Documente Academic
Documente Profesional
Documente Cultură
Executive Summary
Drilling and testing of the East Barrow Gas Field (EBGF) began in 1974. Over the next sixteen years,
eight vertical production wells were drilled. In 1988, initial gas reserves were estimated to be 6.2 BCF
OGIP with 5.7 BCF recoverable. Despite these early predictions, the field has produced approximately 9.1
BCF of natural gas and the reservoir has maintained pressure at ~ 935 psia, which suggests there are still
untapped gas resources within the EBGF.
It was initially believed that continual re-pressurization of the reservoir was caused by an active water
drive. However, no significant water has been produced, even from down dip producers. Reservoir
modeling was performed to better understand reservoir drive mechanisms and to predict future reservoir
behavior. Recent modeling and reservoir simulation strongly suggest that gas hydrate dissociation and
negligible water influx is responsible for the constant re-pressurization within the EBGF. Recent drilling
and production shows that gas production from the EBGF depressurizes the reservoir and raises the base
of the gas hydrate stability zone, unlocking free gas from the hydrate bearing reservoir rock. This is a
significant finding, and validates the presumption of commercial gas production from a hydrate resource.
This paper summarizes the technical studies conducted to understand the reservoir behavior and the
evidence supporting ongoing dissociation and production of methane hydrates from the EBGF.
Field Background
Barrow, Alaska is the northern-most city in the United States. This community of ~ 5,000 people is
located on the coastline of the Arctic Ocean, 320 miles above the Arctic Circle. Securing a reliable energy
source in remote Alaskan villages is an essential living necessity. Barrow depends on natural gas to heat
homes, schools and for everyday electricity generation. Current energy demands in Barrow and the
surrounding communities are met by natural gas supply from Barrows two major gas fields: Walakpa Gas
Field and East Barrow Gas Field (EBGF). Barrows third major gas field: South Barrow Gas Field, is now
nearly depleted and is recommended for use as a subsurface gas storage tank. To secure long term gas
production, Barrow must explore economic ways to sustain production from the methane hydrate reservoir
located in the EBGF.
Remaining Reserves Study for the EBGF
Drilling and testing of the EBGF began in 1974 and continued throughout 1990. During this time, eight
vertical production wells were drilled. To date, the EBGF has produced ~ 9.1 BCF of natural gas, well
SPE-180476-MS
above the initial estimate calculated by Allen and Crouch (1988) of 6.2 BCF OGIP with 5.7 BCF
recoverable. A recent reserves study was conducted in 2007 (Stokes and Walsh) in which reservoir
pressure was observed to be 935 psia. Based on continued reservoir re-pressurization, the EBGF shows
the classic P/Z material balance traits of a strong water influx, or other energy source maintaining high
reservoir pressure as gas is produced (Figure 1). Evidence refuting presence of a strong water drive
include lack of significant water production from the EBGF (1,694 bbls., as of 10/20/2015), which is not
consistent if a strong water influx is responsible for the pressure behavior (Stokes and Walsh, 2007).
Previous estimates expected that the wells would be watered out by now due to water encroachment (Allen
and Crouch, 1988). This suggests contribution from an unknown reservoir drive mechanism, possibly gas
recharge from dissociating hydrates.
Figure 1EBGF reservoir performance plot (modified from Stokes and Walsh, 2007)
A more in depth material balance study of the EBGF was conducted by Singh in 2008. The objective
of the material balance study was to determine the behavior of the EBGF under various reservoir drive
mechanisms and to confirm a reservoir mechanism responsible for the pressure recharge within the field.
The material balance calculation (MBC) was carried out using four distinct scenarios (Figure 2) (Singh,
2008). The objective of the four models was to consider reservoir pressure in relation to the production
history data (2007). The first model considered the response of a volumetric gas tank. The second model
shows a performance study to include the effect of a water influx reservoir mechanism in communication
with a volumetric gas tank. Third, the MBC was refined to include a hydrate tank attached to a volumetric
gas reservoir. In conclusion, individual effects of previously calculated water influx and hydrate mechanisms were evaluated. Both models captured the reservoir pressure response with respect to time, while
in contact with the free gas tank.
