Anda di halaman 1dari 6

SPE 86951

Innovative Reservoir Fluid Sampling Systems – Hamaca Project, Venezuela.

David R. Reddie, SPE, ChevronTexaco - Petrolera Ameriven S.A. Venezuela.


Carlon R. Robertson, SPE, ConocoPhillips - Petrolera Ameriven S.A. Venezuela.

Copyright 2004, Society of Petroleum Engineers Inc.


between the zone being sampled and the PCP intake. Using
This paper was prepared for presentation at the SPE International Thermal Operations and
Heavy Oil Symposium and Western Regional Meeting held in Bakersfield, California, U.S.A.,
this system, single-phase fluid was captured at surface. This
16–18 March 2004. fluid, which had never experienced a significant pressure drop,
This paper was selected for presentation by an SPE Program Committee following review of was delivered to the lab, where PVT analysis confirmed that
information contained in a proposal submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
the formation fluid was at near saturated conditions, at
correction by the author(s). The material, as presented, does not necessarily reflect any reservoir pressure and temperature.
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of In addition to about 4 liters, per zone, of single-phase
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is sample for PVT analysis, approximately 45 liters were
prohibited. Permission to reproduce in print is restricted to a proposal of not more than 300
words; illustrations may not be copied. The proposal must contain conspicuous
obtained from one zone for use in special core analysis
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O. studies, which would have been too expensive to obtain by
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
conventional methods. The costs incurred in obtaining the
large volumes of sample, by the single-phase pump to surface
Abstract technique, were less than prior expenditures on more
The Hamaca Project is a large resource base of Extra conventional downhole sampling operations, where small
Heavy Crude Oil (7 to 9 API), located in the Faja oil belt, to volumes of varying quality sample were obtained.
the north of the Orinoco River, in Venezuela. Petrolera
Ameriven (PA), an operating agent owned by Petroleos de Introduction
Venezuela S.A., ConocoPhillips and ChevronTexaco, is Information regarding the development of Hamaca field is
developing the Hamaca extra-heavy crude oil project. Field available in numerous other papers, e.g. SPE 78990, Ref.1.
start-up occurred in September 2001, with production from a During the appraisal and development drilling phase of
few horizontal wells. Development drilling and field appraisal Hamaca field development, prior to production start-up, a
is ongoing, prior to start-up of the Upgrader and commercial requirement existed for high quality samples of reservoir fluid
production at about 190 Mbopd. to be obtained. Prior surface and downhole sampling
Prior to start up of the field, a requirement existed for high operations failed to provide a conclusive indication of initial
quality samples of reservoir fluid to be obtained, for use in reservoir conditions. Uncertainty remained in the level of gas
PVT and special core analyses, to aid in the prediction of saturation of the reservoir fluids, which had significant
reservoir performance. The general belief that recombination influence upon the reliability of results obtained from
of oil and gas could not achieve a truly representative reservoir simulation and reserve studies.
reservoir fluid sample led to the requirement to obtain Downhole sampling operations in Hamaca wells are
reservoir fluid samples that had not experienced significant complicated by the requirement for artificial lift systems, the
pressure depletion. high fluid viscosity, sediment production from the generally
In the development of the system employed, a rigorous unconsolidated sands, and the limited technology available for
review was made of available downhole sampling equipment. sampling under these conditions.
Instead of taking pressurized samples downhole, a rod driven Conventional surface sampling, for recombination, was
progressive cavity pump (PCP) was used to pressurize the eliminated as a result of the inability to measure gas, oil and
reservoir fluid and deliver it to surface, where samples were water rates accurately. The Faja extra heavy crude oil (EHCO)
collected. The fluid pressure, at all levels above the pump, forms gas-oil and oil-water emulsions that take considerable
was greater than reservoir pressure, insuring high quality time to breakdown and segregate under gravity. Metering
samples. This also required surface equipment to operate at systems, under ideal operating conditions, seldom achieve a
higher than normal pressures, as compared to normal PCP level of accuracy in rate measurement, better than +/-10% of
operations. the actual. In addition, various authorities in the behavior of
Every aspect of the system, and operating procedure, were Faja reservoir fluids, advised that recombination of oil and gas
examined and refined to minimize the possible pressure drop
2 David R. Reddie, Carlon R. Robertson SPE 86951

