.

SPE 50081
F!!l! D

. .

Society of Petroleum Engineers

HIGH RATE GAS WELL DESIGN: ISSUES AND SOLUTIONS

— GOODWYN GAS CONDENSATE, NWS, AUSTRALIA
D. TENG, G. NE~LESHIPl S. HICKING, K. HINDMARSH

Cop@ght1998, .Sceiety of Petroleum Engineers, Inc. The Goodwyn reservoir structure is a large tilted fault block
This papsr was prepared for presentation at the 18?38 SPE Ash Pacifc Oil & Gas @“nference that is open to the north and bounded by major faults to the
and Exhibtin held in Perth, Austmlkt 12–14 October 1SS8.
soutlL west and south-east (Fig. 2). The bulk of the reserves
This paper was selected for presentation by sn SPE Program Cmnm”Ree fdlowfng review of are within the GE and GG sand units. Condensate gas ratio
infmmation contdned in sn ststrsct submitted by the author(s). Contenk of the paper, ss
presented, have not been reviewed by the Sodety of Petroleum Engineers and sre subject to (CGR) increases from south to nom varying from 45
mrrection by the atthor(s). The material, as presented, dces not new=arily reflect anY
Pesfion of the Scciety of Petroleum Engineers, ik officers, or members. Papers presented at
BBL/MMscf in the southern GG unit to 123BBL/MMscf in
SPE meetings are subject to publication review by Ed@rial Cwnmittees of the .Smiety of the northern GD unit. Permeability of the Goodwyn sands
Petroleum Engineers. Ekctronb reprcduciion, distrtsition, or storsge of sny part of this paper
for cwrnmem”sl PUPS without the kwitten mnsent of the Sceiety of Petroleum Engineets is range between 1000 to 3000 md The original Phase I field
prohibtied. Permission to reproduce in print is restricted to an abatrsct cf not more than WI
words iluatrations may not be wpiad. The abstrsct must contdn conspicuous
development proposed 15 wells to develop the main GWGF
scknowfedgment of where and by whcm the paper ws presmtad. Wtite Librarian, SPE, P.O. sand units with the higher CGR gas. This has now been
Box 833835, Richardson, TX 75083-3336, U.S.A., fsx 01-972-952-9435.
accomplished with 11 wells, 7 of which are producem and 4
are currently injectors. Two production wells have also been
drilled to access the further GG and the GD sand units.
Abstract The main export from the Goodwyn ‘gas platform’ is
Goodwyn platform wells produce up to 220MMscf7d of raw currently 100,000 bpd of stabilised condensate. Separated gas
gas yiel~ng UP to 20,000 bpd of stabilised condensate. The is mostly reinfected to sweep the rich gas and maintain
high productivity of the wells has reduced the number of voidage replacement. Dry gas will be produced during the
development wells needed to satistJ platform capacity with blowdown stage.
significant cost savings. This high well capacity was achieved
through the early recognition of the reservoir inflow potential Well Design Considerations
in the completion design and addressing the critical issues of The design philosophy for Goodwyn wells is to maximise the
erosional velocity, Iifecycle sand production, sub-surface and net present value over the lifecycle of the integrated
surface equipment integrity in a high flow, high stress production system. This is achieved by designing wells with
environment. To maximise the well potential the erosional high productivity, but that can also be operated safely and
limit was extended based upon actual experience. The risk of reliably. For high rate gas wells there are obvious technical
sand production was managed through a philosophy covering problems such as sand production erosion and corrosio~ but
sand prediction and its control during operation. Well in an integrated system there are other issues that need to be
hardware was tightly specified and qualified within the taken into account. This is achieved by the production
defined operating envelope. These solutions were applied to technologist working with a multidisciplinary team that
both the initial conventional wells and for the later extended includes reservoir, process, drilling and completion engineers.
reach horizontal wells. Results fkom Goodwyn wells These ‘external’ issues are discussed below.
demonstrate that actual performance match the targeted ideal
and that the issues have been managed. Economic Considerations. Increased well capacity has clear
This paper presents a systems approach to well design benefits – more production or injection per well leading to
where technical and external issues are taken into reduction in well numbers. It can also increase the capacity
consideration. It is a retrospective look at the issues addressed ramp up Me for anew facility, reaching maximum production
for Goodwyn well desi~ with updates to include current earlier. For gas wells, larger well capacity is valuable in future
techniques. when the reservoir pressure is lower, when flow velocities are
higher and fictional pressure losses in the tubing become
Introduction more significant.
The Goodwyn gas field lies on the North West Shelf of However, the pursuit towards increasing gas well capacity
Australia 23km south west of the North Rankin ‘A’ platform should be driven by robusthisked economics. Large capacity
and 138km offshore from Karratha in 13Im of water (Fig. 1). wells may present higher integrity and reliability risks if non-

