Professional Documents
Culture Documents
Abstract
Capturing wellhead pressure for conversion to bottomhole pressure (BHP) requires that
the well is in critical flow with sufficient velocity to not exhibit slugging. This
technology is most commonly utilized in Gas and Gas-Condensate wells. It has greatest
value where single well caissons & unmanned platforms are located. In addition, wells
that are deep, HP/HT, sour, remote, etc... benefit from this technology.
The unique challenges with gathering high frequency, high resolution wellhead pressure
data and accurately converting to BHP for use in PTA will be shown. These challenges
can be broken into 4 main areas: 1. Temperature (ambient temperature effects &
wellbore cooling) 2. Liquid re-injection 3. Operational/Testing procedures 4.
Conversion methods.
Introduction
The most significant advantages of surface testing are reduced costs and the elimination
of risk. A typical surface test will cost a fraction of a downhole pressure test. A surface
gauge can be easily installed in a matter of minutes by field personnel. It does not require
a lift boat, lubricator or the additional manpower normally associated with a downhole
test. The reduction of risk is intuitive. By using a surface gauge, there is no wellbore
intervention, negating the possibility of a lost or stuck tool or losing the well itself.
Elimination of wellbore intervention operations also minimizes the safety risks to field
personnel.
Another advantage of surface testing is that it enables the engineer to test wells in which
mechanical factors make running DHGs difficult or impossible such as highly deviated
wells. HP/HT environments and sour wells require expensive gauges that may not even
survive long enough to complete the well test. Physical obstructions in the wellbore may
prevent running downhole gauges at all. In wells with coated tubing, the operator will
likely want to test from the surface to avoid damaging the coating.
Surface testing also allows the operator to manage the test in real time. By acquiring and
analyzing data in real time, the operator can make adjustments to the test plan to ensure
that the test meets their objectives and that it is done so efficiently and economically.
However, there are technical hurdles to successfully testing from the surface which must
be understood and overcome. These challenges can be grouped into three categories,
instrumentation requirements, candidate selection, and wellbore modeling. This paper
will discuss these technical questions and provide field examples to demonstrate the
viability of surface testing.
Instrumentation Requirements
Accuracy
Gauge accuracy simply quantifies the degree of uncertainty in a given pressure
measurement. For example, if a 10,000 psi pressure gauge has an accuracy of 0.015% of
full scale, then there is +/- 1.5 psi uncertainty in the measured pressure. Gauge accuracy
is not as important in pressure-transient analysis as gauge repeatability, gauge resolution
and thermal compensation.
Repeatability
Gauge repeatability is the measure of error observed when a gauge is repeatedly cycled
between a known pressure and a second pressure (typically atmospheric pressure). It is
expressed as +/- psi maximum deviations from the actual pressure observed after multiple
pressure cycles. These deviations are the combined errors from hysteresis, accuracy and
resolution. If a pressure gauge does not have sufficient repeatability, then the measured
pressure response will be distorted from the reservoir response and the resulting analysis
will be in error.
Resolution
Gauge resolution is the smallest change in pressure that a gauge can reliably and
repeatedly detect. High resolution gauges typically have resolutions of 0.01 psi. They can
detect extremely small changes in pressure and consequently in reservoir response. If a
pressure-transient test is conducted with a gauge that does not have sufficient resolution,
it will not be possible to analyze the data with confidence because the gauge cannot
detect the reservoir response. High permeability wells should only be tested with high
resolution gauges. The following plot illustrates the difference between high and low
resolution pressure data.
Resolution Comparison
2250
2245
2240
2235
WHP
2225
2220
2215
2210
0 6 12 18 24 30 36 42 48
Delta Shut-In Time
Thermal Compensation
Any high-resolution pressure transducer is also a good thermometer. This means that at a
true constant pressure, the transducer element will change properties with temperature. If
the magnitude and direction of this thermally induced shift is repeatable, it is possible to
compensate the transducer for temperature change. Thermal compensation of high-
resolution pressure gauges is very difficult and rarely done effectively. The amount of
compensation required is both a function of pressure and the rate of change of
temperature. DHGs are very susceptible to thermally induced pressure shifts, especially
when making gradient stops. Depending on the gauge, a stop may require 30 minutes or
more to thermally equilibrate before the pressure reading stabilizes. In gas wells,
downhole gauges near the perforations for a build-up may be subject to rapid heating
when the gas stops expanding across the perforations. This loss of Joule-Thompson
cooling distorts the early time pressure readings on many PBU tests. Surface pressure
gauges also require thermal compensation. This compensation is necessary because of
changes in ambient temperature from day to night. It is not uncommon to see 50 psi
pressure swings over the course of 24 hours if the thermal compensation on the gauge is
inadequate. If a pressure gauge is not properly thermally compensated, it is impossible to
distinguish between pressure changes from reservoir response and pressure changes due
to the gauge changing temperature. This will render analysis obtained from that data
questionable. The following plot shows the difference between adequate and inadequate
thermal compensation.
