Reentry Drilling Gives New Life to Aging Fields

A recent burst of technical creativity has produced an abundance of new ways to revitalize old fields and

tap bypassed pockets of oil and gas. However, identifying the best solutions requires a team of experts

with a broad range of skills that cross the traditional boundaries of petroleum engineering disciplines.

David Hill
Eric Neme
Christine Ehlig-Economides
Sugar Land, Texas, USA

Miguel Mollinedo
OXY
Maracaibo, Venezuela

Over the last ten years, new technologies

and field strategies have converged,
enabling operators to give new life to old
wells. Now, reviving production from
declining fields has become a major activ-
ity for oil and gas companies, and one that
requires more support to identify the right
technical solutions. Optimizing well output
and economics are the key goals of these
production enhancement projects and ser-
vice companies are actively participating
in achieving these goals.
This growing demand has pressed compa-
nies in the service sector to diversify their
skills and address a wider range of reser-
voir and production problems. It has also stimulated a flurry of technical creativity. To help operators address these questions,
For example, developments in the area of service companies have reorganized to pro-
reentry drilling alone—coiled tubing vide multiple integrated services.1 With this
For help in preparation of this article, thanks to Olivier
Fabvre, Anadrill, The Hague, The Netherlands; Dave drilling (CTD), slimhole measurements- broader outlook comes an extended range of
Bergt, Jaime Bernardini, Ike Nitis and Pearl Chu Leder, while-drilling (MWD) systems and new capabilities, including identifying underper-
Anadrill, Sugar Land, Texas, USA; Jon Elphick and Andy
Rike, Dowell, Sugar Land, Texas; and Chris Prusiecki, completion technologies for multiple side- forming wells and recommending cost-effec-
Anadrill, Dallas, Texas. track boreholes—have produced a wealth tive interventions to increase well productivity
DESC (Design and Evaluation Services for Clients),
NODAL (production system analysis), PowerPak (steer-
of options for maximizing return on invest- and maximize net present value (NPV).2
able motors), RAPID (Reentry and Production Improve- ment (ROI). But which approach offers the With improved capabilities from new
ment Drilling), Slim 1, VIPER and VISPLEX are marks of
Schlumberger. A-Z PackStock is a mark of Smith
best solution; how should it be applied; drilling technologies, a growing number of
Drilling & Completions. and in which wells? wells are candidates for reentry drilling—

4 Oilfield Review

Anadrill
@,,@
¢¢¢¢¢¢






,,,,,,
RRRRR
QQQQQQ
£££££
Sidetracking
• Section milling
• Window opening


TTTTT
SSSSSS
¥¥¥¥¥
¤¤¤¤¤¤
Completion
• Connectivity
• Isolation
• Access
• Liner hanger
■ Reentry systems. RAPID services cover the key elements of reentry and multilateral
drilling, from pulling old completions to installing the new one and from drilling fluids
to wireline logging.
short- or medium-radius sidetracks and
multilaterals, drilled conventionally or with
coiled tubing. This year, in the USA alone,
more than 1500 reentry sidetracks will be
drilled. By 1999, the number is expected
to increase by 25%.3
Revisiting Existing Wellbores
Reentering wells to gain additional produc-
tion is not new. Since the mid-1950s, oil
companies have reentered old wells and
drilled sidetracks to bypass formation dam-
age or wellbore mechanical problems, and
also to exploit new zones, saving the
expense of drilling entirely new wells.4
Recent expansion of the reentry drilling mar-
ket, however, owes much to improvements
in drilling and completion technology.
Reentry drilling provides a means to
reduce horizontal well costs. In addition to
boosting well productivity, reentry drilling
can also tap bypassed reserves (top right).
Multiple lateral sidetracks can fan out from
an existing wellbore for enhanced access
to reservoirs (middle right). And smaller
isolated pockets of oil and gas can be
tapped by extended-reach wells or multi-
laterals (bottom right). Typically, a hori-
zontal well will triple or quadruple
productivity over a vertical well, and in
some cases, much larger productivity
improvements—up to 17-fold, or more—
Autumn 1996
Producing Bypassed Oil
By-passed zone Short-radius
sidetrack
Completion
• Sand control
Curve drilling • Zone isolation
• Flow control Oil
• Medium radius
• Short radius • Gravel pack
Depleted
Water
zone
■ Improving net present value of old
fields. Reentering wells and drilling hori-
zontal laterals into bypassed zones can
Lateral drilling
tap new reserves from existing wellbores.
• Geosteering
• Drill-in fluids
• Underbalance Optimizing Recovery
have been observed. Additionally, in zones
with underlying water, overlying gas, or
both, horizontal wells can significantly Lateral
sidetracks
increase recoverable reserves.5
Today, service companies use various
approaches to address the growing
demand for reentry drilling. Baker Hughes
INTEQ boosted its reentry drilling services
with support from sister company Baker
■ Multiple sidetracks for enhanced pro-
Oil Tools, and gained a reputation as a duction. Additional drainholes (red) can
reentry specialist in the Gulf of Mexico. fan out from existing wellbores or horizon-
Within Schlumberger, RAPID Reentry And tal trunks and improve reservoir drainage.
Production Improvement Drilling teams
were created to address this fast-growing Tapping Remote Structures
drilling option. Service under the RAPID
umbrella draws on expertise in reservoir
engineering, drilling, directional drilling,
fluids engineering, petrophysics and com-
pletion engineering—the indispensable ele- Extended reach
ments required to plan, drill and complete Subsea multilaterals
successful reentry laterals (above).
Multilaterals
1. Chafcouloff S, Michel G, Trice M, Clark G, Cosad C
and Forbes K: ”Integrated Services,” Oilfield Review
7, no. 2 (Summer 1995): 11-25. ■ Making small fields economical. Innov-
ative drilling techniques can improve
2. Net present value is today’s value of an asset
accounting for all future expenditures and income. asset value by tapping small pockets of
3. Data from Spears & Associates: “Drilling and
oil. Using the latest downhole motor and
Production Outlook,” March 1996. geosteering technology, wells extending
several kilometers from offshore platforms
4. Maurer WC: ”Recent Advances in Horizontal
Drilling,” The Journal of Canadian Petroleum can be drilled, eliminating the need for
Technology 34, no. 9 (November 1995): 25-33. additional structures. Multilateral wells
5. Ehlig-Economides CA, Chan KS and Spath JB: that branch out from a main wellbore can
”Production Enhancement Strategies for Strong access several areas of a field and elimi-
Bottom Water Drive Reservoirs,” paper SPE 36613, nate the need for new wells.
presented at the 1996 SPE Annual Technical
Conference and Exhibition, Denver, Colorado, USA,
October 6-9, 1996.
5
 Well Path
Drainage Slanted well Horizontal well
Volume
Characterization

