Contents Author Index

Overview of engineering and operations

activities conducted as part of the JOGMEC/
NRCan/Aurora Mallik 2007–2008 Gas Hydrate
Production Research Well Program, Part B:
2008 field program
D.I. Ashford1, M. Numasawa2, C.K. Martin3, K. Yamamoto4,
S.R. Dallimore5, J.F. Wright5, F.M. Nixon6, A. Applejohn7, and A.E. Taylor8

Ashford, D.I., Numasawa, M., Martin, C.K., Yamamoto, K., Dallimore, S.R., Wright, J.F., Nixon, F.M.,
Applejohn, A., and Taylor, A.E., 2012. Overview of engineering and operational activities conducted as
part of the JOGMEC/NRCan/Aurora Mallik 2007–2008 Gas Hydrate Production Research Well Program,
Part B: 2008 field program; in Scientific results from the JOGMEC/NRCan/Aurora Mallik 2007–2008
Gas Hydrate Production Research Well Program, Mackenzie Delta, Northwest Territories, Canada, (ed.)
S.R. Dallimore, K. Yamamoto, J.F. Wright, and G. Bellefleur; Geological Survey of Canada, Bulletin 601,
p. 53–65.

Abstract: The Mallik 2007–2008 Gas Hydrate Production Research Well Program was conducted at
the Mallik site in Canada’s Mackenzie Delta, Northwest Territories over two successive winters. Year 1
(winter 2007) operations are described in another paper in this volume. In year 2 (winter 2008), the
Mallik 2L-38 and Mallik 3L-38 wells were reopened following their suspension at the end of the previ-
ous year’s field campaign. Mallik 2L-38 was subsequently recompleted for extended production testing,
while Mallik 3L-38 was equipped as a produced-water disposal well. Operational activities during the
winter 2008 campaign were conducted during a 77-day period from January 16 through April 3, 2008, and
included ice-road and site construction, mobilization of camp facilities and service rig, well re-entry, well
testing, well abandonment, and demobilization.
The Mallik 2L-38 well was perforated (in 2007) over a 12 m interval from 1093 to 1105 m, just above
the estimated base of gas hydrate stability at the Mallik site. Production testing was conducted in three
stages, under three different bottom-hole pressure (BHP) targets. Continuous gas production from the
Mallik gas hydrate reservoir was achieved during a 6-day period from March 10 to 16, 2008. At the con-
clusion of the campaign, all of the Mallik wells that had been drilled during the course of four gas hydrate
research programs (in 1998, 2002, 2007, and 2008) were legally abandoned.

1
62 Sandstone Ridge Crescent, Okotoks, Alberta T1S 1P9, Canada (formerly Inuvialuit Oilfield Services Ltd.),
[email protected]
2
Japan Petroleum Exploration Co. Ltd, 1-2-1 Hamada, Mihama-ku, Chiba 261-0025, Japan
3
Schlumberger Canada Limited, 525–3rd Avenue SW, Calgary, Alberta T2W 0G4, Canada
4
Japan Oil, Gas and Metals National Corporation, 1-2-2 Hamada, Mihama-ku, Chiba 261-0025, Japan
5
Natural Resources Canada, Geological Survey of Canada–Pacific, 9860 West Saanich Road, Sidney, British Columbia
V8L 4B2, Canada
6
141 Clarendon Avenue, Ottawa, Ontario K1Y 0R4 (formerly Geological Survey of Canada)
7
Senior Science Advisor, Environment and Natural Resources, Government of the Northwest Territories, Yellowknife,
Northwest Territories, Canada
8
10227 West Saanich Road, Sidney, British Columbia, Canada

53
 Contents Author Index

Résumé : Le Programme de puits de recherche sur la production d’hydrates de gaz Mallik 2007-2008
s’est déroulé au site Mallik au cours de deux saisons hivernales successives au Canada, dans le delta du
Mackenzie, aux Territoires du Nord-Ouest. Les activités entreprises au cours de la première année (hiver
2007) sont décrits dans une communication qui paraît ailleurs dans le présent volume. Au cours de la
deuxième année (hiver 2008), on a procédé à la réouverture des puits Mallik 2L-38 et Mallik 3L-38, dont
l’exploitation avait été suspendue à la clôture de la saison de terrain de l’année précédente. De nouveaux
travaux de complétion ont été entrepris dans le puits Mallik 2L-38, afin de le rendre apte à servir à des
fins d’essais de production prolongée, tandis que le puits Mallik 3L-38 était équipé de façon à ce qu’il
puisse servir de puits de refoulement d’eau produite. Les activités opérationnelles exécutées au cours de
la saison de terrain de l’hiver 2008 se sont déroulées au cours d’une période de 77-jours, s’étendant du
16 janvier au 3 avril 2008. Elles ont exigé, notamment, la construction du site à l’emplacement du puits et
du chemin de glace pour s’y rendre; la mise en chantier des installations; l’installation de la plate-forme
de maintenance; ainsi que les travaux liés à la réintroduction, aux essais et à l’abandon du puits, et à la
fermeture de chantier.
On a pratiqué des perforations dans le tubage du puits Mallik 2L-38 (en 2007) le long d’une section
de plus de 12 m située entre 1093 et 1105 m de profondeur, soit juste au-dessus de la base présumée
de la zone de stabilité des hydrates de gaz au site Mallik. L’essai de production s’est déroulé en trois
étapes, et en fonction de trois cibles de pression de fond différentes . On a réussi à produire du gaz à
partir du réservoir d’hydrates de gaz Mallik de façon continue sur une période de 6-jours, entre le 10 et le
16 mars 2008. À la fin de la saison de terrain, tous les puits du site Mallik qui avaient fait l’objet de forages
au cours des quatre programmes de recherche sur la production d’hydrates de gaz (en 1998, 2002, 2007 et
2008) ont été abandonnés légalement.

54
 Contents Author Index

D.I. Ashford et al.