samples would fail to produce a fluid, which exhibits the same Pump to Surface
properties, under PVT analysis, as the original reservoir fluid. In the Pump to Surface (PTS) technique, the fluid that
The initial objective was to obtain appropriate volumes of might be trapped downhole, in small volume chambers, at near
reservoir fluid sample, for use in specialized PVT analyses, reservoir pressure, is instead, drawn into the low shearing
appropriate for the gas-oil emulsion forming “foamy” oil, of a PCP. The fluid is further pressurized and pumped to surface.
higher quality than was previously obtained, at an acceptable Wellhead pressure is controlled, so that at all times during
cost. The objective was later revised to include obtaining transfer to surface, the fluid from the pump discharge is
larger volumes of sample, at reservoir conditions, for PVT (4 maintained at above reservoir pressure.
liters per zone) and Special Core Analysis studies (45 liters The small pressure decrease experienced by the reservoir
from one zone). fluid, in flowing from the formation to the wellbore and up to
the pump intake, will allow some gas to be liberated. The
Background behavior of liberated solution gas in foamy oil has now been
The reservoir fluid data available for the Hamaca field in documented in various papers, e.g. SPE 68860 (Ref.2). If the
July 2000, had been obtained from PVT analyses of a limited pressure drop experienced by the crude is low and for a short
number of small volume samples. The samples were obtained, time period, the microbubbles of liberated solution gas will
by conventional downhole sampling operations, from three not coalesce to form larger bubbles, which would segregate
appraisal wells. The fluid analyses indicated a variance in under gravity. With subsequent compression, to a pressure
GOR that was unlikely to exist in the reservoir. Sampling well above reservoir pressure, and the turbulence caused by
operations had been performed over several days, using a passing through the PCP and flowing to surface, the
conventional program of low rate production, to condition the microbubbles of gas may be expected to return to solution.
near wellbore region, with samples being taken in time The initial formation GOR should therefore be preserved.
controlled, small volume chambers. It was found that, because By this system, unlimited volumes of high quality
of the viscosity of the Hamaca crude and the presence of sand, reservoir fluid sample can be obtained. In addition, the
the downhole collection vessels often became plugged and did sediment and water content of the well stream can be checked,
not fill properly. The conclusion reached was that, for future before shipping cylinders are filled. The uncertainty in
sampling operations, an improved system would be required, trapped fluid type, which exists in downhole sampling
which could be relied upon to provide high quality samples. techniques, is therefore eliminated. Critical requirements of
A detailed survey was made of the sampling equipment the system are to:
that could be made available in Venezuela and of sample (a) minimize the pressure drop required for fluids to flow from
taking procedures. The sampling techniques were constrained the formation to the pump intake,
by the requirement to make use of a rod driven progressive (b) maintain the wellhead pressure above reservoir pressure,
cavity pump (PCP). This type of pump has proven ability to while fluids are being pumped to surface, and samples are
efficiently pump high viscosity fluids, with high sediment being taken, and
content. (c) monitor the volume of fluids produced, in relation to the
Downhole sample collection systems tend to be of small tubing volume.
volume. The fluids are trapped by mechanisms that are either For zones that are found to be of low productivity, and
time controlled, annulus pressure controlled or actuated from which cannot be produced with low drawdown, there is the
surface by wireline, etc. Time control requires surface secondary benefit of obtaining atmospheric pressure samples
operations to be co-ordinated and as a result, the pre-sampling at surface.
production period may be extended or cut short, in accordance In addition, for all of the intervals tested, very valuable
with the actuation time. Annulus pressure actuated systems information was obtained regarding the formation condition,
allow sampling to be performed at an appropriate time. the formation fluid type and the ability of the sands to be
Consideration was given to running a carrier, with pressure produced. The PCP intake pressure gauge was also used to
activated sampling chambers, but the high cost, small sample record pressure build-up in specific zones, following well
volume and uncertain sample quality, led to rejection. A closure.
wireline actuated sampling device could not be used, because
of the presence of drive rods in the tubing and because, in Downhole Equipment
most cases, wireline tools cannot be run through the Faja The sampling completion was designed in accordance with
EHCO, due to its high viscosity. Probe type sampling tools, the primary requirement of minimized drawdown. Critical
run on cable, through drilling or completion fluids, were not aspects were identified as the perforations, tailpipe design and
suitable as a result of sealing problems and the pressure drop the proximity of the packer to the test zone. The low
that occurs in the probe tube, with high viscosity oil. drawdown required for sampling was also required to
During the evaluation of alternative fluid collection minimize sand production. The basic components of the PTS
methods for the project, confirmation was obtained that rod sampling test string are indicated in Fig.1 below.
driven PCP and wellhead equipment could be made available
which could allow fluids, at above reservoir pressure, to be
delivered to surface. Consideration of all aspects of the
sampling project, and the quality and volume of samples
obtainable, clearly led to the selection of the “Pump to
Surface” technique, for future PVT sampling in Hamaca.
SPE 86951 Innovative Reservoir Fluid Sampling Systems – Hamaca Project, Venezuela Page 3

