedure, reverse circulation is an alter-
native well-control procedure that will not break down the casing
shoe and will not allow additional formation fluids to enter the well.
SPE Drilling Engineering, December 1992
4. Modifications to the rig circulation system and the BHA (cir-
culating sub) will be required to implement a reverse-circulation
kill procedure successfully.
Nomenclature
P af = annular friction pressure, mlLt2, psi
Pc = well casing pressure, mlLt
2
, psi
Pch = choke pressure, mlLt
2
, psi
Pfm = friction mud pressure, mlLt2, psi
Pio = friction gas pressure, mlLt2, psi
Phs = hydrostatic pressure, mlLt2, psi
Phsm = hydrostatic mud pressure, m/Lt2, psi
Phso = hydrostatic gas pressure, m/Lt2, psi
Pnew = new pressure in gas, mlLt2, psi
Pold = old pressure in gas, mlLt2, psi
Pws = well BHP, m/Lt2, psi
V
new
= new volume of gas bubble, L3, ft3
Void = old volume o ~ gas bubble, L3, ft3
Reference.
1. Goins, W.C. Jr.: Blowout Prevention, Gulf Publishing Co., Houston
(1969).
2. Rhem, B.: Pressure Control in Drilling, second edition, The Petrole-
um Publishing Co., Tulsa (1976).
3. Adams, N.: Well Control Problems and Solutions, The Petroleum Pub-
lishing Co., Tulsa (1980).
4. LeBlanc, J.L. and Lewis, R.L.: "A Mathematical Model of a Gas
Kick," Reprint Series, SPE, Richardson, TX (1973) 6, 178-88.
5. Nickens, H.V.: "A Dynamic Computer Model of a Kicking Well,"
SPEDE (Dec. 1986) 159-73.
6. White, D.B. and Walton, I.C.: "A Computer Model for Kicks in Water-
and Oil-Based Muds," paper SPE 19975 presented at the 1990
IADC/SPE Drilling Conference, Houston, Feb. 27-March 2.
7. Miska, S.Z., Luo, F., and Shafer-Perini, A.: "An Analysis of the In-
flow and Pressure Buildup Under Impending Blowout Conditions,"
J. Energy Resources Tech., ASME (March 1992) 114147.
8. Mills, S.R.: Blowout Prevention, Inti. Human Resources Development
Corp. (lHRDC), Boston, MA (1981).
9. Murugappan, B.S.: "Reverse Circulation Well Control: A New Ap-
proach to Well Control," MS thesis, New Mexico Inst. of Mining &
Technology, Socorro, NM (1990).
10. Bourgoyne, A.T. et al.: Applied Drilling Engineering, Textbook Se-
ries, SPE, Richardson, TX (1986) 2.
SPE Drilling Engineering, December 1992
Authors
Miska Beck Murugappan
Stafan Miska Is a professor and chairman of the Petroleum
Engineering Dept. at The U. of Tulsa at Tulsa. He previously
was a professor and chairman at New Mexico Inst. of Mining
& Technology (NMIMT) from 1981 to 1992 where he was In-
volved In drilling/production research and taught various pe-
troleum engineering courses. Miska holds MS and PhD
degrees In petroleum engineering from the Academy of Min-
Ing & Metallurgy In Cracow, Poland. He Is a member of the
Assn. of U.S. Petroleum Engineering Dept. Heads. Fredar-
Ick E. Beck Is a drilling engineer In the Prudhoe Bay Drill-
Ing Group for Arco Alaska Inc. He formerly was an assistant
professor of petroleum engineering at,NMIMT where he was
1989-90 faculty sponsor of the SPE student chapter. He holds
a as degree In geology and MS and PhD degrees In petroleum
engineering, all from Louisiana State U. a.sker Muftlgappan
Is currently a field engineer for the Middle East Dlv. of Dowell
Schlumberger In the United Arab Emirates. He holds BS and
MS degrees In petroleum engineering from NMIMT.
51 Metric Conver.lon Factor.
bbl x 1.589873 E-Ol m
3
ft x 3.048* E-Ol m
gal x 3.785412 E-03 m
3
in. x 2.54* E+OO cm
Ibm x 4.535924 E-Ol kg
psi x 6,894757 E+OO kPa
'Conversion factor is exact. SPEDE
Original SPE manuscript received for review March 11. 1991. Revised manuscript received
May 4. 1992. Paper accepted for publication Sept. 1. 1992. Paper (SPE 21966) first presented
at the 1991 SPElIADC Drilling Conference held in New Orleans. Feb. 17-21.
253