281
 2 D. TENG, G. NETTLESHIP, S. HICKING, K. HINDMARSH SPE 500S1

standard or unproven equipment is used. Where possible it is much better to let the gas come to the well
A simple indication for comparative purposes is the cost rather than the well going to the gas. The short well is cheaper,
per unit capacity. The incremental unit cost for increasing well more productive and has a lower cost of capacity.
capacity should be lower than the base unit cost, particularly
when technical risks are taken into consideration. However, Top-sides Considerations. Well design does not stop at the
higher unit cost may be justified in situations where the wellhead; it has to be integrated into top-sides design. Flow-
number of available well slots is limited or by reduced safety lines, chokes, pressure relief valves and venting capacities
expomre. should be matched to well capacity. There is little point in
~High rate wells may raise operational concerns as platform gaining well capacily then limiting its impact by having under-
capacity is reliant on fewer wells. Down-time, for example sized top-sides equipment. Benefits of larger flow-lines and
during well surveillance and integrity maintenance activities, chokes or even connecting two flow-lines to one wellhead
has increased impact upon production. However these issues should be considered. Such measures should be identified
are usually minor when compared to the rate benefit from the early, as retro-fitting into an existing system ean be restrictive
well. and takes time to implement. The upper limit to Goodwyn well
capacity is defined by the rating of the flare system and relief
Reservoir Engineering. Prior to embarking on well desi~ it valves on the production train.
is essential that reservoir requirements be taken into
consideration. There is no point pursuing increased well Recognition of Well Potential
capacity if the reservoir requires more wells to access separate It is essential that potential well capacity be recognised at an
blocks. For (-lood~ reservoir engineers identified main sand early stage of the field development. Data from appraisal well
blocks that could be recycled and eventually drained with a test results may under-estimate potential well capacity,
minimum number of wells. Henee there was value in resulting hi over-estimating the number of well slots required,
increasing capacity of wells into these main sand bodies. an expensive item for platform jackets.
For reservoir management it is necessary to balance well
capacities of producers and injeetors in a recycling scheme to Appraisal Well Test Results. Results of exploration/appraisal
maximise condensate reeovery. If insutlicient gas is injeete~ well-tests should be used carefully as they may not be
recycle sweep efficiency is reduced and the dewpoint pressure representative of what can be achieved from production wells.
will be reached earlier, dropping out and leaving condensate in On Goodwyn appraisal well-tests, the combination of 5“ drill
the reservoir. Optimum pmductionlinjection rates should be pipe for the test string and the limited rig thing capacity
determined from a good reservoir simulation model. restricted appraisal flow-rates to between 11 and 53 MMscf/d.
Water coning due to high production rates should also be Well inllow performance from inteqxetation of three
considered in the completion design. This was identified as a Goodwyn appmisal well tests can be seen in Figure 4. The
potential problem for Goodwyn flank wells that are extended extrapolation of these in-flow petiormamx curves are shown
reach and are by necessity horizontal through the reservoir. intersecting the 7“ and 7-5/8” tubing performance curve of a
The extended reach well has a two-fold effect; the horizontal typical Goodwyn well, indicating that these wells would
completion reduces drawdown while the long length of the produce between 60 and 120 MMsef/d. It also shows that the
well limits well capacity. Both act towards reducing the risk of incremental gain from increasing tubing size from 7“ to 7-5/8”
water coning. is between 2 to 11 MMsct7d. Typical ‘pseudo-pressure’ B and
Data gathering requirements (Ref. 1) can also have an F values for appraisal wells are of the order of 300,000
impact. Cables for permanent down-hole pressure gauges that psi2/cP/MMscf/d and 12,000 psi2/cP/(MMscf7d)2 respectively,
run down the length of the well can limit tubing size. In while the absolute open-hole flow potential is 300 MMscf/d.
Goodwyn wells where downhole pressure is required, 7“
tubing is run as 7-5/8” tubing with a cable cannot fit inside the Ideal Inflow Performance. Ideal intlow performance was
9-5/8” casing. The resulting 20% decrease in well capacity has modelled based on Goodwyn reservoir properties, best
to be weighed against the value of information from the down- petioration praetiee and minimal formation damage. This was
hole gauges. validated against operating data from the geographically
An important issue is the along-hole length of the well similar North Ran& gas production field. A comparison
which affwts its capacity, its drilling cost and hence the cost between ideal and appraisal Mow pefiormance cau be seen in
of capacity. These values are plotted against the along-hole Figure 5.
distance to top perforation in Figure 3. Note that costs are The difference in draw-down embles a dramatic increase
relative in this figure. The shortest reach well has a capacity of in production rates. With 7“ tubing in a typical Goodwyn well
220 MMsct7d raw gas while the capacity of the longest reach configuration%well capacity increased 67% from 85MMsct7d
well is 120 Mh4scf7d raw gas. As the planned cost of the using appmisal Mow performance to 142MMsct7d for the
longest well was twice that of the shortest well, the capacity ideal well. The incremental benefit of 7-5/8” over 7“ tubing
cost increased four-fold, an ‘exponential’ effect. Problems increased to 28 MMM7d, making the pursuit of this size more
encountered while drilling the longest reach well caused its worthwhile.
actual capacity cost to be six times that of the shortest well. The poor performance of the appraisal wells can be

282
SPE 50081 HIGH RATE GAS WELL DESIGN ISSUES AND SOLUTIONS - GOODWYN GAS CONDENSATE, NWS, AUSTRALIA 3

accounted by poor petiorating, formation damage and surface operations could cause damage. At the time, risked economic
equipment limitations. Goodwyn appraisal wells were drilled analysis did not just.@ the benefits of plastic coated tubing.
in the rnid-80s, using water-based mud and were pertlorated
under-balanced with 2-1/8” strip guns at 4 spf. The Internal Smoothing. Additional smoothing of the internal
combination of significant formation damage, zero degree surface can be achieved by speci@ing a finer shot-blast finish
phasing and low shot density of the strip guns restricted the at the factory. This method does not offer as smooth a finish as
flow of gas into the well-bore. This created non-Darcy or internal coating, but it is a simple very low-cost option. It can
flow-dependent effeets that become dominant at high rates. reduce roughness down to 20 microns, increasing production
Figure 6 and Table 1 shows the split between non-Darey and by 2’XO for 7“ tubing and l% for 7-5/8” tubing. Again it is not
Darcy contributions to drawdown. known whether the smooth ftish of 13Cr tubing stays in this
state. Shot-blast smoothing of tubing will be used in fhture
Goodwyn wells. It should be noted that there is less-benefit
fkom reduced surface roughness for larger tubing sizes.

Liquid Loading
~ Increasing tubing size raises the minimum production rate of a
Table 1: Drawdownsat 120MMscfldfor appraisaland ideal wells. well. As-well flow-rate decreases, the gas veloeity decreases,
reducing the liquid carrying capacity such that it can no longer
To overcome the pronounced effect of non-Darey flow in lift water or condensate out of the well. Liquids fiis to the
gas wells, it is neeessmy to speci@ the best performing bottom, accumulates and eventually kills the well. The flow-
pefioration charges, run at high shot density with rate at which gas wells load up is dependent on the largest ID
cireumferentkd phasing. Goodwyn wells were perforated with in the completion and the highest density liquid being
4-1/2” guns at 12 spf, 45 degrees phasing and an underbalance produced. In the case of conventional Goodwyn wells, the
of 900psi. The first Goodwyn wells were perforated with largest ID is the 9-5/8” production casing and for the
charges that had an API RP43 seetion I penetration of 17.93”, horizontal wells it is either 7-5/8” or 7“ tubing. Condensed
which under down-hole conditions was predicted to give 7.1” water is the highest densi& liquid produced. The Turner
of formation penetration. Charges with an API RP43 Section I method (Ref. 7.) and the Kutateladze method were used to
penetration of 28.57” and 8.4” formation penetration were calcuIate the flow-rate at which the well loads up. This can be
used on later Goodwyn wells when these beeame available. seen below in Table 2.