SPE-180476-MS
Figure 2An explanatory schematic of the suspected reservoir drive mechanisms in the EBGF (Singh, 2008)
The EBGF MBC study produced compelling results. The volumetric gas tank model compared against
available production data obtained a maximum error of 20%, over predicting pressure drop with gas
offtake. Results of this initial model confirmed the EBGF was not solely under the influence of a
volumetric gas expansion drive mechanism. The performance study of the water influx reservoir
mechanism established two positive pressure support methods (Figure 3) (Singh, 2008). The first was a
small but constant pressure support mechanism, dominated by gas expansion, supplemented by a very
minor water influx component. The second added pressure support was believed to be caused by increased
aquifer activity, or hydrate dissociation (Singh, 2008). The refined MBC model that included a hydrate
tank attached to a volumetric gas reservoir implemented varying hydrate thicknesses. The calculated
reservoir response showed an increase in pressure corresponding to an increase in the hydrate zone
thickness. The effect of water influx and hydrate driven mechanisms were also evaluated. Independently
increasing aquifer strength and hydrate zone thickness failed to match production history data. It is
hypothesized that a combination of both hydrate and water influx mechanisms would bring simulated
reservoir performance closer to that observed in the EBGF (Singh, 2008). The MBC does not have a
provision to include the effect of both hydrate and water influx driven mechanisms in the same model.
Hence, the combination of both reservoir mechanisms will be discussed later.
SPE-180476-MS
Figure 3Performance study of the water influx reservoir mechanism (Singh, 2008)
SPE-180476-MS
The remaining reserves study, MBC study and gas hydrate stability modeling validate the assumption
that constant pressure recharge and excess gas production history from the EBGF results from a
combination of hydrate dissociation and minimal water influx. A state of the art 3D simulation study was
implemented in Computer Modeling Groups-Steam, Thermal and Advanced Reservoir Simulator (CMGSTARS). STARS is designed to simulate complex oilfield production and enhancement processes.
STARS also allows multi-phase multi-component fluid flow while accounting for heat losses, dynamic
rock and fluid properties, and many other applicable features. CMG developed STARS primarily for
conventional black oil studies. However, to simulate gas production from hydrate dissociation, adjustments were made to accommodate hydrate properties (Singh, 2008). In this simulation study conducted
by Singh (2008) the hydrate phase was considered as an oil phase with very high viscosity (1,000 cP) and
suppressed mobility. Reservoir models for the EBGF were built in ROXAR RMS commercial modeling
software (Panda and Morahan, 2004) and were imported into STARS to create a reservoir model of the
EBGF (Figure 5). The main objective of the CMG-STARS simulation study was to develop and compare
historical production data and to advocate the best case scenario for future production results (Singh,
2008).
SPE-180476-MS
Figure 5EB reservoir grid top model created in CMG-STARS (Singh, 2008)
Reservoir simulation and modeling of the EBGF was conducted for six various cases: (1) gas only; (2)
gas aquifer; (3) hydrates gas aquifer; (4) hydrates only; (5) hydrates gas; and (6) hydrates
aquifer (Singh, 2008). To accurately represent the EBGF, a uniform hydrate saturation value of 31% was
used. Simulated reservoir performance was also compared with historical data, as history matching at both
the reservoir level and well level was implemented to replicate EBGF reservoir conditions. Initial
reservoir conditions incorporated into the model are displayed below (Table 1). Once an accurate
simulation model was achieved corresponding to actual EBGF conditions, individual well-level matches
were performed for the best case scenario. Five gas production wells (E.B. 14, E.B. 15, E.B. 18, E.B. 19
and E.B. 21) were integrated into the simulation model. Gas production from the five initial wells began
in 1981; currently E.B. 14 is the only legacy well capable of producing.
