Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Well Testing
Often, the trial-and-error method is viewed as an imperfect surrogate for system- Renzo Angeles, SPE, is an Engineering
atic scientific research. Yet, great advances never would have taken place without Specialist with ExxonMobil Upstream
such an approach. Take shale gas, for example. Despite scarce research conducted Research Company. He has 11 years
by only a few companies, shale-gas plays have undergone massive development of petroleum engineering experience, 3
and have progressed techniques (e.g., horizontal drilling and multistage fractur-
ing) that enabled resources unavailable less than a decade ago. According to a US years as a field engineer for Schlumberger.
Department of Energy report, US shale-gas production is expected to increase from Angeles has conducted and received
1.4 Tcf in 2007 to 4.8 Tcf in 2020. That same year, 64% of the total US gas produc- training in Peru, Colombia, Ecuador, the
tion will come from unconventional gas. USA, Norway, and Canada. Currently,
Well testing is no exception. Although transient-linear-fluid-flow and dual- he works with the subsurface-technology
porosity concepts for fractured horizontal wells have been available for many years,
division supporting ExxonMobils global
painstaking trial and error has been the key to accommodating such extremely-
low-permeability ranges in current prefracture- and post-fracture-well-test designs. exploration, development, and produc-
Similarly, many if not all operators have diligently re-evaluated their appraisal tion affiliates. Angeles areas of interest
programs in an effort to collect data needed for critical parameters (e.g., gas-in- include well testing, formation testing,
place and initial pressure) in a cost-effective manner while accounting for varying near-wellbore modeling, petrophysics,
relationships between matrix and natural- and hydraulic-fracture networks. and subsurface engineering. He holds
It is undeniable that trial and error should not be used to replace the rigor and
thoroughness of systematic scientific research. However, the expanded range of MS and PhD degrees in petroleum engi-
environments that we encounter (from unconventional, to multilayered, to high- neering from the University of Texas
pressure/high-temperature reservoirs) requires rapid learning and adaptation, and, at Austin. Angeles serves on the JPT
therefore, experimentation. The papers selected for this issue are clear examples of Editorial Committee.
best-in-class knowledge development in well testing through first principles and an
old familiar approachtrial and error. JPT
Layered formations are the norm about stimulation effectiveness (skin) Multilayer-Testing Procedure
rather than the exception among oil in each zone. Then, a robust step- The production-logging tool (PLT)
and gas reservoirs. Knowing individ- by-step analysis method that uses contains gauges that measure the flow
ual-layer properties is important for all data and information is applied profile and bottomhole pressure simul-
development strategies, especially for to obtain individual-layer proper- taneously. The procedure for testing a
secondary recovery. A set of best prac- ties. The best practices for use of two-zone reservoir includes the follow-
tices was established in the design, this methodology and the associated ing steps.
execution, and analysis of multilayer uncertainty are detailed in the full- 1. Shut in the well at least 2 days
pressure-transient tests in the Tengiz length paper. before PLT runs are made to stabi-
oil field in western Kazakhstan. Use The challenge in characterizing lay- lize crossflow during the initial build-
of multilayer testing and analysis ered reservoirs is the large number of up. Then, shut-in PLT passes are run
techniques provided a lower level of unknown parameters. The approach upward and downward to measure the
uncertainty in layered-reservoir char- described in the paper resolves stabilized shut-in flow profile and
acterization than can be obtained enough of these unknowns in a sys- bottomhole pressure.
with a simpler commingled-pressure- tematic fashion for a more direct solu- 2. Place the tool above the top layers
transient test alone. tion to be obtained, with a resulting perforations, open the well to flow to
lower uncertainty. reach a constant rate (first low-flow
Introduction rate), and collect bottomhole-pressure
Formations in the platform area of Tengiz Field data until infinite-acting radial flow
Tengiz consist of multiple zones The central platform of the Tengiz (IARF) is observed on a pressure-tran-
with different reservoir pressures. carbonate reservoir is dominated by sient log-log graph (usually 4 to 6
Therefore, crossflow often occurs dur- matrix porosity. The surrounding rim hours). At the end of this drawdown,
ing shut-in periods. In some cases, the and flank regions exhibit evidence record the first set of flowing-period
effect from different layer pressures of fracture networks. Three major upward and downward passes to col-
and skin distributions could result in zones contribute to oil production. lect layer-flow-contribution data.
100% difference in individual-layer The upper reservoir units have been 3. Station the tool above the bot-
permeability estimations. To reduce produced for a longer period of time tom layer. Increase the choke size to
uncertainty, the selective-inflow-per- and, therefore, are more depleted. flow at a higher constant rate. Record
formance (SIP) production-logging All Tengiz wells are essentially ver- the drawdown pressure until IARF is
technique is used to measure the tical and are producing single-phase observed on the transient-diagnosis
bottomhole pressures and flow con- oil with flowing bottomhole pressure log-log graph. Then, run the second set
tributions of individual layers dur- above the bubblepoint, which pro- of flowing-period passes.
ing a commingled-pressure-transient vides good conditions for production- 4. Keep the tool at the same posi-
test. Pulsed-neutron-capture (PNC) log profiling. Because most wells pro- tion as for Step 3 where crossflow
logs are used to gain information duce no water, the connate water has occurs. Shut the well in, and collect
low salinity, and irreducible water the final buildup-pressure data. At the
This article, written by Senior Technol- saturation is approximately 10%, ele- end of buildup, record one shut-in
ogy Editor Dennis Denney, contains vated chlorine-concentration levels upward and downward pass with the
highlights of paper SPE 132596, Best remained in the near-wellbore region PLT across the entire testing interval.