Actual Well Performance. Most Goodwyn wells achieved

ideal well performance, indicating that the measures taken to
r= Turner Load-up
Flow-rate
Kutateladze Load-
up. Flow-rate

r
optimise inflow petiormance were successfid. The actual (MMscfYd) (MMscf7d)
performance of Goodvvyn-02 matched the ideal intlow 5-1/2” 7.2 6.1
performance shown in Figure 5 and Table 1. ‘Pseudo-pressure’ 7,,
B and F factors from production tests were reduced to
typically 100,000 psi2/cP/MMsct7d and 250
7-5/8”
9-518”
11.6
13.9
22.8
I 9.7
11.6
19.1
I
psi2/cP/(MMscf7d)2, while the absolute open-hole flow Table 2: Goodwyn load-up flow-rates for different tubing
potential increased to 1,000 MMscfld. It should be noted that sizes.
the best performing well Goodwyn 03 produces at 220
MMsctYdraw gas with a 20 psi draw-down. These minimum rates are acceptable to produce the
majority of the field reserves. If the predieted load-up flow-
rate had been very him for example 50 MMdd, tubing and
Smooth Tubing casing sizes would need to be re-examined. Siphon strings
In addition or as an alternative to increasing tubing size, well may be required to produce below these rates when reservoir
eapaeity can be increased by reducing tubing internal surface pressures have declined.
roughness. There are two methods to achieve this:
Technical Risks
Plastic Coating. Ceramic or polymer coating reduces surface Erosional Velocity Limits - WMhout Sand. Erosional
roughness from typically 35 microns for 13% Chrome (13Cr) velocity limits as applied to oil and gas wells have been the
steel tubukws down to 5 microns, potentially improving subject of many papers. Referenee 2 provides a review of the
production by 6% for 7“ tubing and 3% for 7-5/8” tubing work done to date by various operating companies in this area.
under Goodwyn conditions. Plastic coatings were originally The relationship as defined in API RP-14E (Ref. 3) has
developed to offer corrosion protection to carbon steel tubing. traditionally been used to define veloeity limits in carbon steel
Use of internal coating on chrome-alloy tubing is less common pipe and in flowlines with bends. It is widely acknowledged
so durability for a 25-year well life could not be established. that this criteria is conservative in oil/gas wells particularly
There was also concern that repeated wireline intervention

283
 4 D. TENG, G. NETTLESHIP, S. HICKING, K. HINDMARSH SPE 50081

when completed with Corrosion Resistant Alloy (CRA) results. While these may indicate expected wear trends,
tubing. Reeognising this, the constant ‘C’ in the API formula there may be areas in the completion or flowline that are
was revised in 1991 tiom 100 to between 150 and 200 for subject to much higher localised erosiom and may not be
CM tubing. ‘C’ used by different oil and gas operating easily detectable.
companies range upwards from 100 to 500 (or none at all), ● In high velocity gas wells, unmanaged sand production
with the majority adopting between 200 to 300 in solids-free can very quickly cut out flow-lines or wellheads, with
non-corrosive environment. These erosional velocity limits do possibility of catastrophic consequences.
not apply in the presenee of solidskmd production or in a ● Most allowable sand limits are based upon a metal loss of
corrosive environment. 0.25 mm/year. Such a metal loss rate is unacceptable for
Experience from the North Rankin field gave the wells designed with a 25-year Iitie.
confidence to fimther extend erosional velocity limits. Kinley ● Even small amounts of sand for example 1 kg/day, can
caliper surveys did not indicate any metal loss fi-om the 7“ accumulate in processing facilities and&come si~cant
9CrlMo tubing in a well that had been producing over 200 with time. Although there is some provision for water-
MMscf7d. This rate is equivalent to au API ‘C’constant of 240 washing Goodwyn production separato~ the operation
at a velocity of 25 rids. Further suppmt was provided by requires a shutdown. Sand can also get into valves and
physical examination of a flowtube from a wireline retrievable metering equipment, causing operational problems.
sub-surface safety valve (WRSV) retrieved from this well tier
two years serviee. The 9CrlMo steel flowtube has an internal The initial philosophy for Goodwyn wells is based on zero
11) of 3.62” and had been subjected to continuous flow continuous sand production. This will be revised if the wells
velocities of over 70 ru/s, at an effeetive ‘C’ of 630. When produce sand due to declining reservoir pressure and/or the
compared to a new flowtube the maximum possible erosion onset of water production.
rate over two years was O.17mm/year.
These results are consistent with worldwide operator Sand Prediction. Cores taken tiom appraisal wells indicated
experience, giving confidence to adopt 35 mh as the that some seetions of Goodwyn sands are friable with low
maximum allowable velocity for 13Cr tubing and 50 IU/Sas Cementation raising concerns that sand may be produced from
the maximum acceptable velocity in duplex flow-lines. Figure Goodwyn wells. For this reasoq sand prediction and its
7 shows the gas veloeity and equivalent API ‘C’ values for management was an important component of the Goodwyn
Goodwyn wells under the highest velocity conditions which is field development plan.
at the wellhead. In the 7“ tubing just below the wellhea~ the Sand strength is primarily a function of quality of the
highest rate Goodwyn well flows at a velocity of 27rn/s at an cementation between the sand grains and the effective stress in
API ‘C’ value of 270. In the 5.5” internal diameter duplex the reservoir during the production life of the well. Friable
flowline of the same well, velocity increases to 33 nds and sands are soseeptible to failure when subjected to lower
API ‘C’ value becomes 330. The first 90° elbow on each reservoir pressures. The near well bore sand failure process
flowline is regularly checked for erosion/corrosiow no metal can be described by two mechanisms:
loss has been detected to date.
. Compressive Failure – This is dependent upon the near
Erosional Limits - With Sand. The erosional velocity limits wellbore vertical effective s@ss.
defined above no longer apply when solids are produced or . Tensile Failure - This is mainly caused by draw down and
where the environment is corrosive. Various models have been flow effeets. It is also at%eeted by the near wellbore
developed within the oil and gas industry to define erosional vertieal effeetive stress.
limits with sand production. These mnge from simple
empirical (e.g. 1 pound per MMscf7d) to sophisticated Near well bore sand ftiure can result in either tranaien~
relationships which take into account sand grain size, shape, continuous or catastrophic sand production into the well bore,
quantity, impingement angle, type of material and flow depending on the vertical effective stress and the magnitude of
regime, see Ref. 2, 4 and 5. Erosional velocity limits with the draw down. Transient sand production can be handled by
sand is not a welldefined subjed, there is wide variation and surface facilities while continuous and catastrophic sand
conservatism within the models. production usually requires down-hole sand control.
It is recognised that some sand production is tolerable, but Laboratory rock strength measurements on Goodwyn core
them is a strong incentive to produce sand-free as long as plugs were used to derive the sand strength correlations
possible, for the following reasons: against wire-line logs. For compressive failure, the TWC
(Thick Walled Cylinder) strength was used in the sand
● Having adopted a maximum velocity limit that strength correlation and for tensile ftilure the UCS
specifically excludes sand production it would be (Unconfined Compressive Strength) was used. Goodvvyn core
incorrect to allow some sand production if the limit is to rock sand strength parameters were difficult to correlate due to
be valid the scatter in the core plug data. This was most likely due to
. Relationships for allowable sand are based upon empirical the variable quality of the core plugs cut from the coarse