Table 1Initial reservoir conditions used for the EBGFs simulated production forecasts (Singh, 2008)
Zone
Hydrate (top)
Pressure
Temperature
975 psi
Saturation
Sh 31%
Sg 14%
Sw 55%
Sg 45%
Sw 55%
Sw 100%
Remarks
(Lw-H-V) close to Equilibrium
*HGC-2050
**GWC-2080
Initial estimates of hydrate-associated gas, free gas and total OGIP were simulated for each reservoir
model (Table 2) (Singh, 2008). The EBGFs historical results for cumulative gas production, cumulative
water production, average reservoir pressure, gas production rates and water production rates successfully
match simulated results for case C. The hydrate gas aquifer reservoir model was determined the best
fitting model to match reservoir performance associated with the EBGF (Singh, 2008). Modeling through
CMG-STARS showed hydrate dissociation from depressurization with minimal water influx was capable
of supplying the necessary pressure support to the free gas zone underlying the hydrate layer. The dynamic
reservoir model validated the assumption that the free gas pool was in communication with a thick layer
SPE-180476-MS
of hydrate-bearing reservoir, supported by weak water influx from the bottom (Singh, 2008). The modeled
change in the GHSZ from 2,050 2,045 TVD reflects the dissociation of gas hydrates at the free gas
interface. A revised OGIP computation, taking into account the presence of gas hydrates produces an
estimate of ~ 42 BCF (Singh, 2008), a 677% increase from Allen and Crouchs (1988) initial OGIP
estimate.
Table 2EBGF simulated results representing all six scenarios calculating OGIP (Singh, 2008)
Case
A
B
C
D
E
F
Reservoir Model
Hydrate-Associated
Gas (bcf)
Only Gas
Gas Aquifer
HYD Gas Aquifer (Best-Case)
Only HYD
HYD Gas
HYD Aquifer
00.00 bcf
00.00 bcf
26.13 bcf
147.62 bcf
26.10 bcf
42.80 bcf
Free Gas
(bcf)
94.21
27.32
15.83
29.31
82.62
08.49
bcf
bcf
bcf
bcf
bcf
bcf
Based on positive results of the described desktop studies, it was proposed by the research team to
design, permit and implement a dedicated methane hydrate drilling and production testing program in the
EBGF (Stokes and Walsh, 2007). High angle gas producers were recommended to allow for additional
production capacity, facilitate characterization of the postulated methane hydrate accumulation, mitigate
well production hydrate issues and lessen associated temperature drop as compared to conventional
vertical producers (Stokes and Walsh, 2007). In 2011, the North Slope Borough approved drilling for two
high angle gas producers (Savik 1 & 2) in the EBGF, although funding for the research effort was
curtailed.
SPE-180476-MS
Figure 6 Savik 1 surface location, well trajectory, nearby wells and seismic lines (McMullen et al., 2011)
Savik 1s planned total depth was not reached due to drilling issues, including gas influx, foamed mud
and fluid losses occurring while drilling (McMullen et al., 2011). Gas influx and foamed mud can be
associated with gas hydrate dissociation. Hydrates dissociate when the equilibrium state is broken, due to
either an increase in temperature or a decrease in pressure. Once the dissociation process begins, mud
weight increases become ineffective as lost circulation and gas cut mud tend to occur. Fluid losses
occurred as the wellbore intersected near or at the top of a porous and permeable Lower Barrow Gas Sand
(McMullen et al., 2011). TD was called at 4,770 MD (2,036 TVD) with a top hole inclination of 90.
The production hole drilled was 1,517 MD, 76% of the total distance planned for the high-angle wellbore,
but more than sufficient to support target production expectations. The Savik 1 wellbore accomplished
58% of the production hole in the Upper Barrow and 78% in the Lower Barrow Sands from the initial
planned trajectories (McMullen et al., 2011). Preliminary log analysis showed a total net pay of 1,348
MD of hydrate and gas sands penetrated, 344 MD in the Upper Barrow (mostly hydrates) and 1,004MD
in the Lower Barrow (free gas) (McMullen et al., 2011).
Savik 2 was the second high angle gas well drilled, designed to tap reserves along a southwest
trajectory in the EBGF. Savik 2 is located 14 west of Savik 1 at the surface. Remaining reserves along
the well path were determined from pre-existing wells where the Barrow Sands had not been efficiently
drained or formation damage had occurred, suggesting sufficient incremental reserves remain. Structural
and stratigraphic controls adopted during well planning and while drilling were supported by 2-D seismic
lines and three pre-existing vertical wells (Figure 7).