Key changes to the 2008 program relative to the original

INTRODUCTION proposal included 1) a single test in the same horizon as
year 1, rather than testing of multiple zones; 2) no geophysical
During the winter of 2007, the Aurora/JOGMEC/NRCan
logging; 3) implementation of sand-control measures; 4) injec-
Mallik 2L-38 and 3L-38 wells (drilled in 1998 and 2002,
tion of produced water into a separate disposal well, enabling
respectively) were reaccessed for the purpose of conducting
reconfiguration of the completion and multiphase test streams;
gas hydrate production testing of the Mallik reservoir. Mallik
5) incorporation of a down-hole heater to prevent gas hydrate
2L-38 was re-entered, drilled, cased, and used for a short-
re-formation in the test string; 6) a shorter duration produc-
duration production pre-test during the winter 2007 operating
tion test; and 7) restructuring of project management based on
season. Mallik 3L-38 was re-entered, deepened, and com-
lessons learned in 2007.
pleted as a produced-water disposal well, to be utilized for
an extended-duration production test scheduled for winter
2008. Both wells were suspended at the conclusion of the
winter 2007 season, pending resumption of field activities the MANAGEMENT OF 2008
following year. FIELD ACTIVITIES
The Aurora Research Institute was the legal operator
for the project, with Inuvialuit Oilfield Services Ltd. acting Management structure
as their project management consultant (Inuvialuit Oilfield The Steering Committee, Technical Committee, tech-
Services Ltd. is a joint venture between the Inuvialuit nical teams, role of the operator, and prime consultants
Development Corporation and Schlumberger Canada Ltd.). selected to co-ordinate field operations for the 2007 season
As the major funding partners of the project, the Japan continued with their roles for the 2008 field season (Ashford
Oil, Gas and Metals National Corporation (JOGMEC) and et al., 2012). However, new decision-making protocols
Natural Resources Canada (NRCan), through the Earth were implemented to improve field oversight. A project
Science Sector (ESS) and the Energy Sector (ES), jointly operations centre was established at the Aurora Research
managed the research and provided operational oversight for Institute facility in Inuvik prior to the start of field operations
the project. in 2008. The purpose of the centre was to provide a strong
This paper describes field activities undertaken in the focus on cost management and operations control, with rep-
winter of 2008, which was the second year of the JOGMEC/ resentation from the primary funding partners, operator, and
NRCan/Aurora Mallik 2007–2008 Gas Hydrate Production project manager, close to the field operations and available
Research Well Program carried out at the Mallik site in the for round-the-clock decision-making. Figure 1 depicts the
Mackenzie Delta. For general background on this program, organization chart for the project and the reporting structures
readers should refer to the companion paper by Ashford established prior to field operations.
et al. (2012), which describes winter 2007 activities as
well as the general goals and objectives of the overall pro-
gram. The original intent of the research program was to Permitting and approval process
establish appropriate site facilities and undertake a short
production test during the first winter field season (2007), As described in Ashford et al. (2012), a number of permits
and then return to the site during the second winter field and authorizations are required to conduct field operations
season (2008) to conduct a longer term production test. in the Mackenzie Delta. For 2008 activities, the National
Unfortunately, as described by Ashford et al. (2012), there Energy Board (NEB) renewed the Operator Licence for
were a number of problems encountered during the first Aurora College to assume responsibility for field oper-
field season, related to the short lead time for program ations. In early February 2008, the NEB issued program
planning, unexpected wellbore conditions, and insufficient approvals to carry out the field program. These included
allowance for the harsh arctic working environment and its permits to Alter the Condition of a Well (ACW) for Aurora/
impact on equipment. As a consequence of these problems JOGMEC/NRCan Mallik 2L-38 and 3L-38, and changes
and associated cost overruns during the first year, the Mallik to well names from JAPEX/JNOC/GSC Mallik 4L-38 and
Steering Committee agreed to a ‘scaled back program’ for 5L-38 to Aurora/NRCan Mallik 4L-38 and 5L-38 (note that
winter 2008 operations, with a substantial reduction in the the well names for Aurora/JOGMEC/NRCan mallik 2L-38
time available for production testing. A single test inter- and 3L-38 had already been changed from JAPEX/JNOC/
val was selected in the same zone that had already been GSC Mallik 2L-38 and 3L-38 as part of the previous 2007
perforated and tested during the previous winter, with the program). The Indian and Northern Affairs Canada (now
goal of achieving a 5- to 10-day depressurization test. Aboriginal Affairs and Northern Development Canada)
Field activities, from start of ice road construction to com- Land Use Permit and the Northwest Territories Water
pleted demobilization, were carried out over 77 days, from Board (NWTWB) Water Licence approved for the win-
January 18 to April 3, 2008. ter 2007 season carried over for the 2008 field activities;
however, the 2007 NWTWB Type B Water Licence did not
permit use of an injection well for disposal of produced

55
 Contents Author Index

GSC Bulletin 601

water during 2007 operations. As a result, the 2007 produc- and 35 cm thick beneath the rig. The area of the 2008 pad
tion testing program required a complex well completion was approximately 2.5 ha, which was somewhat smaller
with reinjection of produced water at greater depths within than that of the previous year given that only a service rig
the producing well. Fortunately for 2008 operations, the was used for re-entry, testing, and abandonment operations
NWTWB allowed pumping of produced water to the sur- (Fig. 2).
face and its subsequent reinjection in Mallik 3L-38, which
Mobilization of field equipment from staging areas in
had been prepared as a water injection well during the pre-
Inuvik to the Mallik site began on February 11, enabling
vious field season (see Ashford et al., 2012 for details on
start-up of the main camp on February 16. Transport and
the completion configuration of this well).
set-up of the service rig was completed on February 24, with
rig-less operations to re-enter the Mallik 3L-38 well being
conducted concurrently during the rig-up period.
LOGISTICS
The 2008 Mallik ice road followed a very similar route
to that of 2007 (Ashford et al., 2012), and a comprehensive MALLIK 3L-38 OPERATIONS
road-sharing agreement was implemented with MGM Energy
The preparation of Mallik 3L-38 as a produced-water
Corp., who were drilling elsewhere on the Mackenzie Delta.
disposal well was undertaken using a slick-line service unit
Ice-road construction activities commenced on January 25,
positioned over the well to retrieve the down-hole suspension
2008. A mobile construction ‘sleigh’ camp was towed from
plug from the wellbore. An injectivity test was conducted in
Tuktoyaktuk to the Mallik site on January 27 behind road-
advance of 2008 production operations, with a 5% potassium
and site-construction equipment. An ice-road profile was
chloride (KCl) brine employed in the final stage of the test,
conducted on February 7, which indicated sufficient thick-
yielding a sustainable injection rate of 250 L/min. at a surface
ness to support heavy loads from Inuvik. Preparation of the
pumping pressure of 4300 kPa. By comparison, the injectiv-
Mallik site commenced on February 1 using a fresh-water
ity test conducted during the 2007 field season recorded a
source located 2 km to the southeast. Construction of the ice
surface pumping pressure of 6800 kPa at the same injection
pad, completed by February 13, required 5.3 million L of
rate and brine formulation. A pressurized circulation loop,
water to create a pad 20 cm thick beneath the camp facilities
installed between 3L-38 and the service rig pump, allowed