Permeability (k), md Vertical, k v Horizontal, k h

Thick and Poor kv Good kv
Shale
homogeneous,
no gas cap or
aquifer Oil

kv kh ≥ 0.1
Shale

Thick and Shale Closely spaced

Gas
homogeneous, Not recommended: parallel wells preferred
with gas cap Oil risk of premature gas
and/or aquifer Water or water production

Layered kv kh ≥ 0.1 Stacked parallel wells,

preferred over vertical with branch flow
conformance

Laminated Intersect as many Not recommended:

Naturally fractured
Naturally fractured
under waterflood
Structural compartment
Stratigraphic compartment
Elongated compartment
(plan view)
Attic compartments
layers as possible risk of disappointing
productivity or
recovery due to low
vertical permeability
■ A skilled team of experts. Members of the initial RAPID team,
Intersect vertical and Horizontal well normal
based in Sugar Land, Texas, (front row, left to right) directional
drilling engineer, Ike Nitis; team leader, with drilling experi-
horizontal fractures to fractures preferred
ence, Eric Neme; completion and drilling engineer, Mark
Stracke; reservoir engineer, Christine Economides, (back row)
multilateral completions engineer, Herve Ohmer and fluids
engineer, David Anderson.
Closely spaced
short parallel
wells normal to The RAPID service was established in 1995
fractures Water injection wells by a business development team in Sugar
One well in multiple Drain each with one Land, Texas, USA (above). The lessons learned
compartments or more wells and the organizational support structure that
has developed are now being duplicated in
locations worldwide, tapping key specialized
skills within all six Schlumberger Oilfield
One well in multiple Drain each with one Services companies. In microcosm, the func-
compartments or more wells tions of the RAPID group reflect the state of the
art in reentry drilling services today.
In some wells, production enhancement
is best achieved without drilling. To
Multiple well paths Single well traversing address this need, a targeted effort on pro-
slanting from single multiple channels
main trunk duction enhancement was also initiated by
Schlumberger. The front line of this effort is
led by an integrated, cross-product-line
team of engineers engaged in identification
Single well travers- One well bed drilled of candidate wells. This Production
ing multiple beds on strike preferred
Enhancement Group, or PEG, is chiefly
responsible for candidate recognition and