Figure 7The Savik 2 surface location, well trajectory, nearby wells and seismic lines (McMullen et al., 2011)
SPE-180476-MS
Savik 1 provided valuable drilling and steering information that was incorporated in drilling Savik 2.
Precautions were taken while drilling to reduce mud temperatures to 50F by opening the mud pit to
ambient temperature. However, while drilling the Upper Barrow interval, mud weight/ECD continually
increased which caused drilling to stop periodically to lower the mud weight/ECD.
TD was called 700 MD short at 4,720 MD (2,008 TVD), with a top hole inclination of 92.2. TD
was called early due to the following reasons: 1) Repeated efforts to reduce mud weight/ECD was
unsuccessful; 2) Time constraints to complete the 5-well drilling program were in effect; 3) The Upper
and Lower Barrows Sands were successfully penetrated; 4) Abundant pay was calculated over 1,256 MD
of the reservoir interval; and 5) The inclination and/or mud weight/ECD issue was projected to cause
further drilling delays (McMullen et al., 2011). Importantly, the production hole drilled was 1,397 MD,
or 70% of the total distance initially planned for the high-angle wellbore. Savik 2s production hole
accomplished 83% of the wellbores planned distances in the Upper Barrow Sand and 53% in the Lower
Barrow Sand. Upon completion it was determined that free-gas net pay was lower than expected since
substantially more hydrates were found than initially expected (McMullen et al., 2011). Based on
preliminary log analysis, a gross sand total of 890 MD of hydrate and gas-filled sands were successfully
penetrated between the Upper and Lower Barrow Sand intervals.
Well Clean Up and Production
After drilling and completion, the Savik 1 and 2 wells were cleaned up and flowback tests were performed
to determine whether liquid buildup was present in either well. Savik 1 cleaned up nicely as the
completion fluids (freeze protectant diesel and brine) flowed through the annulus and to surface. Gas
injection through the production tubing assisted in continual cleanup of the well. Savik 1 was promptly
cleaned of any existing fluids and well head pressures (WHP) recharged efficiently. In the following
months, initial flow tests were performed at Savik 1 for a sixteen day period. The well produced at 800
MSCF/D while maintaining sufficient WHP. Liquid loading was implicated at the end of the initial flow
test. A four point test was then conducted by producing through the annulus while accelerating production
rates and observing WHP. AOF was calculated to be 9.9 MMSCF/D for Savik 1. A conservative well
cycling procedure was later initiated to recover remaining wellbore fluid. Methanol injection was also
performed to promote hydrate dissociation near the wellbore. This process was repeated until no fluids
were recovered and Savik 1 was considered a prolific gas producer, capable of rates tenfold greater than
conventional vertical gas wells in the EBGF.
Savik 2 was also cleaned up as completion fluids were recovered with the assistance of gas injection.
Unlike Savik 1, Savik 2 had complications maintaining WHP during flowback tests as tubing and annular
pressures rapidly blew down, resulting in very low stabilized pressures. The well was shut-in after an
insufficient WHP recharge rate was observed.
Due to inadequate buildup pressures in the tubing and annulus, a hydrate plug was suspected to be
located down hole below the tubing tail at Savik 2. The well was bled down several times in an attempt
to dissociate the hydrate plug. This resulted in liquid buildup in the tubing. The well was then shut in
overnight after 60 gallons of methanol was put into the tubing and annulus. The following day the well
was flowed up through the tubing and annulus for 40 minutes before losing pressure. The well was again
bled down through the tubing and both the tubing and annular pressures moved rapidly toward 150 psi,
which indicated that the blockage was still intact.
Next the well was treated with two hot diesel treatments. The well was flowed back after each
treatment with most of the diesel being recovered, but no gas flow was observed from the formation. It
was concluded that the diesel was going into the Upper Barrow Sand above the suspected hydrate plug.