Gauge Thermal Compensation
12110
12105
12100
12095
Dual Quartz
Silicon-Sapphire
12090
Pressure
12085
12080
12075
12070
12065
12060
12055
06/12/06 9:00 06/12/06 12:00 06/12/06 15:00 06/12/06 18:00 06/12/06 21:00 06/13/06 0:00 06/13/06 3:00 06/13/06 6:00
Time
Candidate Selection
Not all wells can be successfully tested from the surface. If the following criteria are met,
then the well is a good candidate for surface testing:
Constant Mass Flow Rate
Constant Component Flow Rate
Effective Fluid Continuity from the Surface to the Completion (i.e. no gas/liquid
interface in the wellbore)
There are two subsets of wells that meet these requirements, single phase wells and
naturally unloading multiphase wells.
Wellbore Modeling
The final technical hurdle to testing from the surface is accurately modeling the recorded
wellhead pressure (WHP) to bottomhole conditions (BHP). The basic equation for
modeling BHP from WHP is fairly straightforward:
The Pfriction term includes pressure losses in the wellbore due to frictional effects.
Smaller tubing IDs, higher flow rates and lower pressures all increase frictional effects.
The Pgravity term accounts for the hydrostatic head in the wellbore. Lower temperatures,
higher pressures, and denser fluids increase the hydrostatic head. There are many
different BHP models available in the oilfield. Examples of these include Cullender-
Smith and Hagedorn-Brown. At a minimum, the following inputs are necessary for a
good WHP to BHP model:
Gas, Oil and Water Flow Rates
Gas Analysis and/or PVT report
Liquid Yields and Densities
Tubing Schematic
Mid-Perf TVD
Wellhead and Bottomhole Temperatures
In addition to these inputs, an accurate BHP model must account for Wellbore
Temperature Changes, Multi-Phase Flow Regimes, and Phase Redistribution.
From this equation, gas density can be seen to increase with increasing well bore pressure
and decreasing well bore temperature (molecular weight and the ideal gas constant are
both fixed values).
After shutting in a well, the bottom hole pressure increases as the radial pressure gradient
around the well bore equalizes. Therefore the BHP after shut-in is either increasing or
stable; it cannot decrease. If WHP starts to decrease after a shut-in, the term (well depth
X gas density) must be increasing. Since the well depth is fixed, the increase must be due
to changing gas density. Over years, there have been empirical equations developed so to
allow the prediction of temperature with time for these types of wells. These equations
are the results extensive testing and consider the ambient geothermal profile of the well
location, tubing I.D., flow rate, liquid yield, liquid properties and well depth.
The plots show the modeling of BHPs with and without accounting for wellbore cooling.
The BHPs which do not account for wellbore cooling are not consistent with real well
behavior and will be impossible to accurately analyze for reservoir properties.
Phase Redistribution
In wells that are producing within the two-phase envelope only part of the separator
liquid is in solution with the gas. The rest is produced from the well when it is lifted by
the velocity of the gas. When a well like this shuts in, the liquid falls back, accumulates at
the perforations and re-injects into the formation. Liquid re-injection is a common
occurrence in gas wells that produce liquids, either condensate or water. Gas entering the
closed wellbore displaces the fluid in the wellbore back into the formation. There is a
specific pressure signature that is observed when this happens. This occurrence is
prevalent in both surface and downhole gauges, since most DHGs are not set at the top
perforation, but at the end of the tubing (as shown in figure below). Point 1 shows both a
surface gauge and DHG on a flowing gas well that produces some liquid. The well is
shut-in (2) and there is a gas (3)/liquid (4) interface. In the 2nd frame, the surface gauges
pressure readings will not show true reservoir response due to the liquid gas interface
above the top perforation. In the 3rd slide, the liquid is being re-injected, lowering the
column height in the wellbore to a point (5) where now both the surface gauge and DHG
are above the gas/liquid contact level. This masks both gauges ability to see true
reservoir response. In the final frame, the full column above the top perforation has been
re-injected, leaving a homogeneous column to both the DHG and surface gauge. One
must be able to distinguish between reservoir response, operational issue and liquid re-
injection to properly account for this obstacle. During liquid re-injection, the data is not
meaningful, but as long as no boundary has been hit, the data following the injection
period will be valid for analysis of radial flow. To reinforce the data from a PBU test
performed on a well in which liquid re-injection occurs, the subsequent drawdown should
be recorded to confirm the analysis from the PBU.
