Professional Documents
Culture Documents
Abstract
A new method of objectively quantifying formation tester data quality has recently been developed with the aim of
optimizing the generation of fluid gradients and contacts. This paper demonstrates how this new method of processing
formation test data and integration with the open hole log data can be applied to improve job planning for sampling. By
using data that have been quality controlled, sampling operations can be improved by using the optimal sampling tool
configuration which can result in higher operational efficiencies with reduced rig time for complex logging and sampling
operations. In this case study, formation tester objectives included establishing gradients and fluid contact points,
determination of formation fluids and capturing representative samples over multiple horizons in a single 6 ⅛” wellbore. An
additional factor which increased the complexity of this operation was the length and deviation of the wellbore, raising
concerns about differential sticking due to the lengthy stationary sampling times required. The formation tester evaluation
process was divided into two runs: wireline deployed pressure testing and drill pipe conveyed sampling. This culminated in
successfully completing 22 pumpouts with 4 samples captured over 6 days without any incidents. The straddle packer was
used 17 times in varying permeability formations and is considered to be the most number of inflations in the Middle East
region.
The real-time formation tester data acquired were compared to the quality controlled data used for post processing
analysis. The results were then used to identify the optimum procedures to be followed and future best practices. The
lessons learned from the case study presented are used to demonstrate a quality control process with recently developed real-
time methods for evaluating test quality. This allowed the data to be analyzed quickly and with confidence, enabling
improved pumpout and sampling operation planning.
Introduction
Saudi Aramco regularly drills wells that penetrate multiple reservoirs of varying lithology and quality, containing fluids
that can include gas, light oil, heavy oil and water in the same wellbore. Identification and capture of these fluids are critical
to understanding the reservoir and optimizing hydrocarbon recovery. While the open hole logs and pressure testing surveys
can help reduce the uncertainty concerning predicting the fluid type, the results are not always conclusive. Because of these
factors, low contamination formation samples obtained with a Wireline Formation Tester (WFT) tool are common in the
formation evaluation process.
Formation pressure testing and sampling is usually the most time consuming formation evaluation operation and
therefore the most expensive. Efficiency in this operation is required to reduce costs and risks. Tests have shown that by
selecting the optimal drawdown volume and flow rate, the time required to obtain a stable pressure can be reduced by up to
60%. The same goes for sampling; proper analysis of the pressure data can optimize pumping time, reducing rig time (hence
cost) and reduce risks of sticking.
This paper will demonstrate new methods for automatically processing formation test data and integration with the open-
hole log data to enable improved job planning for sampling. The results can be higher efficiencies with reduced rig time for
very complex logging and sampling operations.
2 SPE-177971-MS
Hydrostatic
Pstop
Buildup
Pressure (psi)
tp Δt Time (sec)
Fig. 1 – Pressure time plot for typical pretest showing one drawdown and buildup. Before the drawdown the pressure gauge measures the
borehole hydrostatic pressure and after the buildup the flowline is again exposed to hydrostatic pressure. Most pretests have two or three
drawdown buildup sequences to verify or help improve on the testing results.
By making the assumption of pseudosteady steady-state hemispherical flow, the drawdown mobility is determined as
follows (Moran et al., 1962).
ks qdd
M sdd = = C pf .................................................................................................................................................. (1)
µ ΔPdd
Steady-state hemispherical flow during the drawdown can be observed when the drawdown pressure Pdd has nearly
stabilized and the pretest piston is moving at a constant rate. Assuming the final pressure of the buildup is formation pressure
(Pstop) and the final pressure of the drawdown (Pdd) is relatively constant, the drawdown differential is determined (i.e., ΔPdd=
Pstop – Pdd) and assumed to be the steady-state hemispherical flow pressure differential in Eq. 1 (see Fig. 1). The drawdown
flow rate qdd can be measured directly or estimated by the pretest volume and drawdown time or production time (qdd =V/tp,
cc/sec) and is assumed to be constant. With these assumptions, Eq. 1 can be used to determine the drawdown mobility (Msdd).