After numerous unsuccessful attempts to bring Savik 2 online, a coil tubing unit (CTU) was brought
onsite to clean out the blockage to the base (toe) of the production liner by methanol circulation. The CTU
jetted out the hydrate plug with methanol and 9.3 ppg KCl brine and flowed with Nitrogen Gas for the
10
SPE-180476-MS
night to lift the fluids from the toe of the well. The CTU job was successful in removing the suspected
hydrate plug at 3,960 MD from surface (Figure 8). Small amounts of gas flowed through the wellbore
and to the surface, but only for a brief period of time as the well froze up shortly thereafter. The continued
plugging of Savik 2 suggested that the production liner rests entirely in the GHSZ, which allows no fluids
to flow freely from the reservoir.
Figure 8 Cross-section diagram of the hydrate plug blockage in the Savik 2 wellbore
It is important to note that the deepest vertical depth reached at Savik 2 was 20 shallower than the
shallowest depth of the production liner at Savik 1 (Figure 9). Savik 2s well placement in the GHSZ is
supported by previous efforts to unfreeze the well. When the wellbore was stimulated with methanol and
assisted with a CTU Nitrogen Lift, it showed a short term positive impact on gas flow and pressure
response from the reservoir, only to revert back to its original frozen state shortly thereafter.
Figure 9 Cross-section diagram showing depth relation of the two Savik wellbores
SPE-180476-MS
11
Upon completion, Savik 1 was approved to be tied into the EB Pipeline system to accommodate
Barrows growing energy demands. Savik 1 was brought online (3/20/2012) and produced ~ 500 MSCF/D
for three months, before being shut in until November 2014. More recently (11/1/2014 - present) Savik
1 has been producing at ~ 200 MSCF/D. Limited production rates are attributed to Barrows current
reduced energy demand as the majority of gas is supplied by the Walakpa Gas Field.
Savik 2s production liner is believed to reside in a hydrate zone with the surrounding rock pore space
filled with methane hydrates rather than free gas. A substantial amount of effort and cost has been spent
treating this well with the objective to achieve consistent flow of natural gas to surface. These efforts
proved to be unsuccessful. The well was shut-in and incapable of producing, and the recommendation was
made to produce the EB-14 and Savik 1 wells at rates intended to lower average reservoir pressure in the
field, in an effort to dissociate hydrates at the base of the MHSZ, with the expectation of eventually
unlocking the Savik 2 well.
Savik 2: Breaking Hydrate Equilibrium Conditions
In July, 2014, WHP at Savik 2 spontaneously rose from 650 psi to 850 psi, without stimulation or
enhancement. Constant gas production from the EBGF is thought to have started the depressurization
process by unlocking free gas from the hydrate-bearing reservoir rock and further raising the base of the
GHSZ. Following the sudden WHP build-up, the well was production tested, and commercially viable
rates of gas flowed freely to surface for the first time in Savik 2s history. A blowdown conducted at Savik
2 showed promising results in the wells ability to produce natural gas. The well flowed up the annulus
and the gas was measured at atmospheric conditions at an average rate of 1.2 MMSCF/D. The well was
shut in after 1.5 hours of production, accumulating ~ 75 MSCF of gas. Savik 2 did not show any signs
of wet gas, fluids or freeze up. The WHP and build-up rates post shut in can be seen in Table 3 below.
The WHP post flow returned to the pre WHP conditions indicating good recharge and that the well could
now be connected to the free gas pool (Figure 10).
Table 3Displaying WHP build-up rates for the blowdown conducted on 7/9/2014
Date
Comments
8/18/12
7/9/2014
7/9/14
7/9/14
7/10/2014
7/22/2014
590
850
623
725
779
860
67.5
4.5
0.283
12
SPE-180476-MS
Figure 10 Chart displaying WHP build-up rates ensuing Savik 2s blowdown (7/9/14)
Savik 2 demonstrated similar build-up rates after the previous flowback (7/9/14) compared to the WHP
build-up rates subsequent to the CTU Nitrogen Lift (8/18/12) (Figures 11 & 12). It is important to note
that Savik 2s WHP build-up to ~ 850 psi is believed to reflect hydrostatic pressure gradient. Hydrostatic
pressure is an indication that gas has migrated into the wellbore, displacing the fluid and creating a full
gas column in the wellbore. Slower WHP build-up rates could be attributed to the connection between the
Savik 2 wellbore and the free gas pool. In-situ methane hydrates could still be responsible for blocking
the Savik 2 wellbore from the free gas zone.