Wellbore storage is an issue that some believe is an Achilles Heel to surface testing.
Wellbore storage is the afterflow into the wellbore after the well is shut-in at the
wellhead. Wellbore storage effects last until the pressure is equalized between the
wellbore and the near formation. The duration of wellbore storage is primarily dependant
on the wellbore volume, the formation permeability and the fluid compressibility. In
most cases, this effect does not last long enough to inhibit the detection of radial flow. In
a variety of permeabilities, tests have shown that the radius of investigation at the end of
wellbore storage is roughly between 50 and 200 feet. A surface gauge or bottomhole
gauge would experience these same storage/masking effects for about the same length of
time. The only way to reduce these effects would be with a downhole shut-in tool. The
first two plots below show the occurrence of wellbore storage during a test which had
both a surface gauge and a DHG. The third plot is a Cartesian of surface data that
exhibits the occurrence of wellbore storage.
SPIDR Graph of Company: WBS Exam ple We llname : WBS Exam ple
Starting Date/Time: 3/24/2004 18:00:00 Ending Date/Time: 3/26/2004 00:00:00
WBS Ex ample.MK2
14500
14000
13500
13000
12500
12000
11500
PSIA
11000
10500
10000
9500
9000
8500
8000
7500
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Hours
Procedures
Operationally, several considerations must be made to properly run a transient test. First,
stable flow/shut-in period must be obtained. In the case of the PBU, the stable flow
period prior to shut-in will give the base line for skin calculation. A hard shut-in must be
done. A staged shut-in will lead to an erroneous skin value. In the completions where a
SSSV is in place, it must be locked open during the shut-in. This allows the reservoir
response to reach the surface. During a drawdown, the shut-in period preceding must be
recorded prior to opening the well back up. The drawdown must be done at a constant
choke size, not flow rate. This will allow the analysis of a true transient in the reservoir.
Error
The conversion method is the crux of the viability of surface well testing. By embodying
all of these issues and using a sophisticated conversion program, surface pressure-
transient well tests can be performed. In addition to the software, a skilled engineer must
process the data to ensure all aspects considered, such as flow regimes and tubing
roughness. With any conversion algorithm, there are sources of error. With surface well
testing, the error lies within the inputs (GIGO). An error in flowrate or tubing roughness
will have an effect on the skin value, since it causes a scalar shift in the pressure curve.
On that same note, an error in WHT or gas gravity will have an effect on P*. The
software conversion will only be as good as the inputs used. The following chart shows
the effect that error in the input values have on the calculated values of skin, permeability
and P*. If multiple inputs are incorrect the error is then compounded.
COMPARISON
This section will show 3 field examples of pressure-transient tests using a high
resolution/high frequency dual-quartz surface gauge ran simultaneously with a downhole
pressure gauge. A note to be made for all of these examples is that the surface gauge was
converted to mid-perforation depth while the DHG was hung off at some distance above
the perforation interval and then corrected to mid-perforation depth. This may account
for some scalar difference between the two gauges.
Example 1
This example shows a HP/HT south Louisiana well. The well specifics are as follows:
20,000 feet TVD (onshore)
~ 400 F (205 C)
~ 10% CO2
20 MMcf/D, 2900 BOPD
Layered Reservoir (2-3 zones)
The objective for the test was to find an explanation for the wells poor performance. The
rate had dropped almost 80% from the wells initial rate. For this test, both a high
resolution, high frequency, thermally compensating surface gauge and a downhole gauge
were run simultaneously. The last several hours of stable flow was captured by both
gauges and then the well was shut-in and the build-up monitored. The figures below
show the pressure-transient test, with both the surface gauge and DHG plotted in tandem.
The resultant PTA values are shown.
The surface data was converted to mid-perforation depth. As seen by the plots, the
shapes of the semi-log curves are alike and the resultant values are similar. As seen in
the Cartesian plot, no significant amount of skin was found for this well. The reservoir
pressure has dropped almost 13,000 psi since the last test performed on this well. It is
also noted that the reservoir exhibits channel behavior. These conclusions were made
from both the converted surface data and the downhole gauge data. The surface test cost
US$3,000 while the downhole gauge test cost approximately US$60,000.
Example 2
This example shows a well in the North Sea. The test objectives were not only to
determine the skin, permeability and P* values for the reservoir, but also to locate the
nearest fault/boundary and develop the correlation between the calculated BHP from
surface and the downhole gauge pressures. The test was run as an extended PBU and
both a DHG and a high-resolution, high frequency, dual-quartz surface pressure gauge
were used. The following plots show the Cartesian plot of both gauges and the semi-log.
