SPE-198744-MS

Gas Lift Optimization within Field Capacity Limitations

Uche Chukwunonso Ifeanyi and Samuel Esieboma, OERL; Jennifer Uche, Rego Petroleum

Copyright 2019, Society of Petroleum Engineers

This paper was prepared for presentation at the Nigeria Annual International Conference and Exhibition held in Lagos, Nigeria, 5–7 August 2019.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents
of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written
consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may
not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract
Optimizing oil production with facility constraints has become a challenge to most E&P companies even
as they pursue sustainable resources. The innovative gas lift technique overcomes this challenge. The
conventional gas lift well system has long been in use, but the design most times is limited by gas availability
and pressure which limits the depth of gas lift injection for improved production rates. This challenge may
not be evident in matured producing fields with gas compressors installed with available non-associated
gas source wells, but truly such challenges arise in new fields especially owned by indigenous companies
where much uncertainties at an early field life unavoidably allows you to be more stringent in expenditures
towards development of a field gas lift project. A new gas lift concept was developed and studied in Field
A in an offshore field of the Niger delta in the absence of gas compressors. This design has been proven
to be suitable because it was used to bring four closed wells online even when those wells were removed
from the company annual forecast. The original design consists of a minimum of two unloading valves and
an orifice at a deeper depth, but because of the absence of scrubbers and gas compressors in the facility,
pressure depletion in the reservoirs caused four flowing wells to be closed. The new design then sets dummy
at shallow mandrels and uses a modified size of orifice to optimize available pressure and gas required to
open the closed wells and still sustain other gas lifted wells connected to the same gas lift manifold. This
campaign resulted to an additional 7000Bopd which is the primary discussion of this paper.

Introduction
Gas-lift has been studied and reported for many years (1, 2, 3 4, 5, 6). Several designs in optimizing
production from various fields have also been implemented. D field is an oil and gas producing field located
at offshore Niger Delta, Nigeria. Following a series of appraisal wells, field development plan, first oil
production was realised in early 2000
Main production from D –field is from multi stack major reservoirs A1 and A2, with 70% CO2
concentration. There are a total of two producing platforms at D-field comprising of one central processing
platform (CPP) and one satellite platform. All processing equipments are located at the central processing
platform. Full stream production from the satellite platform undergoes separation processes at the CPP, prior
to collection at the floating storage and offloading (FSO) facility.
2 SPE-198744-MS

Located at the CPP are the separation process unit, water injection module, and produced water treatment
system.
The field is operating with a total of 20 completion strings. Some of the total oil producing strings requires
gas lift to flow indicating that the total field production is highly dependent on the effectiveness of gas lift
in a well. The operator is planning gas treatment facility and gas compressor installation in the next phase
of development and so therefore the gas lift implementation at the moment solely depends on the use of
high GOR and high FTHP wells with zero BSW to gas lift the weak wells in the field. Temporary lines are
used to connect the source high GOR well from its tree cap to the gas lift manifold. The gas lift manifold
is a network of all gaslifted wells on each platform. This modus of operation is replicated in the second
platform which has another high GOR well as its source of gas lift. The gas lift pressure for optimization of
number of gas lift valves and orifice open for gas lift is dependent on the WHP of the source well per time.
Therefore, due to the absence of compressors on site needed for gas lift, proper management of the gas
lift source well requires more attention, and this is put under surveillance thus proper monitoring of the gas
source well is opted as a key factor before progressing with optimization of gas in all gas lifted wells which
will ensure sustenance of production from these wells. This paper will discuss the strategy and methodology
in conducting the surface gas lift optimisation.

Problem Statement
Through the course of production in D field, increase in BSW was observed between year 2000 and 2013.
This phenomenon caused some wells to quit natural flow which then led to investigations and studies to
determine avenues to artificially lift these wells in the absence of gas lift compressors.
Operating this field without compressors to compress produced gas initially posed a lot of challenges
which included production decline and shutting in of wells. In addressing the phenomenon, methodology
of gas lift optimisation was developed

Figure 1—Production history ans log correlation panel for OB reservoir.

