You are on page 1of 6

SPE 101082

Optimized Perforation—From Black Art to Engineering Software Tool


M.R.G. Bell, SPE, and J.B. Davies, SPE, Shell Intl. E&P, and S. Simonian, SPE, FloDynamic

Copyright 2006, Society of Petroleum Engineers


The work also brought into focus critical issues such as the
This paper was prepared for presentation at the 2006 SPE Asia Pacific Oil & Gas Conference industry’s dependence on API Section I data - as a result of
and Exhibition held in Adelaide, Australia, 11–13 September 2006.
which we continue to be hoodwinked by ‘record breaking’
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, as
penetration performance into concrete targets – and the lack of
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to performance metrics in poorly negotiated long-term
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at perforating service contracts.
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
As a result, Shell has initiated research programs to build a
for commercial purposes without the written consent of the Society of Petroleum Engineers is better understanding of perforating phenomena - the results of
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 which will feed back into the software, refining its capabilities
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
and increasing the user’s ability to design and simulate
perforating operations - is driving the development of new
Abstract performance benchmarks, and is rethinking the way
While assisting production engineers in managing the perforating services are procured.
perforating process, Shell recognized the need for an
engineering software tool to guide and advise them. It needed Introduction
to address selecting the optimum perforating system for given When perforating a well - whether initially, for remedial
well and formation properties, and work at log resolution to reasons, or to effect a change in functional specification – the
eliminate problems experienced with existing packages that associated design and equipment selection decisions are
use input values averaged across the reservoir. generally made somewhat arbitrarily, and are often based on
After consultation with staff engineers, the tool was misleading input data.
created in Excel to test functionality and later transformed into The starting point in most cases is a vendor’s catalogue,
a desktop application. It includes a perforating system the service provider’s proprietary software, or the service
database built from published API data. This is not ideal, provider’s representative conveniently stationed in an office
since API Section I data are an unreliable predictor of along the corridor. All of these resources ultimately refer back
performance into stressed rock, but this is the only published to a common data source: API RP-19B Section I test data.
data allowing direct comparison between systems. The API Subcommittee on Perforating defined the test
The tool calculates depth of fluid invasion, stress-corrected protocols described in API RP-19B1 to serve a range of
penetration, and crushed zone properties at log resolution. An purposes. The Section I test, which involves firing a full gun
inflow profile and IPR curves are then generated. The section into a massive concrete target under ambient
analysis can be repeated for multiple guns to compare relative conditions, was designed as a quality control test and to check
performance; for example, to assess whether incremental for charge interference. The Section II test examines single
production predicted for a superior system justifies additional shot penetration into a stressed rock target. The Section III
expenditure. test examines performance under conditions of elevated
The user can also test the sensitivity of inflow performance temperature. Finally, the Section IV test examines single shot
to perforating system performance and input assumptions. If penetration and inflow performance under simulated
system performance is critical, this may encourage laboratory downhole conditions.
tests under representative conditions to verify penetration and Some years ago, Section II testing was used as a routine
inflow. means for comparing the relative performance of perforating
Developing the tool highlighted fundamental shortcomings charges. However, due to increasing variability in the
in the industry’s understanding of perforation, such as: how properties of Berea sandstone – the most prevalent rock used
the crushed zone is formed and removed; the role played by for Section II testing – its use has declined, with most
pressure dynamics; the sensitivity of penetration and tunnel manufacturers now relying on Section I testing as their
quality to formation and system parameters; and, how jet performance benchmark. Driven by the service companies,
perforators perform in targets other than sandstone. This lack the API further reinforced the credibility of the Section I data
of understanding has contributed to technology stagnation, by appointing witnesses to observe test execution and
sub-optimal well performance, poor decision making in the publishing tables of ‘sanctioned’ test results on the API
area of sandface completions, and undue belief in the value of website.
allegedly performance-enhancing perforating services.
2 SPE 101082