Figure 11Chart comparing the annular pressure build-up data from the flowback (7/9/14) and the CTU Nitrogen Lift (8/13/12)
conducted at Savik 2
SPE-180476-MS
13
Figure 12Chart comparing the casing pressure build-up data from flowback (7/9/14) and the CTU Nitrogen Lift (8/13/12) conducted at
Savik 2
14
SPE-180476-MS
Conclusion
Ongoing gas production from the EBGF is believed to have started the depressurization process and
further raised the base of the GHSZ. Detailed MBC, hydrate stability, and CMG-STARS modeling was
accomplished to accurately match current EBGF reservoir conditions. Through modeling and simulation
it was hypothesized that gas hydrate dissociation and minimal water influx was responsible for the
constant pressure recharge mechanism within the EBGF. As a result, two high angle gas producers were
drilled in the hydrate-rich, Barrow Gas Sands. Well logging and drilling operations confirmed a
significant presence of hydrates within the field. Recently, both of EBs high angle gas producers have
shown successful production of free gas from EBGFs challenging but prolific hydrate reservoir. Efficient
gas production from the EBGF will help supply Barrows growing energy demands for years to come.
Acronyms
AOF
bbls.
BCF
CMG-STARS
CSM-HYD
CTU
EB
EBGF
ECD
GHSZ
KCl
MBC
MD
MSCF/D
MMSCF/D
NaCl
OGIP
ppg
TD
TVD
WHP
References
Allen, W.W. and Crouch, W.J. 1988. Engineering Study of South and East Barrow Fields, North Slope Alaska, Alaska.
Technical report prepared for North Slope Borough Gas Development Project, Barrow, Alaska: 156.
Lau, J. 2012. History of Drilling Operations: East Barrow Gas Field Savik Well No. 1. Technical Report Submitted to the
North Slope Borough, Barrow Alaska: 20 29.
Lau, J. 2012. History of Drilling Operations: East Barrow Gas Field Savik Well No. 2. Technical Report Submitted to the
North Slope Borough, Barrow Alaska: 2237.
McMullen, B., Heinrich, A. and Roberts, D. 2011. Savik #1: Geology Operations Report. Technical Report Submitted to
the North Slope Borough, Barrow Alaska: 18.
McMullen, B., Heinrich, A. and Roberts, D. 2011. Savik #2: Geology Operations Report. Technical Report Submitted to
the North Slope Borough, Barrow Alaska: 18.
Panda, M.N. and Morahan, G.T. 2008. An Integrated Reservoir Model Description for East Barrow and Walakpa Gas
Fields. Topical Report prepared under DoE Project No. DE-FC26-06NT42962 submitted to the USDoE, Morgantown,
WV, www.netl.doe.gov: 144.
SPE-180476-MS
15
Singh, P.K. 2008. An Engineering Study to Investigate Methane Hydrate Resource Potential Associated With Barrow Gas
Fields, Alaska, MS Thesis, University of Alaska Fairbanks.
Sloan, E.D. and Koh, C.A. 2008. Clathrate Hydrates of Natural Gases. Taylor and Francis Inc.: 721p.
Sloan, E.D. (1998). Clathrate Hydrates of Natural Gases. Marcel Dekker: 705p.
Stokes, P.J. and Walsh, T.P. 2007. South and East Barrow Gas Fields Reserves Study. Technical Report Submitted to the
North Slope Borough, Barrow, Alaska: 136.
Walsh, T.P., Stokes, P.J., Panda, M., Morahan, T., Greet, D., Singh, P. 2008. Phase I Final Technical Report:
Characterization and Quantification of the Methane Hydrate Resource Potential Associated with the Barrow Gas
Fields. Technical report prepared for the, U.S. Department of Energy: 146.