■ Optimizing production. Reservoirs can be classified by drainage

volume (left). For each reservoir more than one well type—vertical,
OILFIELD
hydraulically fractured REVIEW
vertical, slanted, horizontal, hydraulically
fractured horizontal, AUTUMN
and multiple 96or stacked laterals—may be
effective. Depending on permeability and reservoir characteristics,
slanted and horizontal reentry drilling are two methods for improv-
ing production and recovery (center and right).
Oilfield Review
6
Water Coning
■ Water coning during production and
breakthrough if perforations are too close
to the oil-water contact. Reducing produc-
tion rate decreases drawdown pressure
and mitigates coning.
Water Cresting
■ Effectively producing from horizontal
laterals. Drawdown pressure in a horizon-
tal lateral is lower than in a vertical well-
bore for similar production rates.
solution design. 6 PEG engineers perform
technical and economic analysis of prob-
lem wells and fields, and then design, with
the support of appropriate experts, such as
RAPID teams in the case of reentry drilling,
the optimal solution. Depending on the
production problems encountered, solu-
tions may include new well logs, reevalua-
tion of existing logs, drilling new or reentry
wells, reperforating, well stimulation treat-
ments or other workover techniques.7 The
goal is to provide the best-in-class service
for every production problem.
Autumn 1996
Candidates for Reentry Drilling
Fracturing, reperforating, removing damage
with acid, and recompletion are all widely
used methods to increase production in
existing wells, thereby improving the NPV of
old fields. Now, reentry drilling is generating
high interest for its potential to improve
recovery from damaged or depleted zones,
and tap new zones at lower cost.
So when should reentry drilling be used?
Many times, traditional techniques may
have already been tried unsuccessfully or
may not be advisable. In older wells, reen-
try drilling is the best option when there is
an identifiable reason for a slanted or hori-
zontal well path (previous page, left).
Reentry drilling from an existing wellbore
is less expensive than a new well. And it
has the advantage that borehole trajectory
through the production zone is near the
original wellbore where more is known
about the reservoir from cores, logs, test
measurements and production history.
When the existing wellbore passes
through or near a gas cap or underlying
aquifer, excess gas or water production usu-
ally develops. In the absence of a gas cap, a
traditional strategy to delay bottom water
breakthrough is to perforate near the top of
the productive interval. However, the pres-
sure gradient due to radial flow toward the
well is often sufficient to draw water
upward in the shape of a cone (above left).
Once water reaches the deepest perfora-
tions, it may be preferentially produced
because of higher mobility.
Even in the absence of a higher mobility
contact, the strong bottom waterdrive can
cause excess water production.8 Because
horizontal wells drilled near the top of an
oil zone and above the oil-water contact
produce a linear pressure gradient normal
to the well path, bottom water will rise in
the shape of a crest instead of a cone (left).
The advancing crest-shaped water front
displaces more oil than a cone-shaped
advance, which leads to greater recovery
by virtue of flow geometry.
In formations where sand control is
required, reentry laterals may avoid the
need for expensive gravel-packed comple-
tions to improve production rates while
minimizing sanding problems. Compared
to vertical wells, horizontal wells allow the
same or higher production rates at greatly
reduced drawdown pressures.
Another reason for reentry drilling is to
gain better access to layered reservoirs. If
individual pay zones are
thick enough to be tar-
geted by horizontal wells,
multiple stacked reentry laterals are a
highly effective strategy. To balance pro-
ductivity—barrels per day per unit of pres-
sure drop—from reentry laterals, each
drainhole can be drilled to an appropriate
length inversely proportional to the flow
capacity of that particular layer.
At less cost than stacked horizontal later-
als, a slanted borehole boosts productivity
of layered formations. By
designing wellbore tra-
jectory with more drilled length in less-
productive layers, some conformance
control—balanced productivity from indi-
vidual zones—can be achieved. However,
if early water breakthrough occurs in a
high-productivity layer, the relative ease of
shutting off production from one of the
stacked laterals compared to shutting off
production from a mid-length section of a
slanted well may, in the long run, favor
using a stacked lateral strategy.
A slanted well can produce a marginal
increase in productivity over a vertical well
in laminated formations
where beds are too thin
for horizontal drilling. Often hydrocarbon
zones are missed or not produced in origi-
nal completions. Such intervals can be
reperforated and a hydraulic fracture may
significantly improve productivity. However,
when the interval is thin, reentry drilling of a
horizontal lateral will outperform a
hydraulic fracture.
6. A production enhancement article will appear in a
forthcoming issue of Oilfield Review. Also see the dis-
cussion of the DESC Design and Evaluation Services
for Clients program: Baltz J, Bumgardner S, Hatlen J,
Swartzlander H, Basham P, Blessen A, Sarrafian F,
Schneider M, Clayton D, Frank T, Gordon D, Taylor B,
Kniffin M, Mueller F, Newlands D and White DJ:
“The DESC Engineer Redefines Work,” Oilfield
Review 7, no. 2 (Summer 1995): 40-50.
7. Ehlig-Economides CA, Mowat GR and Corbett C:
”Techniques for Multibranch Well Trajectory design
in the Context of a Three-Dimensional Reservoir
Model,” paper SPE 35505, presented at the
SPE/Norwegian Petroleum Society 3-D Reservoir
Modeling Conference, Stavanger, Norway, April
16-17, 1996.
8. Mobility is the ratio of permeability to viscosity.
Low-gravity crude oils have high enough viscosity
and hence, lower mobility than formation water.
7
 In some reservoirs, stratigraphic compart- Candidate Recognition Analysis
mentalization due to depositional processes Field A B C D E F
may account for bypassed Net thickness, m 14 37 19 6 8 9
hydrocarbons both verti- True vertical depth, m 2400 2400 2800 2800 2500 2400
cally and horizontally. Permeability, md 3 3 1.5 5 4 6
Vertical permeability good in average good poor poor poor
Facies with considerable upper
contrasts in flow charac- Pressure, psi 4000 3000 3800 4000 3689 3615
teristics may serve as bar- Initial pressure, psi 4335 3703 4262 4144 3792 3615
Vertical skin 5 12 10 5 5 5
riers or conduits. In some cases, reservoir
sands may be too thin to be individually
Production rate, m3/d
identified in a seismic section, but have suf- Vertical well (damaged) 24 22 23 38 37 69
ficient areal extent to be visible in seismic Vertical well, skin = 1 41 63 56 59 58 99
amplitude maps for a given structural hori- Vertical well after fracturing 70 94 86 98 97 156
zon. In such cases, horizontal wells may be Horizontal well, forecast 156 95 169 242 236 323
an ideal strategy for producing thin forma-
Production ratios
tions and for extended reach into remote Horizontal/damaged vertical 6.5 4.3 7.3 6.4 6.4 4.7
hydrocarbon sands. Horizontal/vertical, skin = 1 3.8 1.5 3.0 4.1 4.1 3.3
A major application of horizontal wells Horizontal/fractured vertical 2.2 1.0 2.0 2.5 2.4 2.1
has been in naturally fractured formations Payback (days) 20 33 18 13 13 10
like the Austin Chalk in ■ Candidate recognition in the Former Soviet Union (FSU). Field data (top) are used to
south Texas. When hori- calculate production rates (middle) for various well scenarios, including vertical well
zontal wells are drilled normal (perpendic- with original damaged skin, vertical well with skin reduced to unity, vertical well after
fracturing and reentry horizontal well with a skin of unity. Only horizontal wells with
ular) to natural fracture planes, they provide predicted production improvement that was greater than two times the fractured verti-
an excellent plumbing system for enhanc- cal well case were considered as lateral reentry candidates. These wells would have
ing production. Locating natural fractures the fastest payout time.
and determining their orientation are cru-
cial to getting the best well design in these different reservoir bodies as possible. The tion as long as possible. Each well can also
formations. A horizontal well normal to latter case implies drilling in a direction be drilled to the optimal productive length
natural fractures usually provides better normal to the elongation, which, for a flu- within the formation.
productivity than a vertical well stimulated vial reservoir, means drilling perpendicular
by hydraulic fracturing. Although natural to the downhill direction at the time of Reentry Candidate Recognition in Action
fractures are usually vertical, shallower deposition. Another approach might be The Western Siberian region in the Former
reservoirs and overpressured zones may multibranch wells, designed to target chan- Soviet Union contains reservoirs that have
have horizontal fractures open to flow. In nels identified with borehole seismic mea- been produced for 10 to 50 years using
these formations, vertical and slanted wells surements in the horizontal trunk well. conventional vertical wells. Often a simple
are reasonable choices. However, in over- Another application for horizontal drilling workover, such as reperforating, acid stim-
pressured deep formations, it may be advis- deals with a special structural geometry ulation or hydraulic fracture treatment, sig-
able to prop the natural fractures open to called attic compartments. nificantly improves production. But in
avoid loss of productivity as production In these cases, steeply dip- some cases, a better solution is to reenter
proceeds and pore pressure declines. ping beds may be in con- existing wells and drill a horizontal lateral.
Elongated reservoirs can be the result of tact with an up-dip gas cap or down-dip In September 1995, the RAPID team was
fluvial deposition or significant faulting. aquifer. One strategy is to drill a horizontal approached to assist in choosing the best
Both environments are well that passes through several beds, but option for layered reservoirs with thick oil
natural candidates for stays sufficiently below up-dip gas or above columns, where, typically, vertical wells
horizontal drilling. In down-dip water. Although this would seem penetrate the entire productive thickness.
either case, there are to be an efficient approach, it suffers distinct Reservoirs are then progressively drained
apparent drilling strate- disadvantages. Flow is commingled among from the bottom up, plugging back and
gies, depending on the objective for the layers, and gas or water breakthrough will abandoning depleted zones over time.
well. For example, wellbores can be main- interfere with production from other layers. Production from vertically isolated zones is
tained in an elongated reservoir body, or A better strategy might be to drill multiple never commingled in any well.
directionally drilled to encounter as many horizontal wells, each on strike and staying
in a given bed. The advantage of this
approach is that each well maintains an
optimal distance from gas-oil or oil-water
contacts, thus delaying multiphase produc-