during the clean-up, well conditioning and sampling period.


After sampling, with the PCP stopped, the same gauge was
used to monitor pressure build-up, in specific zones of interest.
The PCP selection was also critical to the success of the
operation. The pump had to be capable of producing at low
speed, low flowrate, high pressure differential, with
potentially high sediment content, for many hours. The pump
selected was a single lobe PCP, with a capacity of 12
cu.m/d/100 rpm and 1200m. lift potential. Actual flowrate
achieved was about 25 bpd, at 100 rpm.
The test string had 3-1/2” tubing, with 1” rods. The annular
volume, from the wellhead to the pump discharge, with the
driving rods in place, was determined. This minimum volume
was produced during the well conditioning phase, to ensure
the tubing was fully flushed with single phase fluid.

Surface Equipment
As a result of the abnormally high wellhead pressure
required to be maintained, careful selection of the wellhead
equipment was required. Standard PCP/drive head completion
equipment, installed for normal production, is unlikely to be
able to be used for pump-to-surface sampling, due to the
abnormally high wellhead pressures created.
The PCP drive head and motor were required to be suitable
for many hours of low speed, high torque operation. A
mechanical rotating stuffing box with an internal seal for the
rod string was obtained, which had an appropriate pressure
rating for safe operation at FWHP up to 1500 psi. The
mechanical seal was tested with water and failed at 2500 psi.
Fig.1: Downhole Completion and Surface Systems The highest reservoir pressure encountered in the zones
sampled was about 1150 psia. The rods were driven by a
Perforations were specified which would allow low single 30hp electric motor, connected by belts.
drawdown, while maintaining control over formation Other surface equipment included a choke manifold, a
breakdown and excessive sand/sediment production. Various diesel injection package and flowlines to storage tanks. As a
recommendations were obtained from service companies result of the EHCO viscosity having high sensitivity to
regarding shot density, phasing, charge size, hole diameter and cooling, produced fluids were mixed with diesel at surface,
penetration. The final selection was perforation with wireline downstream of the choke manifold, remote from the single-
run casing guns, 4 shots/ft, 90 degree phasing, 24 gram phase sampling point.
charges, which had expected tunnel diameter of 0.70 inches A conventional test separator was not included as the
and 7.8 in. penetration. These were later changed to smaller flowrate expected was too low for efficient operation. Rate
hole diameter, deeper penetration charges, in an attempt to measurement at surface was performed by monitoring the
reduce sediment production, with unconfirmed success. level of the diluted crude oil storage tank and the pump speed.
The possible use of any kind of sediment screen was As the PCP is a constant displacement pump, rate can be
rejected, due to the potential for drawdown increase from estimated from the pump speed, nominal capacity and the
plugging. The tailpipe consisted of a perforated joint, set volumetric efficiency. At a pump speed of 100 RPM, gauge
across the perforated interval. Some sand production, during tank measurements confirmed the actual displacement rate
the clean-up period, was therefore expected. A downhole sand was about 25 BPD.
control facility was finally installed for one zone, after it was
confirmed that single-phase samples could not be obtained. Developments In The field
Oil had not yet been produced from the zone and confirmation Single-phase sampling at surface operations were
of the fluid type required atmospheric pressure samples to be conducted in Hamaca confirmation wells H-H3-C01 and H-
obtained. C4-CO2, between November 2000 and February 2001. Single-
A packer was required to be run, in preference to a torque phase sampling was successfully achieved in five separate
anchor, to avoid possible gas migration in the annulus and the zones. Formation conditions prevented successful single-phase
associated effect on sample quality. The packer, with the PCP sampling from one zone and tests in two other zones produced
above, was set as close to the top perforation as considered water only, contrary to log evaluation.
safe. No blank pipe was run between the perforated joint and Prior to the production during the sampling, limited
the packer, to avoid possible collection of liberated solution information was available regarding the productivity of each
gas in the annulus under the packer. zone, the level of consolidation of the formations sampled and
A pressure gauge was run to monitor intake pressure in some cases, the formation fluid type, as both wells had no
4 David R. Reddie, Carlon R. Robertson SPE 86951