284
 SPE 500S1 HIGH RATE GAS WELL DESIGN ISSUES AND SOLUTIONS - GOODWYN GAS CONDENSATE, NWS, AUSTRALIA 5

friable sand intervals. A log correlation was obtained for the instruments. Measurements would be performed at regular
better consolidated lower porosity sand intervals, but there intervals or immediately after significant transient sand has
were problems extrapolating and using the correlation for been observed.
muse friable sands where it is most required. . A base calibmtion of the sand detection equipment would
Using an in-house sand prediction tool (Ref. 6), the be carried out. Further edibmtion would only be
minimum sand strength that was required to prevent performed if significant continuous sand production was
continuous or catastrophic sand failure was determined. The observed. Data will be stored to enable retrospective
sand prediction tool provides a probabilistic estimate of sand quantification of sand produced.
fMhwe using the saud strength correlatio~ near well bore
stress regime, well angle and the reservoir depletion plan.
Analysis of Goodwyn log data indicated that good inflow Conventional Well Completions
pefiormance could still be achieved by selecting only
competent sands for perforation Goodwyn Completion Philosophy. The completion
To reduee the possibility of sand tensile ftiure due to philosophy is to have the highest productivity whilst
drawdo~ pressure drop into the well bore should be maintaining integrity and reliability over the wells Metime.
minimised. This requirement complements efforts taken to There will be zero intervention except for reservoir
optimise well productivity as discussed earlier. engineering requirements. To achieve this the following
hardware specifications were adopted:
Sand Management Philosophy. Measures discussed above
gave confidence that sand ffom Goodwyn wells can be . Metal-to metal seal barriers on all components
controlled and that up-front down-hole sand control would not ● Simplicity – minimum down-hole jewellery
be necessary. The following philosophy to manage sand was ● No flow-path from tubing to ammlus - No circulation
agreed for Goodwym devices or sliding sleeves were installed above the
production packer
Sand control ● Equipment with minimal inspection or maintenance
● Up-front gravel packing would not be applied as this has requirements.
every chance of severely curtailing weIl pedormance . Equipment that is qualitied and field proven
through skin and particularly non-Darey flow. ● Speei& suitable material – 13Cr material or greater was
● Through tubing screens would be ordered as a selected for all components, including the xmas tree,
contingency. tubing, safety valve, packer, landing nipples and the liner
● Drawdown would be minimised through best perforation and liner hanger.
techniques to reduce probability of sand tensile failure.
● Only competent sands would be selected for perforation Conventional Well Completion Design. The conventional
using the Goodwyn sand correlation and criteria based well completion design adopted for Goodwyn is kept simple,
upon the in-house sand pmlktion tool analysis. consistent with the philosophy above. To reduce drilling time
and well co% 10-3/4” x 9-5/8” casing was run and cemented
Sairdprodw$bn tinuii in place across the reservoir, eliminating the need for a 7“
● No continuous sand production would be allowed. liner. Atypical completion schematic can be seen in Figure 8.
● Some minor beanup and spurious spikes of irausient sand The main components of this completion (and chokes) are
production ean be tolerated, provided it is monitored. discussed in detail below.

Sand detection. Tubular Size and Connection. The first phase of Goodwyn
● Continuous sand detection monitoring is required on development drilling used Wan&d/conventional bit and
individual Goodwyn wells, backed up by pop-out erosion easing sizes to reduee the risk of problems while drilling and
probes. completing wells. This is important when a facility is brought
● Non-intrusive sand deteetion probes have been proven to online to meet contractual requirements. The surface easing is
be extremely sensitive to small quantities of sand in high 20”, the intermediate easing is 13-3/8” and the production
velocity gas wells (Ref. 4). There was strong preference casing is a 10-3/4” x9-5/8” tapered string.
for non-intrusive sand detection equipment by gas 7-5/8” tubing with slim-line premium couplings with an
operations personnel. A non-intrusive sand detection OD of 8.305” is the largest size found to fit inside 9-5/8”
system was installed for all Goodwyn wells. casing with a drift ID of 8.563”, giving a clearance of 0.258.
● Well rate will be limited to 90% of maximum if sand Slim-line couplings de-rate tubing ratings, particularly
deteetion equipment became unavailable and if the well compressive ratings under combined load. Goodwyn
had not shown any prior sand production. bottomhole and flowing tubing head temperatures are typically
● Metal thickness at su@ace piping bends, subject to the 108”C and 80”C respectively. Thermal expansion from this
highest erosion eff~ would be monitored with ultrasonic temperature profile places tremendous compressive forces on