SPE-177971-MS 3
The constant Cpf is the probe flow coefficient and depends on the size of the probe, geometry of the probe (circular,
elongated, multiple openings, etc.) and the borehole size. Formation anisotropy can also affect this estimate of the mobility
but it is normally not considered. With a 0.10-cc/sec to 1.5-cc/sec drawdown flow rate and a 1.0-cm probe employed, it is
possible to test formations ranging from 1,000 to 0.5 md with pressure differentials ranging from 2 psi to 5,000 psi. By
analyzing the drawdown curve and the build-up curve, it is possible to calculate the mobility of the formation.
Based on these testing QC criteria, formation properties and tool capabilities, a simulator was developed to estimate the
testing time, as shown in Fig. 2. This simulator uses a hemispherical model similar to the one developed by Proett et al.,
2000. The example shown in Fig. 2 is for a “typical” single probe tool but the basic tool parameters for any tool can be used.
The desired buildup stability is the primary criterion used to determine the buildup times. In the case shown below the initial
drawdown flow rate and volume are automatically determined by the simulator, however, they can be specified. In a real-
world setting the mobility is unknown and the first pretest is usually a default setting. As shown in Fig 2, the first drawdown
is short but sufficient information is gained to optimize the next test to enhance the maximum radius of investigation. This is
accomplished by adjusting the next drawdown flow rate and time to maximize the volume withdrawn from the formation
while staying within the operational limitations for the well and testing tool. After each drawdown-buildup the optimized
pretest rates, volumes and buildup times are generated for the next pretest. While the simulator can determine the buildup
time required to reach the desired pressure stability, other operational limitations can be imposed such as specifying the
maximum buildup time. When new parameters are entered or the tool type changed, the simulation is generated showing a
pressure time plot, a stability plot (0-1 psi pressure plot) and the log-log spherical and radial derivatives (Kuchuk et al., 2010;
Bourdet et al., 1989).
100
Initial rate (cc/sec), Blank for Automatic
2000 0.4
Initial volume (cc), Blank for Automatic
Max buildup time (sec), Blank for Auto… 1000 0.2 10
Automatic Inputs
1.1 Qini Initial drawdown rate (cc/sec) 0 0
0 200 400 600 800 1000
16.7 Vini Initial drawdown volume (cc) Tim e (sec) 1
0.2 Q 2nd and 3rd drawdown rate (cc/sec) 1
Stability (psi)
Fig. 2 – Pretest simulator is used to determine “Total Test Time” for a test scenario considering the tool’s specification, operating parameters
and test quality requirements.
4 SPE-177971-MS
The most important result from a planning standpoint is the “Total Test Time” to achieve the desired test quality. By
using a simulator that considers test quality, the complete logging run time can be estimated assuming other factors such as,
formation property variations, number of testing stations, time to change depths, wireline run in and out times, depth
correlations and other operational factors. For example, in a complex pressure survey in a good quality reservoir with an
average mobility of 100 md/cp formation, 100 testing stations would take at least 22 hours for just the pretesting time. If the
testing time or costs need to be lowered, then operational parameters can be changed, such as reducing the number of pretest
drawdowns per station, or selecting a more effective tool technology.
governing the formation’s flow resistance, and in Fig. 5, a 100 md/cp mobility was assumed. The geometric factors for the sources
are determined using detailed simulations (Zefzaf et al., 2006) for the steady-state spherical flow case. In observing the relative
slopes in Fig. 5 it is apparent that the oval probe flow rate is closer to a straddle packer than the single probe. The flow rate is the
primary factor used to determine how quickly a low contamination formation sample can be obtained.
The pressure differential is also an important factor to consider when sampling and it is related to the flow rate by the system
performance. While a higher pressure differential can produce higher rates, this can cause phase changes when sampling and
corrupt the sample. When the pumping differential is increased, the pressure at the source is decreased from hydrostatic. If the
source pressure falls below the bubble point, the sample is no longer single phase and not considered a representative sample.
Notice that the maximum flow rates for any one of the cases shown in Fig. 5 depends on both the pump curve as well as the
source type curve. Therefore, depending on the formation conditions, overbalance and source type, the predicted maximum flow
rates can change substantially. The pumping differential rate can be reduced anywhere below the pump curve line along a
source line, but the rate is also reduced. This can be done during sampling operations if a phase change is detected to preserve
the sample quality, but sampling times are increased. These factors point out how critical the selection of the source and
pump specification can be to obtaining the highest quality samples and keeping pumping station times to a minimum.