prior production.
In general, the downhole completion design did not require
any changes. Other perforation designs were used, but no clear
preferred system was established, due to the unpredictable
variance in the quality, consolidation and productivity of the
zones.
Sand production was found to be of higher significance
than expected. Following clean up of perforation tunnels, with
each increase in PCP speed, a new wave of sediment
production would be initiated. Sediment production did
require some modification of the surface control and sample
transfer procedures.
While initial expectations were that the flowrate and Fig.2 Well Performance Data
FWHP could be controlled by fixed and variable surface
chokes, the small rate and small choke size required, were not In Fig.3, detailed downhole pressure data for the full test
compatible with the initially high volumes of sediment are provided, with the pump speed. The data confirm the low
produced. drawdown that was applied to the formation, during the full
The pre-sampling procedure was revised, so that the well sampling program.
was allowed to clean up, at constant/low PCP speed, with no
restriction at surface. When sediment percentage and grain
size had dropped to a stabilized/ minimal level, the FWHP was
gradually raised to well above reservoir pressure. The tubing
above the pump was progressively displaced with reservoir
quality crude.
In zones where sediment production did not stabilize at a
suitably low level, single-phase sampling at surface was not
attempted and atmospheric pressure samples only were
obtained.
The final established sequence of operations, after
perforation and installation of the PCP string, is summarized
below:
1) Record SIBHP. Start up the PCP to minimal speed
(75/100 rpm). Monitor FBHP/drawdown
Fig.3: Downhole Pressure Data
continuously. Record data at 1 minute intervals.
2) Monitor fluid type, by wellhead samples, and
With the very low drawdown applied in testing, the
flowrate, by gauge tank, at 30 minute intervals.
downhole pressure data clearly exhibit a considerable amount
3) When oil reaches surface, check sediment and water
of ‘noise’, which may result from the gauge being supplied
content. Continue sediment checks at 30 minute
from the same power source as the pump.
intervals. Check drawdown, adjust PCP speed if
As may be expected, in any sample taking operation,
required.
equipment has to be cleaned rigorously to avoid
4) Continue production to clean up, monitoring
contamination. With the very viscous EHCO, this may not be
flowrate, drawdown, sediment, and water. Maintain
easily performed. Between tests, operations were performed to
PCP speed constant.
clean all downhole equipment, using hot water only. The use
5) When sediment has minimized, increase wellhead
of any kind of lighter hydrocarbon or solvent, to clean
pressure gradually, by surface restriction. Take
downhole equipment, was avoided.
atmospheric pressure samples before increasing
FWHP.
Single-Phase Sampling at Surface
6) Maintain FWHP above reservoir pressure, continue
Separate to the development of equipment systems and
production until an appropriate volume has been
procedures for delivering single phase fluids to surface,
produced to displace the full sampling string to
techniques were evolved for single-phase sample collection at
single-phase fluid.
surface.
7) While maintaining FWHP above reservoir pressure,
Sediment production also created problems in sample
take single phase samples at surface.
taking. Sample transfer, direct into single-phase shipping
8) Stop the pump, close in the well for pressure build-
cylinders, proved to be impossible, as a result of sealing
up, if required.
problems with high pressure needle valves and plugging of
Typical data recorded during the clean up, conditioning
small diameter bores. In addition, initial samples, with low
and sampling period are summarized in Fig.2 below, including
(2% to 4%) sediment content, proved difficult to work with in
the wellhead pressure and BS&W.
the PVT lab. Filtration in the lab was not easily achieved.
SPE 86951 Innovative Reservoir Fluid Sampling Systems – Hamaca Project, Venezuela Page 5

The solution developed was to take the sample at wellhead


pressure into an intermediate sample chamber, equipped with
ball valves and large bore ports (Fig.4).