While the Section I test is clearly not a representative Each of the input parameters described in the previous
benchmark by which to evaluate expected in situ performance paragraph presents challenges. Fluid invasion will vary along
(since it measures only depth of penetration into stressed the wellbore according to formation properties, drilling fluid
concrete), the operator community has come to accept it as the type, and exposure time. Depth of penetration and the quality
only source of publically available data for comparing the of the perforation tunnel – described in this case as crushed
relative performance of different vendor’s products. Of zone thickness and permeability – will also vary according to
greater concern is that Section I test data are routinely used as rock and reservoir properties, such as compressive strength
indicators of expected in situ performance, often without any and permeability. By treating each of these parameters as a
meaningful correction for downhole conditions. constant across any significant interval – whether a reservoir
Proprietary service company software estimates downhole interval or the entire wellbore – the model immediately loses
depth of penetration by applying corrections to the Section I the granularity necessary to perform a realistic simulation.
data using empirical correlations derived from test shots made Our second observation is that to create a meaningful
into rock at different levels of confining stress. While these wellbore model linking perforation performance to inflow
correlations are modestly reliable within a reasonable stress performance - and ultimately to well performance - it must
range around the Section I test conditions (typically ~5,000 psi function at the log scale; that is, at whatever incremental depth
compressive strength), their predictive capability declines scale reservoir properties are measured.
rapidly under more extreme conditions. Section II and IV Even when reasonable wellbore models are employed as
tests have shown Section I based predictions can over-estimate part of the perforation design process, no tool is available to
depth of penetration under high-stress conditions by 100% or help the engineer compare between different vendor’s
more. products, or to compare different sandface completion options.
Apart from the inherent inaccuracy of correlations, it Nor can actual well performance data – such as production
should also be noted that the vast majority of test points from rates or inflow profiles from production logs – be readily
which they are derived relate to sandstone targets. Very little compared to predictions in order to validate (or calibrate) the
data exists for perforation into carbonates or other rock types. model for subsequent use.
Furthermore, the Section I test only measures – and the Our third observation is thus that there is an industry need
software tools can thus only predict – depth of penetration and for an engineering software tool to elevate the perforation
entry hole diameter. While perforation length is an important design, selection, and evaluation process from the realms of
factor in determining inflow performance (since it contributes black art to that of engineering practice.
to the effective wellbore radius and determines whether the With these observations in mind, the Perforating Global
perforation tunnels extend beyond near-wellbore damage such Delivery Team at Shell set out to create a software tool that
as drilling fluid invasion), it is by no means the only would ultimately serve several purposes:
significant variable. Within the API Recommended Practice, • Enable rigorous comparison of sandface
only Section IV tests provide an indication of a perforation’s completion techniques;
relative inflow (or outflow) performance. • Incorporate an impairment model to predict the
Our first observation is thus that the industry would do fluid invaded zone;
well to reconsider its reliance on Section I data as a • Rigorously model perforations and the near-
performance benchmark and as a source of input data for wellbore region;
system selection. • Calculate skin and inflow at log resolution;
Perforating system selection is therefore often based on • Allow auditable evaluation of all vendors’
nothing more than a potentially erroneous depth of penetration perforating systems;
estimate. If any further analysis is performed, it will typically • Integrate with vertical lift simulators for
involve inputing ‘perforating parameters’ into a wellbore completion design decisions;
simulator to generate inflow performance relationships (IPR).
These will subsequently be overlaid onto vertical lift profiles Description of the Software Tool
(VLP) corresponding to different completion options, and The resulting software tool, known as SPOT (Shell Perforation
used to select the combination of perforating system and Optimization Tool), was first developed on a Microsoft Visual
completion design that delivers maximum productivity. Basic/Excel platform to facilitate rapid prototyping and
Unfortunately depth of penetration is not the only suspect functionality validation. A significant proportion of the tool’s
input variable in this process. The wellbore model will constituent parts were assembled from pre-existing sources,
typically require additional perforation-related input such as such as spreadsheets, laboratory reports, papers and theses,
depth and permeability of the fluid-invaded zone, thickness and other in-house software. Some original code was
and permeability of the crushed zone formed around the developed to link these modules, as well as a basic graphical
perforation tunnel, entry hole diameter, shot density and user interface.
phasing. The reservoir will generally be subdivided into a Front line engineers were consulted at an early stage to
(limited) number of discrete intervals, for which averaged understand their immediate needs and common workflows. A
properties such as permeability, porosity, and hydrocarbon working prototype was tested with a limited number of
saturation must be entered. On the basis of these parameters, operating unit engineers some five months after project
the model will calculate skin values (a measure of impairment inception. The prototype tool was eventually frozen after six
to flow) for each interval and thereafter an inflow performance
relationship.
SPE 101082 3