Case History
In this case study, the logging operation was challenging due to the well bore
size, deviation and formation conditions. The 6 ⅛” open-hole section for this well
was around 4,000-5,000 ft and had an inclination around 26° which increased to
47° in one section. This presented an operational concern about the possibility of
differential sticking in the borehole when running wireline tools, especially where
extended pumpout times were expected with probe sections and a straddle packer.
The well traversed 11 different formations and not all the intervals penetrated
were at the original formation pressure. Certain horizons were already under
production and, consequently, there was a larger range of overbalance pressures
than would normally be expected. This also contributed to the risk of differential
sticking.
There was also a high degree of variability in the interval’s properties. The
formations varied from clean porous (15-30%) carbonates with low
permeabilities (0.1-20 md) and clean, high porosity (25-30%) clastic formations
with high permeability (± 2000 md), to interbedded and silty sands with varying
porosities (10-25%) and moderate permeabilities (1-100 md). This wide variance Fig. 8 – The tool configuration shown in this
in permeabilities also raised concerns over the risk of sticking, primarily due to figure has the oval probe section placed
the extended sampling times that were required. above the dual probe section. These two
In addition to the standard reservoir parameters of shale volume, porosity, sections can share the quartz gauge in the
water saturation and permeability from conventional tools, objectives of the quartz gauge section. The straddle packer is
below the oval pad and has a dedicated
formation tester runs included the determination of formation fluids, establishing quartz gauge. The order of the tool section
gradients, fluid contact points and capturing representative fluid samples. can be changed if needed.
The formation evaluation process was divided into four runs and comprised a
combination of Logging While Drilling (LWD) and wireline conveyed tools. It was considered prudent to capture the ‘basic’
data while drilling, consequently, a gamma ray, resistivity, density and neutron LWD string was run. Analysis of this data
resulted in a ‘Quicklook’ interpretation to determine the primary zones of interest for pressure testing. Once drilling was
completed, a wireline run comprising spectral gamma, caliper, sonic and neutron spectroscopy was run and logged from total
depth to the casing shoe. No operational problems were encountered in running the tools to bottom and logging up to the
shoe. This was encouraging as it indicated, along with the caliper (four arm, X-Y reading), that the well bore was in a
favorable condition for running a WFT tool. Combining the LWD and wireline data resulted in a more comprehensive
petrophysical analysis to be achieved and this was used to determine the final choice for WFT pressure points. Sample points
were initially picked with the caveat of being refined once the pressure data was acquired.
A dual probe WFT was run on wireline and successfully established 76 pressures combining to give 11 gradients in 10
separate formations. Gas, oil and water were determined from these gradients. Once the pressure data were acquired and
interpreted a refinement of the planned sample points was made. This enabled more definitive fluid identification to be
carried out using a combination of pumpout and sampling techniques, utilizing the oval probe and straddle packer in a
configuration similar to that shown in Fig. 8
8 SPE-177971-MS
The sampling run, which was drill pipe conveyed, successfully completed 22 pumpouts with 4 samples captured and 6
mini DSTs performed within 6 days without any incidents. The straddle packer was used 17 times in varying permeability
formations with 15 pumpouts performed. Two of the straddle packer sets were abandoned after initial pressure testing
proved the points to be too low in mobility for sampling. The number of straddle packer testing and pumpout executions was
considered to be the most performed in the Middle East region at the time. This was a significant milestone, especially when
considering the conditions encountered.
User
Inputs
Data
Quality Setting Range Weight
Ratings
&
Comments Norm.