Practices in Testing and Analyzing of the formation after acid stimulation, 5. Pull the tool out of hole and rig
Multilayer Reservoirs, by Yan Pan, which increased the chance for a PNC down.
SPE, Chevron Energy Technology Com- log to detect acid effects. The bottomhole pressures measured
pany, and Michael Sullivan and Sour-gas injection for miscible at the end of each stabilized shut-in and
David Belanger, SPE, Tengizchevroil, flooding began in late 2007 in the flow period at different tool positions
prepared for the 2010 SPE Western central-platform area. Understanding (Steps 1 thorugh 4) are used to estimate
Regional Meeting, Anaheim, California, individual-zone formation properties the initial pressures of the layers by
2729 May. The paper has not been was important in the success of the applying the SIP technique. The qual-
peer reviewed. hydrocarbon-miscible flood. ity of the layer-pressure estimations
For a limited time, the full-length paper is available free to SPE members at www.spe.org/jpt.
Tight gas wells behave as if controlled Gas adsorption is not accounted for in hydraulic fractures enhance the conduc-
by transient linear flow. The half- this model because most of the data are tivity of a natural-fracture system, which
slope on a type curve indicates this in the transient-flow regime in which is a network of fractures. The calculation
behavior, which enables determining gas-desorption effects are negligible. disadvantage of Model 2 is that the frac-
certain reservoir parameters. Linear- ture spacing, L2, generally is unknown.
flow behavior has been observed in Occurrence of Linear Flow Whether Model 1 or Model 2 is more
shale-gas wells also, but these wells Fig. 1 shows a log-log graph and a appropriate for analysis purposes may
tend to exhibit a significant skin effect square-root-of-time graph for daily pro- depend on the analysts preference. It
that is less common in tight gas wells. duction from Well 114. Transient linear may be worthwhile to compare results
This skin effect can mask early linear flow is shown as a half-slope on the log- from both models. It may be the case that
behavior but may be accounted for with log graph and as a straight line on the one model may be more appropriate than
a modified equation. square-root-of-time graph. However, the other model in a particular reservoir
the early part of these curves, before or for a particular completion technique.
Introduction 200 days, does not seem to represent As Fig. 3 shows, the solution for
Shale-gas reservoirs are being developed transient linear flow. In earlier work, it this linear, dual-porosity model identi-
with horizontal wells and multistage was thought that the early departures fies flow regions that a horizontal well
fracturing. These wells produce in a from a half-slope might represent bilin- may exhibit throughout its production
transient-linear-flow manner, and, in ear flow in a dual-porosity reservoir. A life. Region 4, transient linear drainage
some cases, this flow regime could last more-plausible explanation and analysis from the matrix blocks to the fractures,
for years and might be the only flow were determined with this study. is the dominant flow regime for most
regime available for analysis. A half-slope shale-gas wells in early years.
on the log-log graph of gas rate vs. time Dual-Porosity Linear-Flow Model
or a straight line on the square-root-of- An ideal shale-gas well would produce Linear-Flow Period, Region 4
time graph characterizes this behavior. from a rectangular dual-porosity reser- Because the hydraulic fractures have
Fig. 1 shows an example of a shale-gas voira system of fractures with matrix very high conductivity, the duration of
well in transient linear flow for more blocks flowing into the fractures, with transient flow through these fractures
than 2 years. the reservoir not extending beyond the (Regions 1 and 2) usually is very short
The approach used in this study was fracture system. Thus, this system is a and may end in less than 1 day. Therefore,
to assume that linear-transient-flow linear dual-porosity system, and solu- well performance can be analyzed effec-
drainage out of matrix blocks controls tions have been presented as Laplace- tively by considering the transient linear
production. However, unlike tight gas domain solutions. drainage from the homogeneous matrix
wells, shale-gas wells tend to exhibit a Two conceptual models are shown in blocks into the fractures (i.e., Region-4
significant skin effect, which masks the Fig. 2. They are equivalent in the sense flow). See the full-length paper for details
early linear behavior and which must be that they both represent dual-porosity and examples of the calculations.