JOGMEC AURORA RESEARCH INSTITUTE NATURAL RESOURCES CANADA (NRCan)

PLANNING
(Calgary and client/operator head offices)
Contract
Project
Manager
(Inuvialuit
Oilfield
JOGMEC Services Ltd./ OPERATION MANAGEMENT
Inuvik Schlumberger) NRCan
& COST CONTROL

INUVIK Project Operations Centre

JOGMEC NRCan SITE OPERATIONS

site

QHSE Well-Site Supervisor Logistics

Medic Testing Rig Completion/ Construction Camp Storeman

sand control maintenance & support
Monitors
Artificial
SKK Well services Other non-rig related services
lift

Wireline Other operational services

Figure 1. Organization chart for project management during the winter 2008 season of the
Mallik 2007–2008 Gas Hydrate Production Research Well Program (prepared by C.K. Martin,
Inuvialuit Oilfield Services Ltd.). Abbreviations: QHSE, quality, health, safety and environmental
management; SKK, Schlumberger K.K. (Japan).

56
 Contents Author Index

D.I. Ashford et al.

for continuous circulation of fluids during operations, thus to secure the derrick. Service rig operations commenced on
preventing freezing of the exposed surface lines. Actual February 24, with the first operation being the removal of
fluid injection was undertaken in discrete batches once an the 2L-38 wellhead and installation of the blowout preventer
adequate volume of produced fluid had accumulated in an (BOP). Prior to beginning completion operations, it was neces-
isolated 20 m3 rig tank. sary to retrieve the tubing-retrievable suspension bridge plug
installed the previous year (set at 1072 m). The well began
to flow gas immediately upon release of the bridge plug. A
MALLIK 2L-38 OPERATIONS total of 2.0 m3 of fluid and entrained gas flowed up the tubing
before the well was shut in. A subsequent build up of tubing
The Mallik site is located within 200 m of the Beaufort pressure to 3500 kPa was recorded. The well was circulated
Sea shoreline, at an elevation of approximately 1 m above sea for an extended period, during which time approximately 5 m3
level. Storm surges can cause flooding, so precautions were of 5% KCl brine was bull-headed into the well. By this means,
taken to prevent the well cellars from filling with water and the trapped gas was effectively circulated out of the wellbore
subsequently freezing. During August 2007, relatively large and the well subsequently stabilized with no additional flow
diameter inner tubes from truck tires were installed around observed.
the wellheads (Fig. 3) to minimize the amount of water/ice During bridge-plug retrieval, a BOP test was run for
that could accumulate in the cellars. The well heads were also crew-training purposes and to confirm the function of the
wrapped with steam hoses to afford an efficient mechanism for pipe rams. Unfortunately, the test was not successful, as
thawing the wellhead and cellar excavation, as required. Prior the pipe rams did not close properly. Upon disassembly of
to erecting the service rig, the distributed temperature sens- the BOP, the failure was attributed to a defective ram shaft
ing (DTS) and light electric resistance array (L-ERA) cables (Fig. 5). The ram shaft had been inspected and tested by a
were spliced and connected to surface monitoring equip- third-party specialist in December 2007, and this was the
ment to enable recording of baseline temperature profiles first deployment following its certification. Subsequent
and resistivity data (see Fujii et al., 2008). metallurgical analysis indicated that the failure was caused
Shehtah Nabors LP service rig 203, a free-standing, by torsion overloading in combination with pre-existing
mechanical double service rig rated to 3500 m with 73 mm cracks in the shaft. Following replacement of the ram shaft,
tubing (Fig. 4), was used for Mallik 2L-38 operations. The the bridge plug was successfully retrieved and scheduled
free-standing capability is accomplished by a load beam and operations could proceed.
outriggers, thus eliminating the requirement to set rig anchors

Fd.l. 400.35 m

Thaw
ponds
MALLIK 6L-38 MALLIK 2L-38 drill sump
MALLIK 5L-38

Lease boundary
MALLIK 3L-38 MALLIK 4L-38

Flare pad

MALLIK 2L-38

MALLIK L-38

1.0 m
245.70 m

MALLIK L-38
drill sump LEGEND:
Wellhead Ice pad

Contours (m)
N
1.0 m
GPS Co-ordinates
Description Northing Easting
Fd.I. (east) 7705676.96 513629.42
0 Fd .I. (south) 7705388.09 513352.88
50 50 meters
Fd.I. (west) 7705557.63 513175.22

Fd.l. Lease boundary Fd.l.

m
Access road 20

Figure 2. Plan of the Mallik drill site, showing locations of research wells and ice pad, 2008.