METHODOLOGY
In the plan to stop the decline in field production per day, the operator had to embark on the easiest way to
improve on the field production. A quick evaluation showed that 45% of the wells both on the two platforms
either require gas lift for restart purpose or requires gas lift permanently to sustain flow, by conducting
SPE-198744-MS 3

gas lift optimisation, 60% of the total production can be secured and enhanced. Through this objective, a
methodology was developed to ensure the success of the activity at site.

CONCEPT
Based on the basic concept of gas lift, the capability of a well to flow naturally, the efficiency of lifting in
a gas lifted well and the efficiency of gas utilization through the identification of net oil rate per unit of gas
lift injected were used to gauge the well's performance. In order to ensure that the correct allocation of gas
to a well and the production of the well are optimized, it is important to determine the performance of the
well using those key factors mentioned above. The implementation of the itemized concept in site should
never compromise the current field production rate so an approach was devised based on field tests… The
approach is itemized below.

Gas lift provision for gas lifted wells

In the field of interest, no compressor is available to compress available gas in the facility neither is there a
scrubber available to strip the gas of any non required component. Therefore, gas lift for wells was mainly
dependent on high GOR wells. The source well GOR is in the field is 2800scf/stb and the choke can be
opened up to any level depending on the gas demand. The flowing tubing head pressure ranges from 1300psi
to 1400psi. The well is piped from the tree cap to the gas lift manifold. Normally, when the gas is needed
from this well, swab valve is opened up for gas outflow into the gas lift manifold and the wing valve
is opened to deliquefy the gas. The liquid stream which comes out from the wing valve is sent into the
production separator. The gas which goes out from the swab valve is diredted into the gas lift manifold.
This gas contains oil droplet, but this never caused any integrity issue knowing fully well that the gas has
to be passed into the the annulus for gas lift. During shut down campaign this well is lined into separator
to test the well and determine the gas rate at every FTHP. This pressure and the associated gas available
are used during gas lift optimization. The well is never opened up to a level where the flowing tubing head
pressure available cannot go to the deepest optimized orifice and so the choke opening is usually small to
keep the FTHP as high as possible. Before the source well was selected as a gas lift gas provider, all wells
in the platform were tested and the well with the highest flowing well head pressure, highest delivery gas,
zero water cut and least oil rate was selected. Moreso, the gas reserves which is informed from field study is
also an ideal information required before selection. Selecting a well with oil rate as low as possible makes
the gas available for gas lifting to be as lean as possible after deliquefying it through the wing valve. The
life of this well was managed by ensuring that a proper gas lift exercise is done at every time. A proper
gas lift design and management ensures that no gas is wasted but adds additional production to the field.
When the gas is not managed properly, a FTHP decline sets in and may lead to early failure of this well
for continuous use as a gas lift gas supplier.

Well selection based on high field producers

Based on the production well test data, the oil producers were ranked. The focus of optimisation of gas lift
supply first focused on the top high producers to secure the production of the field. The technique applies
during well start up after a total field shut down or when there is inadequate gas in field for field gas lift
optimization.
4 SPE-198744-MS

Figure 2—Gas lift manifold tied to gas lift gas supplier well

Well selection based on lift efficiency

From scheduled well tests of all the wells, the formation GLR for all wells are tested. This well test
parameters are used in the model to verify the wells ability to flow naturally. Normally, in this field, wells
that have GOR of 900scf/stb and with low BSW flow naturally. This was also confirmed in the model by
validating the model with well test results. In addition, all the well tests were reviewed to determine the
wells which have high total gas liquid ratio. Those wells became candidate for further gas lift optimization,
target being to determine optimal gas lift injection rate above which incremental oil is not obtainable. These
measures were to reduce wastage of available gas which could be used in other gas lifted wells to improve
field oil delivery rate. Moreover, oversufficient gas lift supply to a well could deprive a well from producing
optimally due back pressure exerted on incoming fluid from the formation by the gs injection pressure at
that depth. This is more evident in high producer wells.

Gas lift valve redesign

Based on the analysis result, in the case whereby a deepening of injection point or change of orifice size is
required to further optimise gas lift injection, a gas lift valve redesign and change out is required.