285
 6 D. TENG, G. NETTLESHIP, S. HICKING, K. HINDMARSH SPE 50081

large size tubing. The tubing hanger is a ‘boll weevil’ type a low positive pressure on the ‘A’ annulus. The completion
hanger so tension can not be pulled on the tubing string. The jewellery should also be compatible with the 13Cr tnbnlars or
tubing immediately above the packer spends the majority of its be more noblq otherwise corrosion ean occur.
life in compression, while some of the tubing/connections in
the string are cycled from compression when the well is under Packer Fluid. High pH potassium chloride brine was used for
production to tension when the well is shut in. To ensure Goodwyn. The high pH inhibits the growth of sulphate
integrity under these conditions the coupling was selected for reducing bacteria and also protects the L-80 casing from
its compressive strength performance and was qualified to API oxygen corrosion. The high pH is obtained by adding caustic
5CT. Triaxial stress analyses were performed to verify that soda @aOH) with biocide being the only other additive.
the tubing and packer operated within the integrity envelope Corrosion inhibitor is not required due to the high pH.
under all well operating conditions (F@ure 9). It should be Minimizing the number of chemical additives reduces the
noted that using 7-5/8” tubing inside 9-5/8” casing is not a new chances of chemical incompatibilities in the packer fluid.
concept having been used in wells more than ten years ago The packer fluid density is kept low to reduee the potential
(Ref. 8), for tubing collapse at future low reservoir pressures.
Increasing the packer fluid density for it to act as a well kill
Material Selection. To maintain integrity and reliability over fluid in case of leakage does not sewe any purpose because the
its lifetime, it is important to select materials to avoid volume is too small and the fluid would drain into the
corrosion, particularly as erosional velocity limits no longer formation. During a workover the packer fluid density
apply in a corrosive environment. The following corrosion probably would not match the required kill fluid density and it
mechanisms were considered for Goodwyn wells: would have to be circulated out.
C02 corrosion or sweet corrosion is caused by COZ
dissolved in produced water, reducing the pH. Goodwyn gas Sub-Surface Safety Valves. Flapper type tnbing retrievable
weIls produce condensed water from start of production. The subsmface safety valves (TRSV) were selected over wireline
gas contains 3% molar weight C02 that gives a ptial pressure retrievable valves (WRSV) for their reliability and because
of 135psi. This necessitates the use of 13’XO Chrome tubulars, they do not require maintenance. A non-quaking type of
which are resistant to COQ corrosion. Carbon steel tubnkirs TRSV was selected for its better reliability (as the valve is less
would have unacceptable corrosion mtes without corrosion complicated) and there is less risk of damage during operation.
inhibitor protection. Injecting corrosion inhibitor, however, Goodwyn TRSVS are set at 550 Q below the potential
introduces a flow-path from the annulus into the well-bore, cratering depths caused by internal blow-outs.
which is against the adopted completion philosophy. On conventional wells the production casing is comprised
Sul@dk Stress Corroswn Cracking (SSCC) or sour of 10-3/4” to below the TRSV setting depth and then 9-5/8 to
corrosion occurs under the combined action of tensile stress total depth. The 10-3/4” casing is required to accommodate
and corrosion in the presenm of water and H2S. NACE MR the 7“ TRSV, the hugest standard size available and which has
0175 recommends that sour service materials be used if the a proven record. When this is combined with 7-5/8” tubing a
partial pressure exceeds 0.05 psi. This is equivalent to 10 ppm slight ‘bottleneck’ completion results, which may increase the
H2S at 4500 psi, an order of magnitude above possible HJ3 risk of problems during wireline work.
levels in the Goodwyn well stream. The 7“ TRSV is backed up with a separate insert landing
h the annti the presence of sulphate reducing bacteria nipple complete with an independent control line. This allows
ean genemte serious levels of hydrogen snlphide. This bad a WRSV to be installed should the TRSV fti.
been deteeted in North Rankin well anrndi, so annrdus fluids in
Goodwyn wells are dosed with biocide. In additioq all well Perforating Guns and Packer. The 4-1/2” perforating guns
annulli are sampled yearly for sulphate reducing bacteria and were run in hole attached to the 7“ completion tailpipe by an
hydrogen sulphide. auto-release drop off system. The packer has a large bore to
Stress Corrosion Cracking (SCC) of steel can be caused accommodate running 4-1/2” TCP guns on electric line as
by the combined action of corrosion and tensile stress in the back up should the TCP gun mis-fire, or at a later date for re-
presence of water and Chloride ions. Goodwyn formation completion. The packer is a conventional hydraulic set
water has a chloride content of 12,000 ppm however the 13Cr permanent production packer with nitrile element tubing is
tnbukws seleeted for the Goodwyn wells are not prone to directly screwed into the packer eliminating an elastomeric
chloride stress cracking. seal. Pedlorating underbalance is obtained by circulating diesel
Gabanic corrosz%n occurs when metals of dilYerent into the tubing prior to setting the packer. The TCP guns are
potential are in electrical contact in a conducting fluid. For fired after flowline inslallatio~ off the drilling rig critical path.
Goodwyn wells, the 13Cr tubing is the more noble metal and
acts as the cathode, while the L-80 production casing acts as Chokes. Vibration of orifice chokes were the single most
the anode and will be the first to corrode. The static packer failure prone component on early North Rankin wells. This is
fluid has limited corrosion capacity, but it is still important to understandable given the high velocity through the choke and
ensure that the ‘A’ annulus is prevented from having o~gen the energy it has to dissipate. After modifieations including
ingress to minimise corrosion. This is achieved by maintaining hard-facing critical areas and machining a slight convex facing