57
 Contents Author Index

GSC Bulletin 601

In an effort to avoid the problems encountered during this screen size was to avoid reduced permeability in the near–
winter 2007 operations (see Ashford et al., 2012), one of the well-bore area by allowing passage of fine material into the
first activities undertaken upon re-entry of the 2L-38 well was well bore, while retaining the bulk of formation sand behind
the installation of a mesh screen at the perforations that was the screen.
designed to limit sand infiltration into the production tubing.
After installation of the sand screen, the production test-
Grain-size analysis of Mallik 5L-38 cores and sand recovered
ing string was run into the well. This assembly consisted of
from the 2007 test suggested that the main sediment type in the
a down-hole electric heater, an electric submersible pump
production interval is fine- to medium-grained sand containing
(ESP), and pressure-temperature (P-T) gauges, all on 73 mm
some silt. Although a number of countermeasures were con-
tubing. The ESP and down-hole heater featured combined
sidered to limit sand infiltration, a composite mesh–type sand
power and capillary-tube umbilical lines strapped to the out-
screen was selected due to the short duration of the test. The
side of the tubing and extending from the surface to their
composite screen was 177.8 mm in diameter with a 200 µm
respective down-hole locations. Technical specifications for
mesh, and was installed across the existing perforations from
the ESP and an overview of the placement and strategic util-
1093 to 1105 m (Fig. 6). Packers were set above and below
ization of the down-hole heater are presented in Rojas et al.
the sand screen to anchor it in place. The rationale in selecting
(2012).

Figure 5. Failed pipe ram shaft from the blowout preventer on

Figure 3. Summer site visit in preparation for Mallik 2L-38 well-head Mallik 2L-38 during re-entry of the well in 2008. Photograph by
access during the winter 2008 season of the Mallik 2007–2008 S. Wan, C & M Engineering Ltd. 2012-067.
Gas Hydrate Production Research Well Program. Photograph by
C.K. Martin, Inuvialuit Oilfield Services Ltd. 2012-071.

Figure 6. Stand-alone sand screen of composite mesh type

Figure 4. Shehtah Nabors LP service rig 203, Aurora/JOGMEC/ (177.8 mm diameter, 200 µm mesh size) installed across the
NRCan Mallik 2L-38 gas hydrate production research well, 2008. production-test interval in the Mallik 2L-38 well for the 2008 gas
Photograph by S.R. Dallimore. 2012-068. hydrate production test. Photograph by C.K. Martin, Inuvialuit
Oilfield Services Ltd. 2012-072.

58
 Contents Author Index

D.I. Ashford et al.

While the completion string was being run in, the well differential-pressure orifice meter; and 3) a positive-displace-
again began to flow when the bottom of the completion string ment meter for accurately measuring gas volumes below the
reached a depth of about 800 m. At this point, the annular minimum operating threshold of the orifice meter.
BOP was closed around the tubing and the two umbilical
A program was implemented to periodically sample pro-
lines. However, the annular BOP failed to effectively seal
duced gas, water, and sediment during the test. Water-sample
within operating-pressure range and it was therefore decided
characteristics were measured at the site and subsamples
to cut the umbilical lines to secure the well. Once the well was
retained for further testing in Inuvik and at southern labora-
secured, a stabilized shut-in tubing pressure of 1350 kPa and
tories. Gas and sediment samples were retained for post-field
shut-in casing pressure of 1700 kPa were observed, with the
analysis. All samples were inventoried, stored, and transported
gas kick assessed to be approximately 4 m3 in volume (at stan-
with appropriate temperature control and monitoring.
dard temperature and pressure). The well was successfully
killed by the volumetric well-control procedure (Grace et al., The production test began at 06:00 GMT on March 10,
1994), through bull-heading a total of 6.5 m3 of 12% sodium 2008. The down-hole heater was powered up first to pre-heat
chloride (NaCl) brine into the well in two separate operations. the wellbore fluid before it entered the ESP, in order to pre-
Following this, the umbilical power lines were successfully vent gas hydrate re-formation within the pump and/or tubing
spliced and the completion string was rerun and landed at the during the production stage. Figure 9 depicts the down-hole
planned target depth. The general configuration of the com- pressure regime (at the perforations), as well as the instan-
pletion string and the location of specific components of the taneous and cumulative gas flows for the three distinct
assembly are presented in Figure 7. operating phases during the production test of the perforated
interval (1093–1105 m) in the Mallik 2L-38 well. In phase 1,
the bottom-hole pressure (BHP) was reduced to 7.5 MPa by
PRODUCTION-TESTING OPERATIONS using the ESP to lower the fluid level in the casing annu-
lus down to approximately 330 m below surface. Phase 2
Two completely separate trains (or systems) of surface lowered the annular fluid level to 580 m below surface, thus
production-testing equipment were connected at the Mallik reducing the BHP to 5 MPa. Finally, in phase 3, the annular
2L-38 wellhead (Fig. 8). Train A, connected to the tubing fluid level was drawn down to about 700 m below surface,
outlet at the wellhead, was designed to process the water- yielding a BHP close to 4 MPa.
dominated flow stream produced through the down-hole ESP Production-test operations continued until 12:00 GMT
and tubing to the surface. It included an array of sand filters on March 16. During the 6 days of testing, a cumulative
located upstream of a gas/water separator, with the capability total of 12 300 m3 of gas was produced, with the flow rate
to capture sand down to 50 µm in size. Periodic inspection at the end of phase 3 approaching 2 500 m3/d. The total
of the sand accumulated in these filters provided an indirect water produced was approximately 70 m3 (Fig. 10); this was
indication of the efficiency of the down-hole sand screens in pumped to the surface and subsequently reinjected into the
preventing the migration of larger sand particles from the res- Mallik 3L-38 disposal well. During the test, two separate ice-
ervoir into the well-bore and production stream. Indeed, only plugging events required the gas-water separators to be shut
small amounts of fine-grained residue (very fine sand and in; however, the ESP continued to operate (and the well to
minor silt) were observed on these filters, suggesting that the produce) during these two events, which totalled just over
down-hole sand screens were performing according to design 5 hours of interruption in the collection of surface data. The
specifications. Train A also included a rotary-vane–type water gas- and water-production data were continuously recorded
meter, designed to directly measure the water-flow rate during and archived for subsequent post-field analysis by numerical
production. However, the actual water-production rate real- reservoir simulation, including history matching (Kurihara et
ized was often below the minimal operational threshold of al., 2012; Uddin et al., 2012), prediction of long-term produc-
this meter, so water flow measured directly by this meter was tion trends (Uddin et al., 2012), and short-term geomechanical
considered to be unreliable. Ultimately, produced water was responses of the reservoir to production (Uchida et al., 2012).
allowed to flow from the Train A separator into an external
rig tank, and changes in water level over time were recorded
and subsequently converted to reasonable estimates of the
water-production rate. WELL ABANDONMENT
Train B was connected to the annular space between the
AND DEMOBILZATION
tubing and the well casing, and was designed to handle a Following the production test, the Mallik 2L-38 well
fluid stream dominated by free gas. It incorporated three sep- was killed by pumping 35 m3 of 12% NaCl brine down the
arate gas measuring devices: 1) a Schlumberger multiphase tubing. The remaining fluid occupying the tubing-casing
meter (located upstream of the Train B separator), which annulus was recovered to the surface using the Train B test
measures gas, condensate, and water concurrently, based on equipment and subsequently injected into the Mallik 3L-38
flow velocity and a nuclear densitometer; 2) a conventional disposal well. The ESP, down-hole heater, gauges, and