Review of well data

One of the challenges of operating a field is regarding the well data measurement or integrity issue.
Therefore, to ensure that the well data obtained is valid, review of well data via well modelling is required.
There are many models available in the market, but prosper was used for gas lift optimisation. Because the
wells in this field has no bottom hole gauge, no real time daily pressure dta was available, but slickline
was run in hole maximum two times in a year to acquire flowing gradients, static gradients and bottom
hole pressures.
In addition, each time a slickline goes into any of the gas lifted wells, the top liquid depth is tagged
which is documented and used for effective gas lift design. Because the target in every gas lift design is to
assist in the effective lifting of oil, the liquid depth helps in choosing the shallowest unloading valve depth
in the design to avoid recirculating of gas in depths where there is gas column or just a vaccum column.
SPE-198744-MS 5

The information gathered is used in making a well model and adjustements are made to gas rates when the
rates measured by the meter is not 100% correct. Besides validating the well test data, sensitivity on gas
lift injection rate towards oil or liquid production is also conducted. This is to estimate the optimum gas lift
injection rate which best matches the well at various conditions. The objective of the well data review is to
obtain the value that are representative of the well's condition despite the issue on accuracy. The analysis
will then be confirmed at site and recalibration of well model.

CASE STUDY
For the purpose of this study, two wells out of nine wells which require gas lift to keep them flowing are
used to demonstrate the gas lift philosophy.

D17 Well
This well originally had mandrel depths as indicated in table … with a target production rate of 3175Bopd.
The start up injection pressure inlet of the annulus is 1015psi and becauses of the losses encountered at the
three unloading valves, 865psi pressure is only available for injection at the orifice.
Unloading valve port sizes of 8/64th inch was used each at depths of 889ft, 1279ft and 1553ft while a
reduced port size of 7/64 th inch size was used at the orifice depth of 1750ft. The pressure loss encountered
at the orifice depth cascaded to the gas lift network causing some pressure loss on the manifold which caused
inadequate pressure for effectively lifting other wells attached to the network, insufficient gas required for
optimizing oil from other gas lifted wells.

Number of valves Valve type Measured depth Gas injection rate Pressure lost (psi) Injection pressure Valve port size
(feet) (mmscf/d) at orifice (psi) (64ths inch)

1 Dummy 357

1 valve 889 0.013 8

1 valve 1279 0.013 50 8

1 valve 1553 0.013 50 8

1 Orifice 1750 0.13 50 865 7

Figure 3—Original gas lift design of D17 well.

6 SPE-198744-MS

Review of original gas lift design of D17

For the purpose of optimising oil production in the field using exact volume of gas for gaslifting each well
coonnected to the gas lift network, a review of gas lift desin of this well was made and on the process it
was observed that there is a threshold gas volume required to make same rate of 3175bpd without deffering
production of oil from the field or compromising current field delivery rate. Moreso, it was observed that
loosing unuseful gas results to loss to useful pressure which ends up compromising field delivery rate.
Pertinent to this, the old gas lift design was reviwed by isolating first three unloading valves and leaving
them as dummy and setting orifice only at the last depth of 1750ft. The design gave a delivery rate of
3175bpd while saving about 0.12MMscf/d gas and also reducing pressure losses associated with keeping
a lot of mandrels open for gas lift injection.

Number of valves Valve type Measured depth Gas injection rate Pressure lost (psi) Injection pressure Valve port size
(feet) (mmscf/d) at orifice (psi) (64ths inch)

1 Dummy 357

1 Dummy 889

1 Dummy 1279

1 Dummy 1553

1 Orifice 1750 0.05 50 997 4

Figure 4—Reviewed gas lift design of D17 well

D18 original GL design

This well originally had mandrel depths as indicated in table … with a target production rate of 5000blpd,
but due to the mandrel depth condition, the well was able to make a rate of 3176blpd. The start up injection
pressure inlet of the annulus is 1015psi, but because some pressure were lost in the 2nd and third unloading
valves, the available pressure could not go deep into the last mandrel and therefore the orifice depth was set
at 1700ft which may not be an optimal depth for effective aerating and scooping of oil for a better rate. In
general, a total of 0.2mmscf/d gas injection rate is used to sustain a rate of 3176blpd
SPE-198744-MS 7

Number of valves Valve type Measured depth Gas injection rate Pressure lost(psi) Injectionpressure Valve port size
(feet) (mmscf/d) atorifice (psi) (64ths inch)