Criterion Units Valid >
1 0-‐1 High
Quality Good Valid Fair Low
Quality %
Std
dev. psia ≤ 0.5 10 0.5 ≤ 0.05 ≤ 0.158 ≤ 0.5 ≤ 1.581 ≥ 5 13.2%
Buildup
Stability Pressure psia/min ≤ 0.1 100 1 ≤ 0.001 ≤ 0.01 ≤ 0.1 ≤ 1 ≥ 10 26.3%
Temp F/min ≤ 0.01 5 0.5 ≤ 0.002 ≤ 0.004 ≤ 0.01 ≤ 0.022 ≥ 0.05 13.2%
Mobilities Msdd md/cp ≥ 1 100 0.8 ≥ 100 ≥ 10 ≥ 1 ≥ 0.1 ≤ 0.01 21.1%
Radius
of
Invest. Rinv ft ≥ 1 10 0.3 ≥ 10 ≥ 3.16 ≥ 1 ≥ 0.32 ≤ 0.1 7.9%
Supercharge Δ Psc
Avg psi ≤ 2 4 0.7 ≤ 0.5 ≤ 1 ≤ 2 ≤ 4 ≥ 8 18.4%
Test
Quality Score 4-‐0 2 4 3 2 1 0 100.0%
Table 6 – Automatic data quality rating settings. The quality settings establish the limit desired for a “Valid Test” and the “Range” is used for
scoring the tests from 0 to 4. The “Weight” factor is used to determine the final score considering all of the criterion scores.
Additional
Comments Units Rating Stability
&
Supercharge Mobility
&
Radius
of
Investigation
Mud
Set Overbal. psia < 20
Tight Msdd md/cp < 0.1
Low
Perm Msdd md/cp < 1
Supercharged ΔPsc
Avg psia > 2
Msdd mc/cp < 1
Low
Perm
&
SC Δ
Psc
Avg psia > 2
SPE-177971-MS 9
Hydrostatic
Pstop
Buildup
Pressure (psi)
Drawdown
Linear Regression
tp Δt Time (sec)
Fig. 9 – Pressure time plot showing a linear LSR line fitted to the last 60 seconds of the buildup data to determine the pressure stability.
in this case study. This involves careful consideration of the formation parameters, the WFT tool capabilities, and test quality
criteria. These three factors can substantially impact the total testing time and cost. These issues were analyzed using a
simulator that considered all of these aspects and how they can improve the job planning and real-time operations.
An earlier published method of objectively quantifying the pressure data was used in this case study with the aim of
optimizing the quality of the pretest data and automatically generating gradients and fluid contacts. This quality controlled
data can then be used to select sample points, estimate pumpout time for each point and optimize the tool configuration for
sampling. The choice between an Oval Probe or Straddle Packer for each point can then be made which can ultimately
reduce the overall sampling operation time.
In this case study, the real-time logging data acquired during operations was used for establishing gradients, fluid contact
points and sample points and were later compared against the quality controlled data. By reviewing the real-time operation
procedures and analysis against the post processed quality controlled data analysis, it is possible to identify improvements
that could be made if used in real-time. The lessons learned and quality control process are documented in the paper along
with recently developed real-time methods for evaluating test quality which enables the data to be analyzed quickly and with
confidence, facilitating improved pressure testing and pumpout sampling operations.
XX,050
XX,100
Fig. 10 – A formation sequence showing three layers over a 120 ft interval consisting primarily of sandstone, limestone, shale and silty sands.
Shown in the first track are the lithology volume fractions, followed by the borehole caliper with the API Gamma curves (Total and Uranium
Free) in the third track. The effective water saturation (Swe) track also shows the test points indicated by red arrows and the oil sampling
th th
points with a solid green circle. The 5 track is the interpreted oil-water volumetrics, dominated by oil in this sequence. The 6 track shows the
log interpreted permeability curve and the pretest drawdown mobilities which show a good match to the curve. The gradients were
automatically determined from the high quality points shown in Table 6 for each of the three flow units in track 7. The final track includes the
pressure residuals showing the variance of the pressure points from the gradient regression lines (Collins et al., 2007).
company were automatically labeled as “Supercharged” in the auto comments. Some of the test points were repeated as many as 5 times at
nearly the same depth so there were actually only 11 depth points to be considered for the gradients.
XX,050
XX,100
XX,150
XX,200
Fig. 11 – The K formation interval is primarily a carbonate reservoir with relatively medium to low permeability. There appears to be some
transition from water to oil. There was a clear oil gradient in the upper section where oil clearly dominates the saturation volume. In addition, a
clean oil sample was obtained. In the lower depths the pressures are higher which is likely due to the water flood increased pressures. While
the depth points are very close together for the bottom gradients they do suggest the presence of oil but their accuracy is too low to be
conclusive.