59
 Contents Author Index

GSC Bulletin 601

8 MPa at ~283 m
6 MPa at ~485 m 20", 94 lbs./ft., J55 (103 m)
5 MPa at ~581 m
4 MPa at ~688 m
3 MPa at ~789 m 2.875" EUE, L80 Tbg
Base of Permafrost
at 640 m 13.375", 61 lbs./ft., J55 (677 m)

ESP cable with chemical injection #4

CTS cable (back-up ESP sensor)
Downhole heater #4 cable (NO CHEMICAL INJECTION)
Downhole heater #4 cable Bleeder valve (683 m)
(no chemical injection)
Bleeder valve (739 m)

Bleeder valve with STOP (795 m)

CTS gauge carrier (~798 m)

Chemical injection splice (injection point at 811 m)

EUE: external upset ends

Pump intake (811 m)
Tbg: tubing
Phoenix gauge (~822 m)
ESP: electric submersible pump

CTS: custody transfer system Downhole heater (825–837 m)

FC: float collar

Anular gauge carrier bottom (1083 m)

LB permanent packer (top at 1083 m)

Mesh-Rite sand control screens

Gas hydrate zone (1093–1105 m)

LB permanent packer (top at 1110 m)

Cement plug (1194 m)

Bridge plug (1202 m)

Model S packer (1211 m)

Model S packer (1218 m)

Injection zone (1224–1230, 2.313" landing SXN nipple (1240 m)

1238–1256, 1270–1274 m) Model B shear plug (1238 m) (sheared)

FC (1275 m)

Shoe 9.625", 40 lbs./ft., J55 (1288 m)

Figure 7. Schematic diagram of completion string for the Aurora/JOGMEC/NRCan

Mallik 2L-38 gas hydrate production research well during the winter 2008 production
test (prepared by L. Hernandez-Johnson, Schlumberger Ltd.).

60
 Contents Author Index

D.I. Ashford et al.

tubing were all recovered in good condition. Mallik 2L-38 Mallik site had been completed. A site inspection by Indian
was abandoned by setting permanent tubing-set bridge plugs and Northern Affairs Canada (now Aboriginal Affairs and
at 1050 and 103 m. The 244 mm casing was filled with Northern Development Canada) on August 28, 2008 indicated
cement from the bridge plug at 103 m to the surface. that all aspects of the site and access were acceptable.
The service rig was then moved to Mallik 3L-38 disposal
well to pull the 73 mm tubing, after which the well was
abandoned by setting permanent tubing-set bridge plugs at DISCUSSION
1049 and 103 m. The 140 mm casing was filled with cement
from the bridge plug at 103 m to the surface. Planning
The final surface abandonment of Mallik 2L-38, 3L-38, Much of the knowledge and experience gained during
4L-38, 5L-38, and 6L-38 was accomplished by excavating the winter 2007 field program was successfully applied
around the wells to between 1.0 and 1.5 m below ground to the winter 2008 operations. During the planning stage
level, cutting all monitoring lines, welding casing plates over for the second year, more emphasis was given to detailed
all casings where cover plates were not already in place, time and cost estimates for individual program compon-
welding monument poles onto all wellbores, backfilling the ents, in order to better control schedules and budgets. A
excavations, and fastening well identification plates to the lack of clearly defined roles and responsibilities within and
monument poles. between program management, operations, contractors, and
clients in the winter 2007 program led to problems in mak-
All equipment was trucked from the Mallik site to Inuvik.
ing critical decisions in a timely fashion. This shortcoming
With the exception of construction equipment, all equipment
was addressed by devising a comprehensive communica-
was off the site by March 26. Final cleaning of the ice pad
tions plan that clearly defined the roles and responsibilities
continued until April 3, at which time all activities at the
of the various participants. It also clarified the process of

Well testing unit (A) Relief line to

CINJP WHUPA WHDPA Flare stack N1
WHUTA WHDTA
B
TRAIN A DPU
CR
NRV
Surface Dual-pot 1
X-mas tree Safety valve Choke
manifold
sand trap C Relief line to
PCV
Flare stack N1
P-TANK