286
SPE 50081 HIGH RATE GAS WELL DESIGN ISSUES AND SOLUTIONS - GOODWYN GAS CONDENSATE, NWS, AUSTRALIA 7

on the orifice plates, choke reliability on North Rankin wells be seen in Figure 10.
improved to only requiring annual impactions. The same The sequence of events for completing the horizontal well
orifice chokes installed on Goodwyn wells failed fkquently, is as follows:
sometimes within days of replacement. The increased energy
associated with the higher liquid content of the Goodwyn well . Drill 8-1/2” horizontal hole from 9-5/8” shoe to TD.
stream is the most likely cause of failure. At Goodwyn choke . @ 7“ cemented liner.
auditions, condensate represents 7.5°A of the total volume ● Run TCP guns on drillpipe, perforate overbalanced, kill
and 30°Aof the total mass, whilst for North Rankin the ratio is well and pull spent guns out of hole.
1.4% of the total volume and 8.5% of the total mass. The high . Install 7“ tie-back liner from top of 7“ liner to a depth
viiration generated by oritice type chokes was also identified accessible by wireline.
by reliability engineers to increase the risk of fatigue induced . Set monobore plug with slick-line in the top of the 7“ tie-
failure of small pipework attached to flowlines. back liner.
In an eflort to reduce noisehibration and improve . Run 7-5/8” completio% circulate diesel to underbalance,
reliability, plug and cage type chokes were trialed on North land completion in compression on top of the tie-back
Rankin and Goodwyn. Plug and cage type chokes dissipate liner.
energy by cancellation of flow within the choke body . Install wellhea~ retrieve monobore plug, unload well and
compared to orifice type chokes which discharge two jets bring online.
downstream into the flowline. Transmitted noise and vibration
was reduced by an order of magnitude. All ofice type chokes Details of the horizontal completion are discussed below.
on both North Rankin and Goodwyn have now been replaced
by plug and cage type chokes. Liner Design Issues. To ensure good zonal isolation the 7“
The pressure drop across the fidly open choke can be a liner is rotated during cementing until the liner connection
restriction to production. The highest rate Goodwyn well has a torque rating is reached. Each liner joint is centralised by hvo
total pressure drop Ilom the wellhead to the sepamtor of solid body spirolisers. The liner is hung off with a hydmulic
366psi. This is comprised of 54psi in the 5.5” ID flowline up set liner hanger. The liner material is 13Cr to give a 25-year
to the choke, 196psi across the fully open choke ond 116psi life and to ensure that monobore isolation plugs or straddles
through the manifold and production coolers to the separators. will set and seal.
Increasing the choke Cv from 240 to 488 will reduce the A weight set liner hanger packoff is run above the liner
pressure drop from 196psi to 40psi and increase production by hanger providing formation isolation during post cementiug
aIIOther 13 MMscfld well cleanout. The liner hanger packoff utilises zero extrusion
A disadvantage of some plug and cage type chokes is their seal technoloW to allow high circulation rates during
low tolerance to debris, especially perforation debris. Choke cementing and gives a more reliable seal when set. The liner
designs with a clear path for debris through the plug and cage hanger packoff has a 20 foot polished bore receptacle (PBR),
can alleviate this problem. which the tie-back liner stings into.
Horizontal Well Completion Design Perforating Issues. Underbakmced perforating was the
The success of the initial Goodwyn 7-5/8” conventional preferred technique, but it is difficult to apply in long reach
production wells resulted in reducing the number of wells to horizontal wells. Coiled tubing was considered but it would
satis& platform capacity and developed the reserves close to have required a minimum of a 2“ diameter reel, 7500m in
the platform. Further reserves could be developed by drilling length and then only 200m sections of the liner could be
extended reach wells with targets up to 8 km from the Wtiorated. Several coiled tubing runs would be required to
platform. At these distances, the wells are horizontal by peflorate all sections with all subsequent runs after the first
necessity and can be drilled across strike to access multiple being fired in a balanced condition. Coiled tubing also
sand units. The initial basis of design for the horizontal wells imposes significant platform constmints, for example crane
was and deck load ratings and limited deck space. It would also
increase completion time and would be costly due to the size
o Allow all zones to be cleaned up. and length of coil required. A second option is to perllorate
● Not restrict productivity. underbalance with TCPS and a packer run on drill pipe. This
● Allow PLT access to the toe of the well. would be cleaned up through tempormy facilities on the
● Allow zonal isolation for injection gas breakthrough or platiorm. This option was discounted due to operational
water shut off. complexity and the need to kill the well after clean-up, which
reduces the benefit of under-balanced perforating. A third
These requirements limited completion options. Various option is to leave the TCP guns in place and perforate tier
configurations were reviewed before the final design of a 7“ completio~ but this would restrict flow and prevent access to
cemented and perforated monobore liner, with a 7“ tie-back the petiorated intervals.
liner and 7-5/S” tubing back to surface was chosen. This can The option that was chosen was to perforate with 4-1/2”