accounted for with a modified equation. linear systems. Model 1 is a linear,
dual-porosity transient-slab model. Field Applications
This article, written by Senior Tech- The dominant fracture system in Model The plotting and calculation method
nology Editor Dennis Denney, con- 1 is hydraulic fractures emanating from was applied to several shale-gas wells. In
tains highlights of paper SPE 130370, equally spaced perforation clusters in the many wells, the flow behavior appeared
Application of Linear-Flow Analysis wellbore. The matrix blocks in Model 1 to match patterns described in this
to Shale-Gas WellsField Cases, by are treated as homogeneous, although paper and the calculation results seemed
Hasan A. Al-Ahmadi, SPE, Anas they may contain natural fractures. The reasonable. However, several wells did
M. Almarzooq, SPE, and R.A. main calculation advantage of Model not fit the pattern very well. The main
Wattenbarger, SPE, Texas A&M Uni- 1 is knowledge of fracture spacing, L1, reasons for these wells not conforming
versity, prepared for the 2010 SPE because it depends on the perforation- were production interruptions, recom-
Unconventional Gas Conference, Pitts- cluster spacing. pletions, liquid loading at lower rates,
burgh, Pennsylvania, 2325 February. Model 2 is a linear, dual-porosity and interference from nearby well-frac-
The paper has not been peer reviewed. transient-cube model. In this model, turing treatments and production.
For a limited time, the full-length paper is available free to SPE members at www.spe.org/jpt.
[m(p)m(pwf)]/qg , psi2/cp(Mscf/D)
10,000 2.0E+6
Slope=1/2 1.8E+6
1.6E+6
1,000
Gas Rate, Mscf/D
1.4E+6
1.2E+6
100 1.0E+6
8.0E+5
6.0E+5
10
4.0E+5
2.0E+5
1 0.0E+0
1 10 100 1,000 10,000 0 5 10 15 20 25 30 35 40
Time, days Time1/2, days1/2
Fig. 1On the left, a log-log graph of field production data shows a half-slope line indicating linear flow
after approximately 200 days. On the right, the corresponding square-root-of-time graph shows a straight line
indicating linear flow after approximately 200 days.
Discussion
Hydraulic- Approximately 400 shale-gas wells
Hydraulic Unknown were analyzed, and most exhibited
Fracture Natural-Fracture
Fractures L2
Spacing L1 Network clear Region-4 behavior. They did not
show early linear and bilinear flow
(Regions 1 and 2) because of the high
Fracture Half-Length
E03
parameter values to obtain agreement
E04 between two different OGIP values.
E05
Region 4 Conclusions
E06 A calculation procedure has been pre-
Region 5
sented to aid analysis of shale-gas-well
E07 performance. The procedure is based
E06
E05
E04
E03
E02
E01
E+00
E+01
E+02
E+03
E+04
E+05
E+06
E+07
E+08
E+09
E+10
For information on becoming a Rhino Industrial Applicator call: in the USA 877-509-4603, in Europe, Africa and
Middle East: +49 6103 936474, in Latin America: 1-858-450-0441 ext. 6021, in Australasia: +61 7 5585 7000
*All Rhino urethanes are 100% solids. Some epoxy and polyaspartic products may contain less than 100% solids. All businesses independently owned and
operated. 2010 Rhino Linings Corporation. All rights reserved. JPT 1010 5865
WELL TESTING
Introduction
Traditionally, well testing is a reliable 0
dynamic method to acquire a variety of 100 150 200 250 300 350 400 450
information from the reservoir, includ-
Temperature, F
ing permeability, formation damage,
bottomhole pressure and temperature,
fluid samples, and production quan- Fig. 1Database of well tests performed in an HP/HT environment.
TCP=tubing-conveyed perforating.
tification. Over time, methods and
equipment have evolved, expanding
well-test capabilities to different sce- experience in similar cases will be well most of these cases, a deep-well simula-
narios and adapting to more-demand- received because it provides relevant tor was used to test equipment above
ing environmental and safety con- information that will help to make the expected conditions.
straints. In recent years, high-pressure/ appropriate decisions. With HT wells
high-temperature (HP/HT) wells have defined as having bottomhole tempera- Design Consideration
become more common. Documented tures in excess of 300F, extreme-HT Elastomers. The following service
wells are defined as wells having bot- conditions must be considered before
This article, written by Senior Tech- tomhole temperatures between 350F selecting the proper elastomers.
nology Editor Dennis Denney, con- and 400F. Fluids to be sealed: Consider the
tains highlights of paper OTC 21060, type of well fluid, including drilling
Challenging Well-Testing Operations Background mud, aggressive brines (such as formi-
in High-Temperature Environments As shown in Fig. 1, many wells around ates or calcium bromides), acids, and
Worldwide Experiences and Best the world with bottomhole tempera- alcohols.
Practices Learned, by Alejandro tures above 300F have been tested. Temperature range: Include mini-
Salguero, SPE, Edgar Almanza, and Tool configuration has changed lit- mum and maximum operating condi-
Josmar Haddad, SPE, Halliburton tle because equipment reliability has tions along with thermal cycling and
Energy Services, prepared for the 2010 improved. The lessons learned are potential exclusions. Some seals are not
Offshore Technology Conference, extremely valuable for planning and appropriate for low temperatures (i.e.,
Houston, 36 May. The paper has not design. Extreme temperatures repre- seabed in offshore wells or cold-fluid
been peer reviewed. sent the main challenge for the equip- injection).
ment, particularly for electronics, elas- Pressure range: Maximum and
Copyright 2010 Offshore Technology tomers, explosives, and some surface minimum operating range must be
Conference. Reproduced by permission. equipment, including flexible hoses. In considered along with the compression