D
LCV

DPU
SP

CR
Surface
Safety valve
(B)
WTA GASPA
GASTA GASPC
WQA GASTC
GASDPA
GASQA GASDPC
GASQC
Well testing unit (B)
Low gas
measurement skid
TRAIN B DPU DPU Relief line to
CR CR
Chemical
A Phase tester
NRV
1
Flare stack N2

pump C
Choke Tester
PCV PCV
SP
Relief line to
Flare stack N2
P-TANK B

D
LCV

DPU
To 400bbl tank
SP

CR
Emergency
down stations
PIN GASPB
POUT CSGUP CSGDP WTB GASTB
DPV CSGUT CSGDT WQB GASDPB
TL QASQB
LEGEND PIN: Phase tester pressure in WHUP:Well-head upstream pressure on train A/B
POUT: Phase tester pressure out WHUT: Well-head upstream temperature on train A/B
D Liquid line DVP: Venturi differential pressure WHDP: Well-head downstream pressure on train A/B
C Gas line TL: Line temperature WHDT: Well-head downstream temperature on train A/B
CSGUP: Casing upstream pressure CINJP: Chemical-injection-pump pressure
B Pressure relief line CSGUT: Casing gas differential pressure
PCV: Pilot circuit valve
A Well effluents GASP: Separator pressure on train A/B/C
LCV: Liquid control valve
GAST: Separator temperature on train A/B/C
Fluid Water GASDP:Separation gas differential pressure NRV: Non-return (check) valve
Gas Relief line on train A/B/C SP: Sampling point
QASQ: Gas flow rate on train A/B/C DPU-CR: Differential pressure unit
Hydraulic signal
pressure signal WT: Seperator water temperature on train A/B
WQ: Water rate on train A/B th
Updated 10 January 2008

Figure 8. Schematic diagram of test equipment configuration for the winter 2008 production test of the Aurora/
JOGMEC/NRCan Mallik 2L-38 gas hydrate production research well (prepared by L. Hernandez-Johnson,
Schlumberger Ltd.).

61
 Contents Author Index

GSC Bulletin 601

Phase 1 Phase 2 Phase 3

13 000 13 000

12 000 12 000
Initial gas flow to surface
11 000 11 000

10 000 10 000
Commence gas metering CUMULATIVE GAS
9 000 9 000

Gas produced (m )
Pressure (kPa)

3
8 000 8 000
7 000 7 000

6 000 6 000
BOTTOM-HOLE PRESSURE
5 000 5 000
4 000 4 000

3 000 3 000
INSTANTANEOUS GAS RATE
2 000 2 000

1 000 1 000
0 0

12:00 12:00 12:00 12:00 12:00 12:00 12:00

March 10 March 11 March 12 March 13 March 14 March 15 March 16
Date and time (GMT)

Figure 9. Down-hole pressure and gas production during the three phases of depressurization
of the Aurora/JOGMEC/NRCan Mallik 2L-38 well, winter 2008 production test.

120 100
Phase 1 Phase 2 Phase 3
110 90
100
Train B metering 80
period begins
90
70
80

Water produced (m )
3
Bottom-hole pressure
70 60
Pressure (kPa)

60 50

50 40
40
30
30
'Zeroed' to start of Train A
data-collection period 20
20
10
10 Ice plug at surface
0 0
16:00 16:00 16:00 16:00 16:00 16:00 16:00
March 10 March 11 March 12 March 13 March 14 March 15 March 16
Date and time (GMT)

Bottom-hole flowing pressure (memory gauge)

Cumulative water production (flow meter)
Cumulative water production (P-tank level)

Figure 10. Water production during the three phases of depressurization of the Aurora/
JOGMEC/NRCan Mallik 2L-38 well, winter 2008 production test.

62
 Contents Author Index

D.I. Ashford et al.

consultation and authorization for critical decision-making, communications were problematic in 2007, partly because
including the timely management of activities and schedules of misunderstandings by the contractor and its field oper-
in response to unforeseen conditions and events. An import- ators regarding the requirements of the various program
ant part of the communication strategy was the establishment participants. Although still not perfect, changes to the sys-
of a Project Management office in Inuvik (which was staffed tem implemented for the 2008 program resulted in better
throughout the field-operations phase) to facilitate effective capacity and reliability than had been achieved the previ-
liaison between the operator, field supervisors, contractors, ous year. In addition, the operator provided an independent
and clients. Perhaps just as important was the involvement of satellite phone system for its representatives in 2008, con-
all participants in each planning stage to ensure that all ser- siderably relieving the strain on the core communications
vice providers were clear on the goals and objectives of the systems.
client (e.g. the need to measure all gas from the beginning of
a production test, sampling requirements, and model predic-
tions to inform the establishment of operational parameters). Well operations
Detailed decision trees were constructed for critical activities
to capture a wide range of possible operational scenarios and Two well-control events (mentioned in the ‘Mallik 2L-38
to provide guidance for dealing with each, thereby ensuring operations’ section) were encountered, emphasizing the need
appropriate contingencies for situations where the risks were to re-evaluate standard well-control procedures as applied to
deemed significant. completion and production operations in gas hydrate reser-
voirs. The first event occurred during retrieval of the bridge
As a result of the previous year’s field operational experi- plug used to suspend the well at the end of the 2007 field pro-
ence and the implementation of a more rigorous management gram. Given that the natural in situ reservoir salinity in the
approach, considerably more realistic time and budget contin- targeted production zone (around 1100 m depth) was in the
gencies were built into the winter 2008 program. In the 2007 range of 45 ppt (Wright et al. 2005), the gas hydrate within
program, the drill rig was oversized and had been in storage the production-test interval was very close to its instabil-
for some time. This, together with inefficiencies inherent in ity threshold. It is thought that the relatively warm 5% KCl
the use of a drill rig instead of a service rig for production fluid (about 50 ppt KCl) used to kill the well following the
testing, resulting in excessively long rig-up time and a com- 2007 production test increased local salinity and temper-
paratively awkward deployment of equipment in the 2007 ature enough to promote gradual dissociation of residual gas
program. Furthermore, a number of mechanical issues were hydrate in the vicinity of the perforated interval. This likely
encountered with the Shehtah Nabors rig 414 used in 2007 continued for some time after the bridge plug was set and
(Ashford et al., 2012), some of which were related to insuffi- the well suspended, with free gas accumulating below the
cient lead time for procurement. In order to minimize rig bridge plug until local pressure increases, combined with
deficiencies in the 2008 program, a number of meetings of stabilization of local reservoir temperatures, stopped the dis-
the Project Operations Team and Nabors Production Services sociation reaction. Upon removal of the bridge plug at the
took place before the Shehtah Nabors rig 203 was selected. start of 2008 operations, the residual free gas thus accumu-
This service rig and the accessory buildings and connection lated would have flowed up the well to the surface, causing
piping for the Mallik 3L-38 produced-water disposal well an observable gas kick. The solution applied was to force
were assembled at the Nabors facility in Leduc, Alberta prior more 5% KCl into the formation, which may once again
to transport of the equipment to the Mackenzie Delta. have shifted near-wellbore reservoir conditions outside the
gas hydrate stability field; this manifested as a second gas-
kick event when the completion string was run into the well
Road and site construction several days later. Again the response was to force brine into
activities and logistics the formation, this time 12% NaCl.