287
8 D. TENG, G. NETTLESHIP, S. HICKING, K. HINDMARSH SPE 50081

TCP guns 200psi overbalance, with the well immediately . The completion is Wled using proven equipment.
killed by a kill pill spotted above the TCP guns. The high . The monobore plug is another well control barrier.
permeability of the Goodwyn formation combined with a ● The upper completion can easily be retrieved.
horizontal well means that the productivity of the well is less
susceptible to formation damage and sub-optimal perforating. Disadvantages of the tie-back liner design are:
The kill fluid has to be formulated for the following
properties: . Two trips are necessary to -install the completion
increasing completion time.
. Prevent si@ficant losses over a wide range of . The seal assembly is potentially exposed to movement
permeability and ‘lock up’ the well quickly. and does not conform to the conventional well completion
● Be non-damaging. philosophy of metal to metal seals.
. Allow minimum filter cake liftoff pressures. ● The sueeess of the completion operation is dependent
. Be nnaffeeted by aetnal firing of the perforating guns. upon recove~ of the 5.72” diameter bridge plug through
the completion string.
The kill fluid chosen was a calcium carbonate particulate . Retrieving the tie-back liner will not be simple as it is
system in a viseosified carrier fluid. A top down firing directly conneeted to the permanently set liner hanger
sequence is used with a time delay of several minutes behveen packoff.
each detonation. As each gun section pefiorates the 7“ liner the
kill fluid moves down and kills the newly perforated zone. To address the potential problem of seal movement during
There have been no well control problems with this system. the life of the well, compressive loading of approximately
After the well is kill@ the spent guns are pulled out of hole 25,0001bs is slacked off when the tubing is landed. To
for the completion to be run. overeome the problem of plug retrieval a pump open deviee is
installed below the bridge plug allowing the well to be placed
Tie-Back Liner and Completion. The tie-back liner has two online for clean-up/production if the plug ean not be accessed
fimctiony it is the bottom part of the completion string sealing with wirdine or retrieved. To date, these bridge plugs have
at the liner hanger and it gets the completion string back to a been reeovered on all oeeasions but aeeess problems due to
depth for wireliue access. debris has necessitated prior well clean-up through the plug on
A problem in installing a monobore completion is that the two horizontal wells.
7“ liner hanger is typically at an angle of 80”, beyond wireline Four horizontal wells have been successfidly drilled and
reach. To overeome this problem the tie-back liner is run with completed in the Goodwyn reservoir. The longest reach well is
a standard weight set liner hanger pack-off at the bottom that Goodwyn 13 which reached a TD of 8420 mah (2870 mTVD)
seals at the liner hanger. The liner hanger packoff is with a horizontal section of 2000 metres.
effectively being used as a completion packer that is set
without wireliue. The tie-back liner is pressure tested down the Horizontal Well Intervention Issues. The Goodwyn
annulus after the packoff is set. horizontal wells intersect multiple sands. The monobore
The tie-back liner is long enough to get the top of the tie- eoneept allowed simpler and cheaper through tubing
back liner to a depth accessible by wireline, @ically 65°. A intervention for re-completion. Wells were designed based on
slickline set monobom plug is then installed in the top of the coiled tubing intervention which was eonsidemd to be the
tie-back liner to aet as a well control barrier. This has three most appropriate technology at the time. However well tractor
purposes: technology is now available and is being evaluated for
application in Goodwyn horizontal wells. It will be the
● The upper completion is run with a barrier in place. preferred option beeause it is a simpler operation and should
● Diesel can be circulated prior to stinging the upper be more cost effective.
completion into the top of the tie-back liner. Concentic workovers using a workslring run through
c The upper completion can be internally pressure tested. tubing may be the only technique that can access the latest
extended reach horizontal wells, beeause the wells have gone
The top of the tie-back liner has a 20 foot PBR that the beyond the reach limit of coiled tubing and well tmctors.
upper completion stings into. The upper completion has non- Concentric workovers maybe performed with the well killed
elastomeric seal stinger at the bottom with a protective shroud or under pressure with a hydraulic workover unit.
assembly to eliminate seal damage whilst running in hole.
Where To From Here
Using the tie-back liner has the following advantages: The Goodwyn formation is highly produetkq this ean be seen
in Figure 11 which is a plot of formation permeability
. Coiled tubing is eliminated from the completion process. thickness or kh against production capacity for dWerent tubing
. There are no circulation devices in the completion string, sizes in a ~ical Goodwyn well. Formations above 20,000
or extra nipple profiles in the monobore completion. mdm are tubing limits and as Goodwyn formations have

2&3
SPE 50081 HIGH RATE GAS WELL DESIGN ISSUES AND SOLUTIONS - GOODWYN GAS CONDENSATE, NWS, AUSTRALIA 9

permeability thiclmksses up to 170,000 md.m with an average References:

of 60,000 mc@ the extm capacity can only be tapped by
larger sized tubing such as 9-5/8” and beyond. 9-5/8” tubing is 1. Scott, M., Evaus, M.,: ‘Use of Pressure Monitoring to Improve
not universally applicable for every part of the Goodwyn Reservoir Management in an Offshore Recycled Reservoir -
reservoir as increased flowrate may cause water coning Goodwyn Gas Condensate Fielt NWS, Auskalia’ paper SPE
problems. However where it can be used, it is a cost effective 50077 to be presented at the 1998 SPE Asia Pacific Oil& Gas
method of gaining extra tubing limited capacity, providing Conferenceand Exhibition,Per@ Australi~ 12-14Oct. 1998.
technkal risk can be managed. 2. SalaIU M.M.: ‘An alternative to API 14E Erosional Velocity
The change from 7“ to 7-5/8” tubing for the first phase of Limits for Sand Laden Fluids’paper OTC 8898 presented at the
Goodwyn wells created many concerns, especially when weli OffshoreTech. Conference,HoustorLTexas,4-7 May 1998.
productivity, sand production and erosion were unknown 3. API: ‘API RP 14E Recommended Practice for Design and
Installation of Offshore Production Platform Piping Systems:
quantities. In hindsight it may’seem a trivial step, but it is now Fifth Editiou p.23, American Petroleum Institute, Washington
a stepping stone to more productive wells such as 9-5AY D.c. 1991
completions. They are being targeted for the Perseus 4. Castle, M.J., and Teng, D.T.,: Extending Gas Well Velocity
development being drilled from the North Rankiu platiorm Limits Problems and Solutions’ paper SPE 22958 presented at
and eventually will be used on Goodwyn for developing the 1991 SPE Asia-Pacitic Conference, PertlL Western
southern sections of the reservoir. Australia, 4-7 November, 1991.
The envisaged well design is shown in Figure 12. The 5. Salama, M.M., and Venkatesk E.S.,: ‘Evaluation of API RP 14E
design utilises existing 7“ equipment (xmas tree, TRSV, Erosional Velocity Limitations for Offshore Gas Wells’ paper
OTC 4485, Proc of the 15ti Offshore Technology Conference,
packers etc). The main change to the drilling of this well is
Housto~ 1983
running the 9-5/8” casing as a liner instead of all the way to 6. SIEP, ‘SrmdManagement Guide’ EP97-5250, Shell International
surface. The 13-3/8” casing beeomes the production casing – Exploration and Production B.V., October 1997.
requirkg it to have premium threads. This allows space to run 7. Turner, RG. et al., ‘Analysis and Prediction of Minimum Flow
9-5/8” tubing up to the sub-surface safety valve, where it Rate for the Continuous Removal of Liquids from Gas Wellsj’
narrows back to a standard 7“ well design. These hybrid 9- JPT (Nov. 1969) 1475-82: Trans., AWIE, 246.
5/8” completions will be capable of flowrates up to 275 8. Barril, R.J., and Gay,L.G., ‘Controlling Sand Production in High
Mh4sWd from the Perseus field. The increased rate will result Rate Gas Wells’ World Oil September 1983.
in higher velocities through the 7“ tubing giving flow
velocities of 38 rids. Procedmes will be developed to monitor
for erosionkorrosion at critical points in the weII. The same
philosophy of zero continuous sand production will be
adopted.
The hybrid design is an intermediate stage to a full 9-5/8”
completion capable of over 300MMscf7dj see Figure 13. This
will require development and testing of new equipment such
as full-bore TRSVS, WRSVS and xmas trees which will incur
long lead times and increase reliability risk. The fidl 9-5/8”
completion cm then be used to evolve even larger completions
to exploit the exeellent productivity of gas fields in the North
West Shelf of Australia and elsewhere.