8 Oilfield Review
 To accommodate this request, a question-
naire was designed to collect data from
several reservoirs. Six wells were selected
that appeared to be particularly promising.
For each of the six cases, productivity
improvement expected from a horizontal
lateral was calculated (see “Evaluating
Productivity Improvement,” right). Because
vertical wells had been drilled through the
entire productive oil column, shallow
zones were damaged during drilling as
mud weight was increased to reach total
depth. The sensitivity to skin damage was
investigated to compare production rate
improvements that could be achieved from
a vertical workover, hydraulic fracture and
horizontal lateral.
To evaluate the potential productivity
improvement from a horizontal well reen-
try, a lateral drainhole length of 750 ft
[229 m] was assumed for all cases. An ideal
target skin of unity in the lateral was
assumed for productivity comparisons (pre-
vious page). Only horizontal wells calcu-
lated to be twice as productive as
fracture-stimulated vertical wells were con-
sidered as candidates for lateral reentry.
The most favorable production enhance-
ment plan called for medium-radius drilling
with VISPLEX drilling fluid, and completion
of the lateral section with a predrilled liner.9
Proof of the validity of this approach will
come from results of the drilling program,
scheduled to begin later this year.
An interesting application for reentry
drilling in difficult structures occurred in
north Texas, where, the operator, TRIO, was
drilling vertical wells through mound-
shaped reefs. The reefs are seen on 3D seis-
mic surveys, but hydrocarbons have
migrated into traps, caused by dolomitiza-
tion, which cannot be identified by seismic
surveys. Wells are usually drilled into the
center of the reefs, but this is somewhat of a
hit-or-miss proposition.
9. VISPLEX mud (containing a mixed-metal hydroxide) is
a high shear-thinning (thixotropic) drilling fluid, pri-
marily used for milling windows in casing, which is
also used as a drilling and a drill-in fluid. The mud-
cake produced is easily removed from the formation.
Autumn 1996
Evaluating Productivity Improvement
7500
Two 300-ft
stacked laterals
Bottomhole flow flowing pressure, psi
6500
2400
ft
Tubing head pressure = 1678 psi 1200
ft
600 ft
5500 300 ft
Tubing head pressure= 300 psi
4500 Single slanted
laterals
3500
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
Liquid flow rate, B/D
■ Stacked laterals compared to slanted wellbores. A NODAL sensitivity analysis compares two stacked lateral
boreholes to various lengths of a single slanted wellbore path through two thick, clean sands in a Gulf of Mex-
ico reservoir.
In the Gulf of Mexico, there are many clean sands inflow performance relationship (IPR) to slanted
with high permeabilities—often in excess of 1000 or horizontal wellbore length. Because of fric-
md—but completion designs must provide sand tion-induced pressure drop in the small internal
control. A typical example illustrates the use of flow diameter, the IPR curves converge for longer
reentry drilling under these conditions. tunnel lengths, and there is little productivity
A previously drilled well path was deviated at gain between drilling a 1200-ft [366-m] and a
about 35˚ through the productive sand and 2400-ft [732-m] hole. The red curve is the total
hydraulically fractured for stimulation and sand productivity of two 300-ft [91-m] stacked laterals,
control. The post-treatment well test indicated a one in each layer. Because of the shorter length,
high skin of 40 and a permeability of about 180 md. and therefore less frictional resistance, the two
Because the reservoir contained two approximately stacked short (300-ft) laterals should outperform
40-ft [12-m] thick, clean sands separated by a one long (2400-ft) slanted well.
shale bed, the question was whether to design a This illustrates the impact of tubing diameter
slanted reentry well or two stacked laterals. on reentry laterals in high-permeability forma-
Since the design was for a reentry well, lateral tions. Since drilling horizontal or slanted wells
diameter was limited to 6 in. [15 cm]. The lateral increases production rates, frictional pressure
completion called for a prepacked screen and drop in the tubing or lateral can limit production
gravel pack for sand control, leaving the internal potential. In this case, another solution could be
flow diameter at just under 2 in. [5 cm]. A NODAL to plan to produce the lateral or laterals at a
sensitivity study for this case shows two families lower drawdown pressure. This solution could
of curves (above). The green curves show the avoid the need for expensive sand control mea-
effect of lowering surface pressure on vertical sures—prepack screen or gravel pack. Net pre-
flow performance. The steep climb at high rates sent value analysis, accounting for the costs of
suggests, to experienced reservoir engineers, various options and coupled with production fore-
that larger tubing would allow higher flow rates. casts for each design, can provide a way to select
However, the cost of replacing tubing was prohib- the optimal solution.
itive. The blue curves show sensitivity of the
9
 Well Performance Analysis
Tubing size, in.
Flow diameter, in.
500
400
Wellhead pressure, psig
300
200
100
0
0 1000
■ Well performance analysis. In RAPID reservoir analysis, selecting a production
improvement plan begins with well performance matching. In this example, the well
inflow performance relationship (IPR)—wellhead pressure versus flow rate—includes
several tubing uptake curves. Flow rate can be significantly increased by changing to
larger diameter tubing.
After a dry-hole vertical well was drilled,
Anadrill was approached to plan a sidetrack
from the vertical well, building angle
quickly to laterally traverse the reef and
increase the chance of intersecting areas of
vugs—large spaces in the formation—that
hold oil. The well had been drilled with a
77⁄8-in. vertical hole through the reef, but
because of the small areal size of the struc-
ture, only a maximum 500-ft [152-m] hori-
zontal displacement was available for a
lateral borehole. It is difficult to get a long-
or medium-radius sidetrack turned in such a
short distance and it is also a challenge to
kick off with a small-diameter drill bit inside
such a large open hole.
The proposed solution was unique. The
hole was plugged with cement to about
100 ft [30 m] above the planned kickoff
point (KOP). A smaller, 63⁄4-in. pilot hole
was drilled to the KOP with a 43⁄4-in. bot-
tomhole assembly (BHA). Then a 61⁄2-in. bit
was placed on the BHA with a 43⁄4-in., 3°-
bend motor. The smaller bit was used to
prevent damage to the cement pilot hole
while running in to the KOP with the bent
motor. The BHA drilled the curved section
at a rate of 27°/100 ft and found hydrocar-
10
2 7/8 3 1/2 4 1/2 5 1/2 7
2.441 2.992 3.958 5.012 6.276
(installed in well)
Vertical well inflow
performance curve
Tubing uptake curves
2000 3000 4000 5000
Gas rate, Mscf/D
bons at about 62° inclination. The reentry
sidetrack turned a $230,000 vertical dry
hole into a well that produced 200 BOPD.
Sidetrack cost, including completion, was
about $140,000.
Another example comes from a major
oil company in Houston, Texas, that asked
the RAPID team for horizontal drilling
recommendations in the difficult condi-
tions of a south Texas gas field. The reser-
voir was depleted to 300 psi [2070 kPa] at
a depth of 10,000 ft [3048 m]. Even
drilling with air would result in severe
overbalance conditions that could dam-
age the reservoir. Although coiled tubing
drilling was the only practical drilling
technique, anticipated production would
not justify the cost of this option.
The RAPID team examined well condi-
tions and field performance, and discov-
ered that the 15-year-old completion design
used in the 80 producing wells of this field
contained a flow restriction that limited
production rates. Well performance analy-
sis indicated that reengineered completions
using larger tubing would double produc-
tion rates (left). The implemented solution
cost 95% less than horizontal drilling with
coiled tubing and was immediately avail-
able for every producing well in the field.
Gas production from wells worked over
according to this recommendation doubled
from about 1 to 2 MMscf/D.
Reentry Drilling Systems
When reentry drilling is the optimal solu-
tion, one of the first decisions is to choose
between conventional and coiled tubing
drilling (CTD). Through-tubing reentry and
underbalanced CTD is an economical solu-
tion for drilling and workover operations on
rigless platforms. Underbalanced drilling
minimizes formation damage and increases
drilling penetration rates.
The majority of older wells will be reen-
tered by conventional drilling with long-
radius—greater than 500-ft [152-m]—or
medium-radius—200- to 500-ft [61- to
153-m]—sidetracks. However, there is a
major trend toward reentry drilling with
short-radius—40- to 100-ft [12- to 30-m]—
drilling.10 Short-radius sidetracks require
articulated drilling systems, which are highly
effective in competent formations that can
be completed without liners or other com-
pletion hardware. Short-radius drilling tech-
niques, whether by conventional means or
with coiled tubing, allow drillers to turn well
trajectories in a much shorter distance than
was previously possible. This allows kicking
off below well hardware, if required, or
drilling a curve and lateral section com-
pletely within a reservoir to avoid problems
with overlying formations.
Multilateral drilling, an increasingly popu-
lar drilling strategy in new wells, uses multi-
ple horizontal sidetracks from a primary
trunk in a parent well. This technique can
make small fields economical and reduce
the number of wells needed to drain a reser-
voir. Fewer wellheads significantly reduce
the cost of subsea completions and tie-back
operations. The multilateral geometry can be
simple opposing laterals in the same hori-
zontal formation for better penetration, or
stacked laterals to gain access, in multilay-
ered reservoirs for example, to formations at
different depths. A multilateral pattern can
be used in the same horizon to drain larger
Oilfield Review