months of development for translation into a C++ based production log can be imported and added to this
Windows application. plot for calibration or verification purposes.
The tool consists of the following key modules: • Any of the along hole plots in combination with
• Well data entry & log import open hole logs for visual correlation of results to
• Perforating system database formation intervals
• Drilling fluid invasion estimator SPOT also features a scenario manager, which allows the
• Perforation performance analysis engine user to test result sensitivity to variations in specific input
• Inflow performance simulator parameters. This might be used to investigate the relationship
• Graphical and tabular output between, for example, borehole size or static underbalance and
The well data required by SPOT are similar to those used inflow potential.
by any wellbore simulator except that rock and reservoir Finally, having assessed the inflow potential delivered by
properties are obtained directly from open hole log data, rather different systems, SPOT includes a basic economic analysis
than averaged or constant values. module to assist the user in developing the business case for
The perforating system database contains specifications deploying a particular perforating solution.
and performance data for a wide range of vendors’ systems.
Unfortunately the bulk of this information is still derived from Examples
Section I tests, but stressed rock (Section II) and inflow data Although the SPOT project was only initiated in April 2005, a
(Section IV) are provided where available, and can be beta version of the Windows application has been undergoing
augmented whenever the user obtains test results under local fieldtesting since late March 2006. Particular focus has been
conditions. Note that, should SPOT be made available to given to matching SPOT outputs to well histories that include
users outside Shell, proprietary data – such as Section IV pre- and post-perforating production logs. Unfortunately such
results – would have to be restricted to users from the complete datasets are uncommon, but sufficient confidence
originating company’s staff. has been gained from the available data for SPOT to be
The drilling fluid invasion model uses a proprietary applied more widely.
algorithm to produce an invasion profile across the interval of
interest, accounting for drilling fluid type and rheological Example 1 – Optimized Perforating Interval
properties, exposure time, and formation properties. This In our first example, an asset team was faced with
model has been developed and refined by Shell over several perforating a long, heterogeneous, gas-bearing sandstone
years, and has been calibrated to field and laboratory data. interval. While it would be desirable to perforate as much of
To initiate the analysis process, the user either selects one the interval as possible in order to guarantee meeting gas
or more perforating systems directly from the database – delivery targets (expected to be > 300 million scf/d per well),
which can be filtered according to a number of different this could mean a total perforated length in excess of 1,100ft.
criteria to reflect prevailing geometric, contractual or other Such an operation would require either coiled-tubing
constraints – or chooses ‘quicklook’ mode, in which SPOT deployment using CIRP connectors to allow retrieval under
will analyse all viable systems and present those offering pressure, or drill pipe deployment using gun anchors to allow
greatest potential. The user may also compare other for slickline retrieval.
completion methods such as barefoot, slotted or pre-drilled By virtue of working at log resolution, SPOT provided a
liner, gravel pack etc. much more detailed inflow profile prediction than was
Using rock strength and other log-derived parameters, previously available. By comparing total inflow potential for
SPOT estimates downhole penetration, crushed zone thickness different perforating systems and lower porosity cut-off
and permeability, and entry hole diameter at each log depth for values, the team showed that a reduction of over 60% in
each perforating system. Combining this with the invasion perforated interval would result in less than 5% reduction in
profile, SPOT then calculates component skin factors and well potential. The new total perforated length would be
inflow potential for each log-scale increment using established around 400ft, making multiple electric wireline runs a viable
Darcy and Forcheimer equations. The incremental inflow solution, greatly reducing the cost, complexity and risk of the
values are finally summed across the entire interval to perforating operation.
generate inflow profiles and inflow performance relationships
for each system. Example 2 – Validated Well Performance
The relative performance of different completion options is Our second example features a field where a horizontal
displayed on a series of plots, which include such information sidetrack had been perforated on the same interval as an offset
as: vertical well but delivered no measurable improvement in
• Depth of penetration and depth of invasion, productivity despite its geometric advantages. The team was
overlaid against along hole depth (see Figure 1) planning a new horizontal well and was concerned that some
aspect of the completion process had impaired the earlier
• Sand failure risk, showing near-wellbore
sidetrack. A SPOT analysis was performed to try and match
effective stress overlaid with rock strength
the existing well performance and optimize the completion of
• Inflow performance relationships for each
the new well.
system, overlaid on the same plot
Starting with the simplest case, a SPOT model was built
• Inflow profile for each system, overlaid against
for the vertical well, including perforating with the obsolete
along hole depth (see Figure 2). Note that a
gun design in use when the well was drilled (the SPOT system
4 SPE 101082