Fig.6: Sampling – Stage 3

A large number of 600cc single-phase sample cylinders


were filled using this process. This type of cylinder has two
Fig.4: Sampling – Stage 1 internal pistons, with pressurized nitrogen (5,000 psi) on the
back side. This type of cylinder allows the reservoir fluid
The sample was isolated from the wellstream, warmed and sample to be maintained in single phase at all times, while
slowly transferred, at elevated pressure (5000 psi), through a complying with shipping regulations.
fine filter (100 micron), to a shipping cylinder (Fig.5). At all After delivery of the single phase sample cylinders to the
times, the sample pressure was maintained well above PVT lab, the samples were transferred. Subsequent analysis
reservoir pressure, during the transfer process. With filtration, indicated that the bubble-point pressures of the samples
the sediment content in the shipping cylinder was reduced to obtained were consistently close to reservoir pressure,
about 0.2%, with particle size being very small. If the confirming that the reservoirs sampled were at saturated
sediment filter plugged during the transfer process, it could be conditions.
cleared by briefly reversing the flow, displacing fluid from the Approximately 4 liters of sample were required for the full
shipping cylinder to the intermediate vessel, before resuming suite of PVT analyses for one zone. In addition to the PVT
the transfer process. samples, approximately 45 liters of single phase fluid were
delivered for use in Special Core Analysis work.
The total cost of obtaining the single-phase samples, from
five zones, in two wells, was less than the cost of prior
sampling operations, where a few 600 cc samples of uncertain
quality were obtained.

Atmospheric Pressure Samples


In some zones, the low productivity, and associated high
drawdown and sediment production, prohibited single phase
sampling at surface from being performed. In some zones,
water was the produced fluid. In these cases, the same
sampling/test string was used to obtain formation fluid
samples that were confirmed to be free of drilling and
completion fluid contamination, before production operations
were terminated.

Pump to Surface Bailer


More recent sampling operations in the southern areas of
Fig.5: Sampling – Stage 2
the field, where the formations are shallower, the temperatures
lower and fluid viscosity higher, encountered severe problems
After filling the shipping cylinder, the sediment filter was
flushed using produced well fluid, maintaining pressure in as a result of high levels of sediment production. Without a
sampling lines above reservoir pressure (Fig. 6). screen tailpipe, sediment level was excessively high, leading
to PCP/production string plugging. With a screen tailpipe,
inflow to the pump intake was restricted to almost no flow
conditions. Following a period of re-evaluation, the rod driven
PCP sampling string was changed to a pump to surface bailer.
6 David R. Reddie, Carlon R. Robertson SPE 86951

This consists basically of two opposed check valves in a


telescopic section, at the end of the tubing string, which is
reciprocated by use of the rig. With no tailpipe screen, the tool
successfully lifted produced fluid to surface, with sediment
levels in excess of 80%.
Surface filtration was attempted but could not be
performed successfully. High sediment content fluid samples
were therefore delivered to the lab for sediment extraction.
Various atmospheric pressure sample analyses have been
completed, confirming the higher than expected reservoir fluid
density and viscosity.

Acknowledgements
The authors wish to thank Petrolera Ameriven and its
parent companies, ConocoPhillips, Petroleos de Venezuela,
S.A. and ChevronTexaco, for the opportunity to publish this
paper. We must also thank our former colleague, Darryl
Williams, SPE, and numerous service company personnel, for
their assistance in developing the systems described in this
paper.

References
1. Gipson, et al, “Hamaca Heavy Oil Project – Lessons
Learned and an Evolving Development Strategy”, SPE
paper 78990, prepared for presentation at the SPE
International Thermal Operations and Heavy Oil
Symposium and International Horizontal Well Technology
Conference held in Calgary, Alberta, Canada, 4-7
November 2002.
2. Mastmann, et al, “Predicting Foamy Oil Recovery”, SPE
paper 68860, prepared for presentation at the 2001
International Thermal Operations and Heavy Oil
Symposium, held in Porlamar, Margarita Island,
Venezuela, 12-14 March, 2001.