,
,


,
,


,
,

|z



,




,




,

y
{




,



,
,


,
,


,
,
|z
|z
{



,

y



,

y
{
|,






z
y
|


z






y
,

{

,

,

,
|
{


,

|



|


y
,

{



,

{
A

angled laterals in a fan-shaped pattern.

Reentry Well Engineering

Preparing a well for reentry drilling can
B

involve a range of services from supplying

the workover rig, pulling the old comple-
tion and cement squeezing old perforations
to fishing debris from wells and cased-hole


reservoir areas through parallel laterals or as
C

milling systems.12 Complete engineering

and technical support come from Smith
specialists, but crosstraining allows Anadrill
drillers to run Smith equipment.
Sidetracking out of casing begins with a
gyro survey of the existing hole to precisely
determine location of the casing. A correla-
tion log pinpoints the target formation.
Using these data, kickoff depth and position
D

and retract for tripping. Length of the

milled section depends on several fac-
tors—nominal ID and casing coupling
diameter, bit diameter, and bent housing
E

■Section milling. A specialized bottomhole

assembly cuts through the casing and into
the cement at a chosen depth (A). Cutter
blades extend from the tool when needed

motor angle (B). After milling (C), cement

logging for corrosion and formation evalu-
ation. Depending on well design and con-
ditions, there are several possible reentry
scenarios ranging from kicking off in open
hole or cased-hole sidetracks using a whip-
stock to cut a window through the side of
the casing—window milling—to cutting a
complete section out of the casing or
liner—section milling.11
To provide efficient section milling
and window opening capabilities,
Schlumberger formed an alliance with
Smith Drilling & Completions. This partner-
ship allows the RAPID group to provide
worldwide sidetracking services, including
permanent and retrievable whipstocks, and
of the milled section are chosen. A cement-
bond log shows whether there is good
cement behind the proposed milled section.
If not, an underreamer is run between
milling and plug-setting operations to clean
up bad cement and enlarge the borehole.13
For section milling, about 60 ft [18 m] of
casing is milled if the kickoff is to be steered
magnetically out of a vertical well (above).
The milled length of casing can be reduced
if a gyro is used to steer the BHA. A compe-
tent cement plug is then set across the
milled section. To avoid magnetic interfer-
ence, the plug is dressed with a bit to the
kickoff point 20 ft [6 m] from the lower cas-
ing stub. The disadvantages of section milling
are that it requires a secure cement plug for
proper sidetracking, and there is a risk of not
is placed across the open interval and
new hole is drilled by kicking off of this
plug (D). When milling is complete, the
lower section of the original well is perma-
nently isolated from the sidetrack (E).
10. Maurer, reference 4.
11. Ehlig-Economides C: ”Improving Production Using
Re-Entry Drilling Techniques,” Petroleum Engineer
International 68, no. 9 (September 1996): 32-33.
12. Bell S: “Milling Applications Demonstrate
Versatility,” Petroleum Engineer International 66,
no. 3, Supplement (March 1994): 12-15.
13. Hill D, Askew W, Tracy P and Koval V: “A
Predictable and Efficient Short Radius Drilling
System,” paper IADC/SPE 35049, presented at the
1996 IADC/SPE Drilling Conference, New Orleans,
Louisiana, USA, March 12-15, 1996.

Autumn 1996 11
 A B C D E

■ Window milling. Operations to

cut an opening out of the casing
begin by running and orienting a
retrievable whipstock, which is
used to guide mills in the lateral
direction (A). After the whipstock
anchor is set, the attaching pin is
sheared and a starter mill initi-
ates the window cut a few inches
into the casing (B). The window
mill does the bulk of the milling,
and is run together with string, or
watermelon, mills that open up
and smooth out the new opening
through the casing wall (C). Once
milling is completed, lateral
drilling can start (D). The whip-
stock is used to guide BHAs and
completion equipment into the
lateral sidetrack (E). After the lat-
eral is completed, the whipstock
can be removed to allow access
to lower formations (F and G).