Acknowledgments
We thank the North West Shelf Development Participants
and the management of Woodside for permission to publish
this paper. The Participants are Woodsid~ Energy Limited
Lt& BHP Petroleum, BP Developments Australia Ltt
Chevron Asiatic Ltd, Japan Australia LNG (MIMI) Pty. Ltd
and Shell Development (Australia) Pty. Ltd.

289
rlg.z: uooawyn Keservou

250.

A w
1? 200-
% A Iii
A ~g
~ A A
7j150 . A
.
A ;$
“3
g, loo ,$ %
~ - A
/3 ~z
E
z 50. L s
3

0,
0 2000 4000 6000 8000 10000
Well distance (m)

A Well Capacity _Planned ‘wA] co st + Unit Capacity co st

Fig. 3: Well length vs productivity

290
 4500 .—... 4500 .-— . —--
T ~“”-’”or--”l 0
0 I
I I I t
-.. .
1
-.+ 0
+
0
4000- -.. 4000- “-----
... -.. .
+ 0 5-li2”tbg
‘. o 5.4/2bg ...
+++1 . 7“ tbg
‘., +0 ‘...: . +
+ rt.kg 0
.,. o 7-51s”tbg
m n 7-N8”@ ‘. -0
+
+ 9-5/0”Ibg
3500- 3500-
‘. —Aw 1 0 ‘., +
0 m o —Ideal
+
n —Aw2 + . . . . . .. Apprl
‘\\ ‘T 0
<>
‘? ?\ ‘r . . . . . ..AKr3 +
- 0 “~ +
3000- 0
0
‘.,+
-e
on .0
+++ ““.
6 s. .+
d. i?+
2500. .
150 o 50 200 250
o 50
Fl~fiats (mm’&!/d)
Flowrate (mmarf/~T

Fig. 4: Inflow performance of appraisal wells ‘Fig~5’: Ideal Well potential ‘for varioui”t&b&g sizes

4500- , ~== ....~.+

“---”-
IR
2500
o 50 100 150 0 & 100 200 250 300
FIowra&%Macfld)
Flowrata (MMscf/d)

Fig. 7: Flo;veloci@=n~ ‘API “C” vs flowr~te = -

Fig. 6: Darcy and”non-Darcy fiOWContribution’ ‘- ‘

—- . , —.- .:. r - ,.,

GOodwyn Design
250
DESIGN LIMIT PLGT POR PIPE SECTION 2 I I
W-5
e. “.. —-..,,-, , 2?0

E
F

I/Iv
I
N
; 4.
R
N
A
L

:
E
‘.i /11
s
s
I [
50
k,
w -a . . tubing
-.----.—- I I

I
1/
: 5 leak
Icollapse 1.250
6 bullhead kill -_-.7” .......7“by7-518”. –— 7-518J - S-518”

.. aas
7 ———.
ini ect ion 1
I___ \
5 overpull 07
9 test below I 10000 i00000 ‘ ioooooo
-8. NOTE : LIMIT 1000
0. 250. 500. 750.
-750. -500. -2<0. FormaNon Productivity (millldarcy.metres)
EFFECTIVS TENSIW, kiPs

Fig 9: Tubing Tfiaxial Stress Analysis Fig. 11: Well pote~tial vs Kh for various tubing sties

291
 m
l’=—’’”-’””-- II

f —.-—..
-., x,m-. —
WEu CLWPLETDS DLAORAM
OOOOWWi QWA42(15P23
L-., ‘_’uu
y. ———..
UWPLETCUOL40R4M
0C9U#YN
-—-”

OWA4X09P4H

. 8: Goodwyn-02
. Deviated Well Design Fig. 10: Goodwyn Horizontal Well Design

Variant Big Bore Completion with 7“ Xmaa Tree True Blg Bore Completion 9-5/8” Xmaa
& TRSV Trea & TRSV

b b
F Xmas Tree,
9.5/6” Xmas Tree,
6.375” bore
6.75” bore appmx.

—.:,,,.
7“ 13Cr Tubing ,J t .

9-516” 13Cr Tubing

T WRSV Nipple (5.87LY ID
9-5/6” WRSVNk@e(11 .7Y
approx,)
!3-318” L-80 7“TRSV (5.750” ID appmx.) %%;$?I%%D,
13-318” L-80
Casing 72 lb/ft 6.250” ID appmx.)
Casing 72 Ibtft
:, ,;!
9-5/8” 13Cr Tubing ,~; ,, 9-516” 13Cr Tubing
,.?.,;“i
.“ ,>.

7 13Cr Tubing
k 7“ 13Cr Tubing
@ 3500m
@ 3500m
9-5/8 Permanent
9-5/6” Permanent
Packer 6.0” ID
Packer 6.0” ID
9-518” L-SO I 5.75” Landing Nipple
5.75” LandingNipple
13Cr Liner 47 lb/ 5.625” ID approx, 5.625” ID approx.
& 13Cr Liner MuleshoelStinger Assy Muleshoe/Stinger Assy
- (no seals) & 13Cr Liner
Hanger @ - (no seals)
Hanger @
5300m 5300m
9-5/8” Shce 9-516” Shoe
L4-1/2°TCP Guns(Fired 4-li2- TCP Guns (Fired
& Released) & Released)
?Qw @6m~ Anchor Mounted Anchor Mounted
bigbore2 bigbore2
/ 3LWCmlVD / 3000mTVD
.vsa .Vsd --

Fig. 12: Variant 9 5/8” well design Fig. 13: True 9 5/8” well design

292