References
Beik, H., Proett, M., Torres-Verdín, C., van Zuilekom, T., Engelman, B., and Sepehrnoori, K., 2010, “Effects of Highly
Laminated Reservoirs on the Performance of Wireline and While-Drilling Formation-Tester Sampling with Oval,
Focused, and Conventional Probe Types,” Paper SPWLA-2010-93360 presented at the SPWLA 51st Annual Logging
Symposium held Perth, Australia, June 19-23.
Bourdet, D., Ayoub, J., and Pirad, Y., 1989, Use of pressure derivative in well-test interpretation, SPE Formation Evaluation,
Vol. 4, 293-302.
Collins, C., Proett, M., Storm, B., and Ugueto, G., 2007 , “An Integrated Approach to Reservoir Connectivity and Fluid
Contact Estimates by Applying Statistical Analysis Methods to Pressure Gradients,” Paper SPWLA 2007-HH presented
at the SPWLA 48th Annual Logging Symposium, Austin, Texas, USA, 3-6 June.
Jones, C., Alta, W., Singh, J., Engelman, B., Proett, M., and Pedigo, B., 2007, “Collecting Single-Phase Retrograde Gas
Samples at Near-Dewpoint Reservoir Pressure in Carbonates Using a Pump-Out Formation Tester with an Oval Pad,”
Paper SPE110831, presented at the 2007 SPE Annual Technical Conference and Exhibition held in Anaheim, California,
U.S.A., 11–14 November.
Kuchuk, F., Onur, M., and Hollaender, F., 2010 Pressure Transient Formation and Well Testing, Elsevier, Developments in
Petroleum Science Vol 57, page 35.
Moran, J.H., and Finklea, E.E., 1962, Theoretical analysis of pressure phenomena associated with the wireline formation
tester, paper 177-PA, Journal of Petroleum Technology, Vol. 14, 899-908.
Proett, M. A., Chin, W. C., and Mandal, B., 2000, “Advanced Dual Probe Formation Tester with Transient, Harmonic, and
Pulsed Time Delay Testing Methods Determines Permeability, Skin, and Anisotropy,” Paper SPE 64650, presented at
the SPE International Oil and Gas Conference and Exhibition held in Beijing, China, November 7-10.
SPE-177971-MS 13
Proett, M. A., Musharfi N., Meridji, Y., Gill, H., Eyuboglu, S., 2014, “Objectively Quantifying Wireline and LWD Pressure
Test Quality,” paper SPWLA 2014-QQQ, presented at the SPWLA 55th Annual Logging Symposium held in Abu
Dhabi, United Arab Emirates, May 18-22.
El Zefzaf, T., El Fattah, M. A., Proett, M. A., Engelman, B., Bassiouny, A., 2006, “Formation Testing and Sampling Using
an Oval Pad in Al Hamd Field, Egypt,” Paper SPE 102366, presented at the SPE Annual Technical Conference and
Exhibition held in San Antonio, Texas, U.S.A., September 24–27.
Acknowledgement
The authors would like to thank the management of Saudi Aramco and Halliburton for their support and permission to
publish this article. The authors also acknowledge the contribution of the Saudi Aramco Reservoir Description Division and
Halliburton operations team for their efforts in providing the field examples.
Nomenclature
Cpf ......................... probe flow coefficient (md⋅psi⋅sec/cc)
ct .......................... formation total compressibility (1/psi)
ks ...........................formation spherical permeability(md)
Msdd ....................... spherical drawdown mobility (md/cp)
Pdd ........................................... drawdown pressure (psia)
Pmh .................. wellbore mud hydrostatic pressure (psia)
Pstop ..................................... final buildup pressure (psia)
ΔPdd ....................... drawdown differential pressure (psi)
ΔPpo .......................... pumpout differential pressure (psi)
qdd ............................................... drawdown rate (cc/sec)
qpo .................................................. pumpout rate (cc/sec)
Range .................. geometric scaling range for the Score
Score ............ value of the rating for a test criterion (4-0)
Δt ......................................................... buildup time (sec)
tp .............................. drawdown or production time (sec)
Vpt ...................................................... pretest volume (cc)
µ .................................................................. viscosity (cp)
φ ..............formation porosity (pore volume/total volume)