1 Dummy 368

1 valve 1100 0.017 50 8

1 valve 1400 0.017 50 8

1 orifice 1700 0.17 915 7

1 Dummy 2250

Figure 5—Original gas lift design of D18 well

Review of original gas lift design of D18

Similarly, D17 original design has two unloading valves at the depths of 1100ft and 1400ft and an orifice
at 1700ft. With the original design about 0.2MMscf/d gas was required to produce this well at the rate of
3175bpd liquid and an associated pressure loss of over 100psi which degenerated to an injection pressure
of 915psi at a mandrel depth of 1700psi. The gas was not able to go deeper into a mandrel depth of 2250psi
because of such unavailable pressure from the manifold. To save gas and pressure for use in other gas lifted
wells, a redesign was conducted for this well and the sole objective of this redesign is to take away any
gas which does not give additional production in the form of oil and allocate it for other wells connected
to the gas lift network. In this case, having the idea where the top liquid is at, the first to the fourth gas lift
mandreal were blinded and an orifice was secured at the last mandrel depth. In this case, there is 1015psi
injection pressure which is able to inject from the orifice depth at 2250ft compared with the original design
which could not go deeper because of pressure losses associated with injecting at the shallower unloading
valves. Because of the higher injection pressure available for this redesign, a little incremenatal oil gain was
observed though not much, but shows an effective approach which can be able to gain a huge incremental
oil in fields with high potential to produce oil.
8 SPE-198744-MS

Number of valves Valve type Measured depth Gas injection rate Pressure lost(psi) Injectionpressure Valve port size
(feet) (mmscf/d) atorifice (psi) (64ths inch)

1 Dummy 368

1 Dummy 1100

1 Dummy 1400

1 Dummy 1700

1 orifice 2250 0.12 1015 7

Figure 6—Reviewed gas lift design of D18 well

Execution at Site
Towards the target of optimizing fluid delivery from this field, the gas lift optimization strategy deployed
in the two wells which are illustrated above gave rise to improved injection pressure available at the gas lift
network which was made possible through conservation of gas coming from the choice reservoir which also
helped to manage the depletion of reservoir pressure.. The most interesting part of the whole exercise is the
fact that the volume of gas read off from the design is not the total volume of gas saved, because if it takes
a well five days to stabilize before cutting off the gas lift gas, therefore the volume of gas lost increases
as the it takes days to start up a well and the pressure losses associated with this keeps increasing per day.
After the volume of gas injected in both wells were regulated, the pressure on the manifold improved from
about 700psi to over 1000psi. This approach created the opportunity to open two other wells which were
closed because of gas and pressure unavailability.

Conclusions
Gas lifting in a new field with neither a compressor nor a gas scrubber was made possible by
a. Identifying a source reservoir with very high GOR and reserves that could last till end of field life.
b. Identifying a well producing from such reservoir with zero or insignificant water cut.
c. Identifying wells which require assistance with gas lift.
d. Determining the minimum gas requirement to meet same target rate for each of the wells.
e. Identifying the least number of unloading valves and the minimum size of valve port size/orifice
required in every gas lifting design to meet a target rate.
SPE-198744-MS 9

Following this approach leads to conservation of gas and pressure required to have more wells to tie to
the gas lift network and then improving on daily delivery rate for the field.

References
1 F. Bertuzzi et al, Truo AI ∼ (1953)$v" 198$ PO 271o
2 K. E. Brown & G. H. Fancher, Jr *∼, SOC. Petro zrs (1963)∼V. 228,p. II–59.
3 K. E. Brown & A. R. Hagedorn, See. Petr. Engrs. (19651S V. 234, p. 1–475.
4 K. E. Brown, T. C. Doerr, J. P. Brill, Soo. Petr. Engrs. (1967), V. *O, p. 1–419.
5 F. H. Poettmann & P. G. Carpenter, Drill. &Prod. Prac., API (1952), 257.
6 N. C. J'. Ros, SOC* Petra Engrs* (1961) Sv. 222s p. I–1037*
7 M. R. Tek, Sot. Petr. Engrs. (1961),V. 222, p. 1–1029.