F G
12 Oilfield Review
 Vertical Profile Plan View
5000

0 ft

5100
True vertical depth, ft

Lease
5200 N
Top of sand line

Actual wellbore profile

467-ft drilling limit
5300
Bottom of sand
1000 ft

5400
-50 300 500 1000 1500 1750
Lateral section displacement, ft

■ Short-radius drilling in Texas. Frequently the challenge is to drill a short-radius reentry 1600 ft
lateral into a small target zone and remain within the hydrocarbon pay while avoiding
lease boundaries. In this well, lease restrictions and state regulations defined a narrow Actual wellbore path
116-ft [35-m] target for the first 700 ft [213 m] followed by a turn to the left (right). The
target was entered with a 77-ft [23-m] radius and a whipstock set at 5159 ft [1572 m]. 0 ft 500 ft 800 ft
Drilling continued horizontally with the same BHA. Ability to rotate the Anadrill short- E
radius drilling system resulted in excellent directional control along the horizontal lat-
eral. In addition to avoiding lease boundaries, the wellbore was maintained within the
pay zone for most of its 1600-ft [488-m] length (above).

being able to reenter the lower casing stub
after drilling the lateral. Drilling penetration
rates are often limited by the ability to clean
cuttings from the well, and once the well-
bore turns horizontal, cuttings removal is
even more difficult. Modern milling tools are
designed to create small, nonclogging cut-
tings that are easily removed from the well.
Polymer muds are more effective for milling
than clay-base muds. Oil-base muds are not
recommended for milling operations.14
An alternative to section milling is to cut a
window in the casing. This requires setting
an oriented whipstock and milling an open-
ing in the casing (previous page). After the
whipstock is set in position, the bolt con-
necting the starter mill to the whipstock is
sheared. Then rotation is started and car-
bide tips on the nose of the starter mill cut
into the casing wall. In the next stage, a
window is cut into the casing using a win-
dow-milling bit, which is forced into the
casing and the formation by the angle on
the whipstock face. The window is
enlarged or polished using the window
mill and one or more watermelon mills run
directly below the drill collars.
Section milling offers several advantages
over window milling. It can eliminate the
need for gyroscopic orientation, moves the
kickoff depth closer to the target for a given
curve radius and requires only one milling
operation. Window milling, on the other
hand, uses a whipstock that provides a
positive sidetracking mechanism, but
requires several gyro runs to orient both
the whipstock and drilling assembly.
Cutting a window also requires multiple
milling operations and a shallower kickoff
depth due to the rathole needed for the
subsequent drilling assemblies.
Whichever system is used, once entry to
the formation is gained, there are more
choices to be made. Besides standard
medium-radius drilling, several recently
■ A roadmap for the driller. Before reentry
introduced options for reentry drilling sys- drilling begins, a detailed plan is
tems can make well reentries more cost- designed. At the Schlumberger Sugar
effective. 15 Short-radius drilling, coiled Land District in Texas, Catherine Ortiz,
a Drilling Planning Center engineer,
tubing drilling, and multilaterals are each reviews a trilateral plan with Steve
candidates for thorough cost-benefit Thurston, a well planning engineer,
analysis (right). before her crew leaves for a job sched-
uled to start within 24 hours.
Short-Radius Systems
Short-radius wells are drilled to avoid tra-
versing problem formations that would oth-
erwise require a liner to isolate, or because
wells must be kicked off below hardware,
such as an external casing shoe.16 In some
14. “Tips for a Successful Re-Entry,” Petroleum Engineer
formations, the kickoff and lateral can be International 66, no. 3, Supplement (March 1994):
kept entirely in the pay zone, avoiding 8-9.
15. Ehlig-Economides et al, reference 5.
shale beds and reducing the risk of stuck
16. Leazer C and Marquez MR: “Short-Radius Drilling
pipe (above). Expands Horizontal Well Applications,” Petroleum
Engineer International 67, no. 4 (April 1995): 21,
23-24, 26.

Autumn 1996 13

Atlantic Ocean
Venezuela
Cerrejon
Pacific Ocean
Alturitas
Boca
Escalante
Rosario
■ Location of the Alturitas field in Venezuela, South America.
The curved section is drilled with a specifi-
cally designed short-radius system. The
short-radius BHA consists of a drill bit, artic-
ulated motor, flexible nonmagnetic drill col-
lar housing and MWD systems. High-
strength drillpipe is run immediately above
the BHA for easy passage through the curved
section. Drillstring in the vertical well sec-
tion usually contains standard drillpipe.
The curvature of borehole drilled by a
conventional—long- or medium-radius—
downhole motor is defined by three points
of contact between the BHA and borehole
wall—generally the drill bit, the near-bit
stabilizer and the first stabilizer above the
Power section
Lower
articulation
Adjustable standoff
1.5 ° Fixed bend
Rotating near-bit stabilizer
14
Peninsula
de Guajira Peninsula de
Paraguana
Gulf of Venezuela
Guasare
El Mamon
Las Palmas/Tiguje
A R G E N T I N A
Monte Clar
Bolivar fields
San Lorenzo
Lake Poco
Maracaibo
motor. On a short-radius system, however,
the three points of contact have to be posi-
tioned below the motor knuckle joint.
Articulations are needed to allow the
motor to pass around sharp bends and
have no effect on angle build rate. They
also allow rotary drilling. Both roller cone
or polycrystalline diamond (PDC) bits can
be used at the operator’s discretion to han-
dle different formation characteristics.
Upper
articulation
Stinger
Rotor/stator
■ Short-radius drilling system. The
Anadrill PowerPak XF short-radius drilling
system consists of an articulated BHA
with the Slim 1 slim and retrievable MWD
system. Build angle is controlled by a sur-
face-adjustable standoff at the top of the
short rigid motor section behind the bit.
The Anadrill short-radius drilling system
uses a 4-ft [1.2-m] rigid motor section with
a surface-adjustable standoff as the third
point of contact to control radius of curva-
ture (below). This system maintains contin-
uous contact with the borehole, allowing
predictable build rates and easy control of
the horizontal section. This also avoids the
need to prepare different motors for each
section of the well.17 Directional control is
monitored with a Slim 1 retrievable MWD
system that includes a gamma ray measure-
La Vela ment for geological correlations. This MWD
de Coro tool was designed to communicate with the
surface through mud-pulse telemetry during
angle-build drilling to a 40-ft minimum
radius of curvature. The directional sensor
has been placed in the lowest position,
directly above the motor power section, for
enhanced trajectory control.18
One recent example of production
enhancement through short-radius drilling
took place in OXY’s Alturitas field, 30 miles
[48 km] west of Lake Maracaibo, Venezuela
(left). The target Marcelina reservoir lies
below a coal stringer that is difficult to drill
at any inclination other than vertical, which
made horizontal drilling uneconomical
Fishing head
Mud pulser
Battery module
Gamma-ray detector
Direction and
inclination systems
17. Hill et al, reference 13.
18. Hutchinson M: “Innovative Short Radius Drilling
System Demonstrates Greater Flexibility and
Directional Control,” Petroleum Engineer
International 68, no. 10, Supplement (October
1996): 14-15.
Oilfield Review
until short-radius drilling technology Vertical Profile
10900
became available.
Alturitas 22 was producing 300 BOPD 10920
[47 m3/d], so the objective was to increase