Effective construction and maintenance operations for In hindsight, this conventional well-control strategy may
roads and facilities was achieved for the winter 2008 pro- have led to recurrent dissociation of gas hydrate in the near-
gram by using experienced northern contractors that had wellbore area, and to difficulty in managing the well under
performed well during the 2007 field program. This led to reservoir conditions that were close to the boundaries of
negotiation of more beneficial contracts and improved atten- gas hydrate stability. Therefore, a thorough understanding
tion to details (e.g. well-site topography), which in return of stability conditions in a particular setting, together with
resulted in more efficient road and facilities construction. careful monitoring of pressure and temperature conditions
Materiel management was also improved in the 2008 pro- in the well, is critical to successful management of produc-
gram, as dedicated stores staff were employed to implement tion operations in gas hydrate reservoirs. In addition, the use
a comprehensive shipping, receiving, and inventory system. of nonsaline, freezing-depressed suspension and completion
An efficient, experienced expeditor was retained from the fluids should be considered as an alternative to conventional
2007 program, as was the camp sewage and waste-water bull-heading of saline fluids for well-control operations.
system that had performed well in the first year. Remote

63
 Contents Author Index

GSC Bulletin 601

Production test did occur, although they were generally quickly addressed
and had only a minor impact on production operations and
Perhaps the most critical operational knowledge gained related data acquisition.
during the winter 2007 operation was from the planned
short-duration production test, the purpose of which was Finally, the 2008 extended-production test was con-
to evaluate equipment and production strategies in prepara- figured to pump produced water to the surface (via the
tion for an extended-duration production test the following production tubing) and through the surface-testing equip-
year. In practice, this test was intermittent and ultimately ment of Train A, with subsequent injection into a dedicated
abbreviated as a result of pump-performance problems and produced-water disposal well (Mallik 3L-38). Indeed, the
eventual pump failure due to excessive sand entering the cumulative rate of water production was sufficiently low that
casing (see Ashford et al., 2012). Post-production analysis continuous injection could not be undertaken, so the pro-
of the pump and related equipment, in conjunction with duced water was instead held in an external rig tank until
recorded pump-performance data, indicated that, although a sufficient volume had accumulated for periodic injection.
pump capacity was appropriate for the expected volumes The experience and knowledge gained during the 2007
of produced gas and water, the implementation of effective program were indeed valuable and undoubtedly contributed
sand-control measures would be crucial to the successful greatly to the efficient, safe, and environmentally respon-
realization of sustained, long-term production. Modelling sible operations of the Mallik 2008 program, including
and laboratory tests, using archived sediment samples col- the completion of 6 days of continuous production testing.
lected during the Mallik 2002 Gas Hydrate Production This experience, however, was gained at considerable cost.
Research Well Program (Dallimore and Collett, 2005), Due to the significant budget over-runs encountered in
provided a basis for the evaluation of various sand-control 2007, together with the failure of major components of the
options (e.g. gravel-pack versus sand screens). In the end, a externally mounted down-hole monitoring systems and end-
composite mesh–type, stand-alone sand screen was selected of-season difficulties in meeting environmental commitments,
and successfully employed in the 2008 extended-production the major funding partners of the program were motivated to
test. This selection was validated by good overall pump per- minimize perceived risks in the following year. Consequently,
formance during the test, the lack of pump wear observed in the original 2008 program plan (20 days testing up to three
the post-test inspection, and the very low volumes of sand production zones, incorporating pre- and post-test cased-hole
produced (supported by grain-size analysis of produced sol- logging) was scaled back to a nominal 5.5 day test (ultim-
ids), all of which confirmed that effective sand control was ately a 6-day test was achieved) of the already-perforated
achieved. zone, without any well logging. As a result, a number of
The 2008 testing program utilized a more conventional the improvements proposed for the 2008 testing program
pump configuration (compared to 2007), with the ESP intake (e.g. oriented perforation with optimized sizing and pene-
positioned well above the production perforations and sand tration, more robust logging operations, and enhanced data
screen (Fig. 7). In addition, a down-hole electric-induction communications) could not be implemented and evaluated.
heater was positioned between the perforations and the pump
(see Rojas et al., 2012). The heater was designed to pre-heat
produced fluid prior to it entering the pump intake, so as to SUMMARY
minimize the risk of gas hydrate re-formation (and possible
blockage of flow pathways) within the pump assembly and tub- The Mallik Gas Hydrate Production Research Well
ing. This requirement was based on the analysis of the 2007 Program was conducted in a safe and environmentally
test results and subsequent modelling, which suggested that responsible manner from January through April of 2008.
re-formation of gas hydrate could not be ruled out, given the Gas hydrate was continuously dissociated and produced
down-hole pressure and temperature conditions anticipated to the surface during a 6-day production test by reservoir
during the production test. depressurization, and measurable gas and water were pro-
duced to surface. Production-flow data were captured under
In order to accommodate the requirement to measure (and
three distinct bottom-hole flowing pressures. Produced
sample) multiphase fluids under low flow volumes and rates,
water was injected into a sand formation located well below
the 2008 surface-testing facilities featured two independent
the gas-hydrate-bearing intervals at Mallik, using the recom-
flow/test streams consisting of separate surface equipment
pleted 3L-38 as an injector well (Fig. 2). Two well-control
for flows dominated by gas (annulus) and water (tubing;
events were successfully managed. The operations team and
Fig. 8). Insulation and tracer heating of exposed surface flow
the Technical Committee built adequate contingency into the
lines was also implemented to minimize the possibility of
schedule to allow for weather delays and operational prob-
ice plugging and flow interruption in this extreme environ-
lems. The program was completed within the allotted time
ment, in which air temperatures commonly fall to -30 to
frame and under budget. All gas hydrate wells drilled in the
-40°C or lower for extended time periods. In spite of these
course of the four Mallik research programs conducted over
measures, a number of freezing events and pipe blockages
the previous 11 years were abandoned according to relevant
federal and territorial regulations.