database includes obsolete systems for exactly this situation). production, tunnel collapse or plugging)
Initial results suggested a productivity index (PI) much higher Calculations to assess performance after some
than was actually observed, but careful review of the log data period of production currently require multiple
identified spurious multi-Darcy permeability streaks as the static runs at different pore pressures
source of unrealistic potential. The field geologist confirmed representative of depleted conditions;
that such streaks were artifacts of the porosity-permeability • The need for a better understanding and
transform and were not observed in cores. After conditioning representation of potential interference effects
the log data to remove these artefacts, a PI of 10 sm³/d/bar was between adjacent perforations;
obtained, respectably close to the value of 10.8 sm³/d/bar • The need to represent the influence of natural
interpreted from well tests. fractures, other pertinent geological features, and
A SPOT model was next built for the horizontal sidetrack, mechanical problems such as poor cement bond
applying the log conditioning learning while maintaining other or damaged casing;
variables consistent between the two wells. A good • Independence from other completion operations –
productivity index match was once again obtained, confirming need to link to other workflow components, such
that the vertical and horizontal wells did indeed exhibit similar as stimulation, fracturing, and sand control
productivities. The primary reason for the apparent under- simulators;
performance of the horizontal sidetrack was found to be its • Independence from other elements of the
reduced conduit size – 4½” liner in 6” hole versus 7” liner in production system – need to link to reservoir,
8½” hole – rather than any impairment mechanism. near-wellbore, and vertical lift simulators to
Applying the cumulative learnings from these two models, produce an integrated engineering tool with
the team next constructed a SPOT model for the planned new appropriate levels of detail in each component;
horizontal well. It would feature an 8½” lateral with a 7” It is clear that SPOT will not necessarily become the
liner, and would be perforated with the incumbent service ubiquitous tool for perforation engineering. At the time of
provider’s premium 4½” tubing-conveyed perforating system. writing, discussions are ongoing regarding the potential for
For this case, a productivity index five times that of the releasing our in-house tool to a wider user group. Irrespective
horizontal sidetrack was predicted. While such a PI might not of whether such a step is ever taken, however, we expect the
be achieved in practice without encountering other problems, benefits demonstrated by SPOT will encourage the providers
such as water coning or upper completion constraints, the of existing perforating software packages to enhance their
analysis gave the team confidence that the proposed well capabilities along similar lines. It remains to be seen whether
design and completion strategy should yield significantly they will incorporate competitor system performance data in
greater potential than might have been feared in light of the order to overcome the inherent disadvantage presented by only
horizontal sidetrack’s poor performance. being able to analyse one vendor’s products with a given
These examples illustrate only two of many possible SPOT package.
applications within completion design and analysis workflows. In researching the need for representative perforating
SPOT is continuously being adapted and improved by the system performance data, we noted with interest that many of
team to cater for new applications as they are uncovered. the manufacturers are acutely aware of the inadequacy of
using Section I test data as either a performance benchmark or
Discussion input parameter for simulations. However, they are reluctant
While SPOT is a major step forward in the quest to rigorously to move to a stressed rock benchmark in part because their
model the perforation process and its impact on completion products will appear to underperform relative to those of
performance, a number of fundamental issues remain to be competitors who do not adopt a similar approach. This
solved. We view SPOT as a vehicle, or framework, within situation is propagated by ignorance among the customer base,
which discrete modules can be substituted as more where many perforating products are purchased on the basis of
sophisticated components are researched and developed. Section I test performance only. Until the problems with this
Areas for further improvement of particular interest or approach are more widely understood and accepted by the
concern are (in no particular order): operator community, it is difficult to envisage a general shift
• Dependence on Section I data – may be mitigated away from Section I data.
by providing alternative input data better However, we see an even more significant paradigm shift
representing in situ perforator performance; on the horizon. At recent industry meetings on perforating, a
• The use of a somewhat arbitrary “perforation consistent call has been heard for systems and test protocols
efficiency” parameter – requires improved optimised and designed for inflow, rather than depth of
understanding of fundamental processes during penetration. With this comes a supporting request for
perforating and tunnel cleanup, for multiple target perforating product specifications addressing the performance
media and mechanical geometries; that can be expected of a given product under a variety of
• SPOT is a static model and should ultimately operational conditions. These challenges will take time to
include dynamic processes, both during properly understand and address, although work on perforator
perforating (e.g. dynamic pressure and flow product specifications has been started before, probably
effects) and later in perforation life (e.g. effective stalling for lack of general industry attention.
stress changes, water breakthrough, sand
SPE 101082 5