True vertical depth, ft

production by drilling a horizontal lateral 10940

using the Anadrill short-radius drilling sys- 10960

1
2
tem. The plan was to set a retrievable whip-
3
stock in the 95⁄8-in. casing, mill a window, 10980
4
drill the curve and lateral, and then place 11000 5

the well on production. The retrievable

whipstock allows the original completion 11020

to be reentered, if necessary, or more later-

11040
als to be added at a later date. 0 200 400 600 800 1000 1200 1400 1600 1800 2000
An A-Z PackStock was set at 10,895 ft Lateral section displacement, ft
[3321 m] and a 20-ft [6-m] window, ■ Alturitas well profile. The drilling plan called for a short-radius lateral to the bottom of
including about 9 ft [3 m] of formation, the pay, then the lateral was allowed to angle up—cutting through each sand layer
was milled using a gel mud to improve before turning back down, again cutting through each of the layers. Once the bottom
sand was reached, drilling stopped—1933 ft [589 m] from the kickoff point. This lateral
removal of cuttings. Inclination at 10,915 ft produced a sevenfold improvement in production rate over the original vertical well.
[3327 m] was 3°. The mud system was
changed to oil-base and the BHA was VIPER Coiled Tubing Drilling System
replaced with an Anadrill short-radius
drilling system. In another 84 ft [26 m] of
drilling, 90° inclination was achieved,
placing the lateral well within the target
depth of 10,988 to 11,003 ft [3349 to
3354 m] (above right).
Drilling continued horizontally through
the reservoir, which consisted of a series of
sandstone layers. The horizontal lateral was
VIPER BHA with PowerPak motor
allowed to angle upward from the lower-
most layer, crossing all the sandstone mem-
bers for about half the lateral length. The A B C D
wellbore was then steered downward
again, staying within the pay. Drilling was
stopped after the well path had descended A B
back through the entire sand sequence—a Drilling head module Logging tool

horizontal distance of 1933 ft [589 m] Coiled tubing Check Pressure transducers, Gamma-ray sensors
from the kickoff point. and wireline valves and electronics and electronics
Success of this project can be measured
by current production and cost. The lateral
Wet Electrical Telemetry Directional and inclination
was left as an openhole completion flow- connect disconnect and power sensors and electronics
ing 2000 BOPD [318 m3/d]—nearly a sev-
enfold rate increase over the production of
a typical vertical well in this field. The cost
of this workover was $3.2 million, com- C D
pared to an original well cost of $2.4 mil- Orienting tool Steerable motor
lion, a nominal increase in cost relative to
Motor, gear train
the improvement in production. OXY plans and bearing section Articulation
to drill more wells of this type.

Coiled Tubing Systems Motor/power Nonmagnetic Surface-adjustable

One of the newer technologies developed
for the reentry market is coiled tubing
drilling (right). This approach is attractive
electronics power section bent housing
■ Coiled tubing drilling system. The VIPER system consists of a wireline-powered BHA
that includes an instrument package for directional control as well as gamma ray,
temperature and pressure measurements, a PowerPak downhole motor, an orienting
tool that can rotate continuously and a circulating sub. Data and downhole commands
are transmitted via a cable that is pumped down inside the coiled tubing.

Autumn 1996 15
 when drilling rig mobilization costs are
prohibitive. The most successful applica-
tion of CTD is through-tubing reentry com-
bined with underbalanced drilling. Coiled

Anadrill
tubing allows more precise control of low
downhole hydrostatic pressure. Not having
to pull production tubing and kill the well
Shallow or makes this technology attractive.19
depleted reservoirs
New coiled tubing directional BHAs pro-
vide improved directional control and effi-
ciency. One such system, called VIPER
technology, is a wireline-powered BHA that
includes a downhole orienting tool for
directional control and MWD system for
Layered
reservoirs directional measurements. Both are oper-
ated from surface via wireline-supplied
power and signals. Without the wireline,
signal transmission is impossible in under-
balanced drilling environments where
foamed, aerated or nitrogenated mud is
used. The wireline system also increases
Fractured the data transmission rate by several orders
reservoirs of magnitude over mud-pulse systems,
allowing surface control of sensors.
Another VIPER system benefit is improved
coiled tubing drilling efficiency. The elec-
■ Multilateral drilling for improving productivity. In depleted zones, a network of later- tric motor in the orienting tool offers higher
als increases the length of wellbore in contact with the reservoir (top lateral), which torque, as well as accurate and uninter-
also reduces adverse pressure drawdown effects. Several isolated layers can also be
tapped from the same wellbore (middle laterals). In a fractured reservoir, dual laterals rupted directional control. Continuous
intersect twice as many fractures (bottom laterals). slow rotation of the motor drills a smoother
borehole profile, allowing longer-reach
drilling by reducing friction and dogleg
curves. The ability to continuously monitor
5700 downhole pressure during drilling, tripping
and circulating ensures accurate mainte-
True vertical depth, ft

KOP
nance of underbalanced conditions.
6100
10°/100 ft build
Top of Multilateral Systems
6500 Austin Chalk Multilateral drilling places more than one
Bentonite
Eagleford drainhole into one or more hydrocarbon
intervals (above left). Improved recovery and
6900 False Buda
Buda reduced well construction costs, through
Georgetown reuse of the parent borehole and surface
7300 equipment, make multilaterals an attractive
0 800 1600 2400 3200 4000 4800
option. The cost of preparing an existing well
Lateral section displacement, ft is the same regardless of how many laterals
are drilled. Multilaterals, therefore, cost less
■ Typical Austin Chalk quadrilateral openhole (barefoot) completion drilled for Union per lateral than single lateral wells.20 Slot
Pacific Resources. management is improved, and the expense of
drilling additional parent wellbores is elimi-
nated.21 Additional reservoirs can be tapped
by drainholes that could not have been