The well test was a success for the operator and they were very pleased with the results
given by the surface gauge. The calculated BHPs were 15 psi different from the
measured BHPs. The skin, permeability and P* are equivalent and both data sets show a
boundary at approximately 30 hours. The main difference between both variations of
testing was the cost. The surface test cost US$7,500 while the downhole test cost over
US$25,000.
Example 3
In this example, a surface gauge was run to affirm a previous test that showed the
completion to have a high skin value. The reservoir depth was around 18,000 feet and
the gas stream contained 3% CO2. The well was flowing at 11MMcf/day and produced
sand at higher rates. The objectives for this well test were to verify the previous tests
evaluation of high skin and compare the calculated BHPs with a DHG.
As seen by the figures above, the converted surface gauge data tracked along the same
curve as the DHG. The calculated flowing BHPs were 150 psi lower than the measured
flowing BHPs, which may have been due to an error in flow rate measurement. There
was a minimal difference during the build-up phase of the test. As seen by the results
shown in the above semi-log plot, the decisions that were to be made on this well would
be not be different whether a DHG or surface gauge was used. One major note on this
test is seen in the plots. As the surface gauge pressure line continues the DHG pressure
curve stops around a total test time of 15 hours, 3 hours into the build-up. During this
test 2 of the 3 DHGs failed and all 3 had to be fished out of the wellbore which cost the
operator a substantial sum. Since the results were comparable during the testing, the
operator determined that he could conduct future pressure-transient tests from the surface
thereby reducing cost and risk.
CONCLUSION
This technology has also moved beyond just testing a dry tree and has been adapted to
subsea completions as well. Deepwater/subsea operators who have had to deal with
permanent downhole gauge failures or completions without a PDHG are now taking
advantage of surface measurements for PTA.
Surface well testing has come a long way since its inception and has gained industry
acceptance. By using surface acquired pressures for pressure-transient testing on a well
that fits the candidate selection criteria, an engineer can save money and decrease
operational risk while still receiving highly accurate results.
APPENDIX
Critical Unloading Rate for Small Diameter Tubing via Turner's Equation
Critical Unloading Rate for Medium Diameter Tubing via Turner's Equation
6000
16000
14000
5000
12000
4000
Critical Rate (MCF/D)
3000 8000
6000
2000
4000
ID = 2.992 (3 1/2 in, 9.3 lb/ft)
ID = 2.750 (3 1/2 in, 12.95 lb/ft) ID = 4.892 (5 1/2 in, 17 lb/ft)
1000
ID = 2.441 (2 7/8 in, 6.5 lb/ft) ID = 4.408 (5 in, 15 lb/ft)
2000
ID = 2.323 (2 7/8 in, 7.9 lb/ft) ID = 3.958 (4 1/2 in, 12.75 lb/ft)
ID = 1.995 (2 3/8 in., 4.7 lb/ft) ID = 3.548 (4 in, 9.5 lb/ft)
0 0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
Wellhead Pressure (PSIA) Wellhead Pressure (PSIA)
3000
25000
2500
20000
Critical Rate (MCF/D)
Critical Rate (MCF/D)
2000
15000
1500
10000
1000
0 0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 0 200 400 600 800 1000 1200 1400 1600 1800 2000
Wellhead Pressure (PSIA) Wellhead Pressure (PSIA)
ACKNLOWDGEMENTS
Dr. Nathan Waldman: SPE, Data Retrieval Corporation President
Rodney Kelly: Data Retrieval Corporation V.P of Sales and Marketing
REFERENCES
SPE 77701: Chris Fair, SPE, Data Retrieval Corp., Betsy Cook, SPE, Nexen Petroleum
U.S.A. Inc., Tom Brighton, BG-Group, Michael Redman, Data Retrieval Corp., Stacy
Newman, SPE, Data Retrieval Corp. (2002), Gas/Condensate and Oil Well Testing -
From the Surface
SPE 77626: Nathan Waldman, SPE, Data Retrieval Corp., Chris Fair, SPE, Data
Retrieval Corp., Charlie Tyrrell, Oceaneering International, Inc., Rakesh Kumar, Shell
(2002), Subsea Well Testing at the Subsea Tree
Cullender, M.H. and Smith, R.V. (1956), Practical Solution of Gas-Flow Equations for
Wells and Pipelines with Large Temperature Gradients
Turner, R.G., Hubbard, M.G., and Dukler, A.E. (November, 1969), Analysis and
Prediction of Minimum Flow Rate for the Continuous Removal of Liquids from Gas
Wells