Ultimately, with our engineers better educated and


properly equipped with design and decision making tools, we
see operators preferentially purchasing those perforating
systems – at the high-performance end of the product range, at
least – that are supported by credible product specifications
and representative tests demonstrating inflow potential. This
may result in a differentiated market featuring premium
systems with certified inflow performance (as well as depth of
penetration and other familiar parameters) and second-tier
commodity systems sold on the basis of lesser benchmarks.

Conclusions
In response to an observed industry need, the Perforating
Global Delivery Team at Shell has created an engineering
software tool focused on perforation design, selection, and
evaluation.
The tool serves several purposes including: comparison of
different sandface completion techniques; using an
impairment model to predict the fluid invaded zone;
rigorously modeling perforations; calculating skin and inflow
profile at log resolution; and allowing an auditable evaluation
of different vendors’ perforating systems. Work is ongoing to
integrate the package with vertical lift simulators to facilitate
completion design decisions.
Early applications of the prototype package have shown
significant benefits through: superior resolution, revealing
effects previously masked by layer averaging; the ability to
model, match and diagnose existing well performance;
rigorous comparison of all available technical solutions and
systems, irrespective of vendor; detailed sensitivity analysis
on input variables previously assumed to be constants; and
enhanced operational learning and performance as a result of
following structured design and decision making processes.
Further work is required to address known shortcomings in
the tool. Many of these are symptomatic of industry
paradigms that must change if perforating systems are to be
properly optimized and applied. Work is also required to
develop a better understanding of the fundamental processes
and phenomena associated with perforating, notably into
targets other than sandstones.
We hope that by publishing the rationale behind the
development of this tool and some of the methodology around
which it has been constructed we will trigger further debate on
the underlying issues, a renewed interest in perforating
fundamentals, and a general move to enhance the perforating
tools and products available to front-line production engineers.

Acknowledgements
The authors wish to thank Shell International Exploration &
Production for allowing publication of this paper. We would
also like to acknowledge Dr. Sam Simonian of FloDynamic
for his assistance in developing SPOT and in the preparation
of this paper, Dr. Hans van Velsen for developing the drilling
fluid invasion model, and the staff at QinetiQ (UK) who
helped develop perforation characterisation models.

References
1 American Petroleum Institute, “Recommended Practice for
Evaluation of Well Perforators”, Recommended Practice
19B, 1st Edition, 28 Sep 2001
6 SPE 101082

Figures

Figure 1 - Graphical Output Example Showing Depth of Penetration & Depth of Invasion Against Depth

Figure 2 - Graphical Output Example Showing Calculated Inflow Profiles Against Depth

You might also like