64
 Contents Author Index

D.I. Ashford et al.

Rojas, M., Martin, C.K., Hernandez-Johnson, L., Ashford, D.,

ACKNOWLEDGMENTS Wright, J.F., Yamamoto, K., Numasawa, M., Dallimore, S.R.,
and Isted, R.E., 2012. Electric submersible pump as an effect-
The authors acknowledge the partnership that undertook ive artificial lift method to control bottom-hole pressure in a
the 2007–2008 Mallik Gas Hydrate Production Research producing gas hydrate well, JOGMEC/NRCan/Aurora 2007–
Well Program: the Japan Oil, Gas and Metals National 2008 Mallik Gas Hydrate Production Research Well Program;
Corporation, Natural Resources Canada, and the operator in Scientific Results from the JOGMEC/NRCan/Aurora Mallik
2007–2008 Gas Hydrate Production Research Well Program,
for the program, Aurora College. Mackenzie Delta, Northwest Territories, Canada, (ed.)
S.R. Dallimore, K. Yamamoto, J.F. Wright, and G. Bellefleur;
Geological Survey of Canada, Bulletin 601.
REFERENCES Uchida, S., Soga, K., Klar, A., and Yamamoto, K., 2012:
Geomechanical study of the Mallik gas hydrate production
Ashford, D.I., Mizuta, T., Dallimore, S.R., Yamamoto, K., field trials; in Scientific results from the JOGMEC/NRCan/
Nixon, F.M., Imasato, Y., Wright, J.F., Taylor, A.E., and Aurora Mallik 2007–2008 Gas Hydrate Production Research
Applejohn, A., 2012. Overview of engineering and operations Well Program, Mackenzie Delta, Northwest Territories,
activities conducted as part of the JOGMEC/NRCan/Aurora Canada, (ed.) S.R. Dallimore, K. Yamamoto, J.F. Wright, and
Mallik 2007–2008 Gas Hydrate Production Research Well G. Bellefleur; Geological Survey of Canada, Bulletin 601.
Program, Part A: 2007 field program; in Scientific results
from the JOGMEC/NRCan/Aurora Mallik 2007–2008 Gas Uddin, M., Wright, J.F., Dallimore, S.R., and Coombe, D., 2012.
Hydrate Production Research Well Program, Mackenzie Delta, Gas hydrate production from the Mallik reservoir: numerical
Northwest Territories, Canada, (ed.) S.R. Dallimore, history matching and long-term production forecasting;
K. Yamamoto, J.F. Wright, and G. Bellefleur; Geological in Scientific results from the JOGMEC/NRCan/Aurora
Survey of Canada, Bulletin 601. Mallik 2007–2008 Gas Hydrate Production Research Well
Program, Mackenzie Delta, Northwest Territories, Canada,
Dallimore, S.R. and Collett, T.S. (ed.), 2005. Scientific results from (ed.) S.R. Dallimore, K. Yamamoto, J.F. Wright, and
the Mallik 2002 Gas Hydrate Production Research Well Program, G. Bellefleur; Geological Survey of Canada, Bulletin 601.
Mackenzie Delta, Northwest Territories, Canada; Geological
Survey of Canada, Bulletin 585, 140 p. and 2 DVD-ROMs, Wright, J.F., Dallimore, S.R., Nixon, F.M., and Duchesne, C.,
<http://geoscan.ess.nrcan.gc.ca/cgi-bin/starfinder/0?path=geoscan. 2005. In situ stability of gas hydrate in reservoir sediments
fl&id=fastlink&pass=&format=FLSHORTORG&searc of the JAPEX/JNOC/GSC et al. Mallik 5L-38 gas hydrate
h=R=220702> [accessed October 10, 2012]. production research well; in Scientific results from the Mallik
2002 Gas Hydrate Production Research Well Program,
Fujii, K., Yasuda, M., Cho, B., Ikegami, T., Sugiyama, H., Mackenzie Delta, Northwest Territories, Canada,
Imasato, Y., Dallimore, S.R, and Wright, J.F., 2008a. (ed.) S.R. Dallimore and T.S. Collett; Geological Survey of
Development of a monitoring system for the JOGMEC/ Canada, Bulletin 585, 11 p., <http://geoscan.ess.nrcan.gc.ca/
NRCan/Aurora gas hydrate production test program; in cgi-bin/starfinder/0?path=geoscan.fl&id=fastlink&pass=&form
Proceedings of the 6th International Conference on Gas at=FLSHORTORG&search=R=220702>
Hydrates (ICGH 2008), Vancouver, British Columbia, July [accessed October 10, 2012].
6–10, 2008, Paper 5830, 6 p.
Grace, R.D., Cudd, B., Carden, R.S., and Shursen, J.L., 1994.
Advanced blowout & well control; Gulf Publishing Company,
Houston, Texas, 397 p.
Kurihara, M., Sato, A., Funatsu, K., Ouchi, H., Yamamoto, K.,
Fujii, T., Numasawa, M., Masuda, Y., Narita, H.,
Dallimore, S.R., Wright, J.F., and Ashford, D., 2012. Analysis
of 2007 and 2008 gas hydrate production tests on the Aurora/
JOGMEC/NRCan Mallik 2L-38 well through numerical
simulation; in Scientific results from the JOGMEC/NRCan/
Aurora Mallik 2007–2008 Gas Hydrate Production Research
Well Program, Mackenzie Delta, Northwest Territories,
Canada, (ed.) S.R. Dallimore, K. Yamamoto, J.F. Wright, and
G. Bellefleur; Geological Survey of Canada, Bulletin 601.

65