16 Oilfield Review
Open Hole
Sidetracking and drilling
from Anadrill
Drill-in fluids from Dowell
Cemented Liner
Packers, liners,
completion fluids
from Dowell
Tubing-conveyed
perforating from
Wireline & Testing
Prepacked Screen or Slotted Liner
External casing packers,
completion fluids,
Maraseal and screen
from Dowell
Packers from
Wireline & Testing
Autumn 1996
drilled previously, and production rates per
wellhead can be much greater.
The most basic multilateral application is
openhole, or barefoot, completions in
competent carbonates like the south Texas
Austin Chalk (previous page, bottom).
Anadrill has drilled more than 50 such
wells to date. Lateral drainholes intersect
natural fractures, increasing production
from a single well. Inability to perform
workovers, however, is a drawback. These
are essentially throw-away wells with com-
mingled flow and no chance of turning off
water production.
Completing Multilateral Wells
In general, three completion options are
available for reentry multilateral wells
(left). Wells can be left open as in the
Austin Chalk, cased and perforated, or
completed with some variation of a pro-
duction screen.
Soft formations that produce from matrix
permeability require normal completions,
such as slotted liners and gravel packs in
each branch, connected mechanically to the
main wellbore trunk. This connection has to
be pressure-tight to maintain zone isolation.
Furthermore, when different reservoir types
are produced through the same multilateral
well, selective accessibility to each lateral
may be necessary throughout the life of the
well. Complete control of each drainhole is
essential to avoid jeopardizing production
of the entire multilateral system when one
drain is depleted or produces excessive
water or gas.
Today, most lateral connections are built
downhole and rely on good cement to pro-
vide a seal and isolation. Schlumberger is
developing hardware systems that allow
separate completions for each branch of
the well. These systems include surface-
built junctions that can extend into any
portion of the well—vertical or horizon-
tal—and each branch can be easily and
selectively accessed. With such systems,
there are no reductions in internal diame-
■ Reentry completion options. Reentry lat-
erals may be left as openhole completions
in competent formations like the south
Texas Austin Chalk (top). Alternatively,
laterals can be cased, cemented and per-
forated (middle). More complex comple-
tions, such as gravel-pack completions,
are also available (bottom).
ter in the trunk, which allows lateral
branches to be drilled in any sequence,
and allows standard tubing and packer
completion strings to be run. An outlet port
will support a liner hanger and packer,
making it possible to run any type of stan-
dard completion in the lateral, and
enabling good sand control practices, iso-
lation and flow control.
The Outlook
An explosion of new technologies coupled
with a collapsing of conventional bound-
aries between different oilfield services has
given operating companies the widest pos-
sible range of solutions to increase recovery
in aging fields. A comprehensive toolbox
for production optimization through reentry
drilling and completion can be provided by
groups like the RAPID team. The potential
value of these services is dramatic.
Thousands of wells have been drilled and
completed conventionally. Using reentry
techniques to increase production from just
a fraction of these wells will be equivalent
to discovering several giant new fields.
—RCH, JMK, AM
19. For more information on coiled tubing drilling see:
Bigio D, Rike A, Christian A, Collins J, Hardman D,
Doremus D, Tracy P, Glass G, Joergensen NB, and
Stephens D: “Coiled Tubing Takes Center Stage,”
Oilfield Review 6, no. 4 (October 1994): 9-23.
20. For some examples showing some cost details of
reentry multilateral drilling: Hall D: “Multi-Lateral
Horizontal Wells Optimizing a 5-Spot Waterflood,”
presented at the SPE Permian Basin Oil & Gas
Recovery Conference, Midland, Texas, USA, March
27-29, 1996.
21. On offshore wells, a slot is a space that accommo-
dates one wellhead in a template secured to the
ocean floor. A template has a limited number of
slots, which cannot be changed once the template
is installed. If one well waters out or is dry, that slot
is already used up. Reentry drilling, however, gives
new life to the slot because it allows bypassing
unproductive zones with a new drainhole.
17
 Quo Vadis, Extreme Overbalance?

Perforating and surging techniques using very high wellbore pressures promise dramatic, cost-effective

improvement in initial well productivity—under the right conditions. But what are those conditions, and how

do the procedures work? Leading investigators look at the basics of these new completion methods, and

examine lessons learned to date from Prudhoe Bay, Alaska and the North American midcontinent.

Larry Behrmann In the vast majority of wells today, the
Klaus Huber moment of truth—do we have a producer,
Bryan McDonald or a hole in the ground?—is revealed
Rosharon, Texas, USA through underbalance perforating. When
perforating guns fire, pressure in the well-
Benoît Couët bore is below that of the reservoir, creating a
Ridgefield, Connecticut, USA pressure differential that helps clean the per-
foration tunnels. Formation fluids rush into
John Dees the tunnels and flush out metallic charge
Consultant debris, surrounding crushed rock, and sands
Dallas, Texas or clays that were driven into the tunnels. If
the drawdown is large enough, inflow can
Ron Folse sweep away enough debris to open the most
Marathon Oil Company conductive natural path between the forma-
Lafayette, Louisiana, USA tion and wellbore.
Two-way communication along this path
Pat Handren is essential for optimal well completion and
Oryx Energy productivity. When a well goes straight into
Dallas, Texas production, clogged perforations will limit
inflow of hydrocarbons. If intervention is
Joe Schmidt planned, perforations need to be clear to
ARCO Alaska, Inc. accept treatment fluids carrying proppant
Anchorage, Alaska, USA for fracturing, gravel for sand control, or
acid. Hydraulic fracturing and prepacking
Phil Snider perforations ahead of gravel packing benefit
Marathon Oil Company from removal of crushed sand that can
Houston, Texas reduce injectivity and elevate fracture initia-
tion pressures or lead to early screenout of
proppant during fracture stimulations.1
For help in preparation of this article, thanks to Bob
Cooper, Dowell, Houston, Texas, USA; Ray Dickes,
Underbalance perforating works across a
Glen Edwards, Joe Hromas, Bjorn Langseth, Al Salsman broad range of rock properties and reservoir
and Mike Selewach, Schlumberger Perforating & Testing conditions. Its applicability decreases, how-
Center, Rosharon, Texas; Roger Card, Jack Elbel, Mark
Mack and Ken Nolte, Dowell, Tulsa, Oklahoma, USA; ever, with a decline in reservoir pressure,
Roger Keese, Dowell, Anchorage, Alaska, USA; George permeability or rock strength. The trick is to
King, Amoco Exploration and Production Technology
Group, Tulsa, Oklahoma; and Terry Green, Dowell,
Sugar Land, Texas.
NODAL and X-Tools are marks of Schlumberger.
achieve enough underbalance to generate
sufficient flow rate for cleaning, but not too
much to collapse the perforations and drive
sand into the well. Theoretical and applied
studies have focused on defining the opti-
mal underbalance for ranges of reservoir
pressure, permeability and rock strength.2
With a good theoretical foundation and a
record of favorable results, underbalance
perforating reigned as the unchallenged
champion until a few years ago, when a
handful of investigators turned underbal-
ance on its head (next page ). Building on
experimental work by the US Department of
Energy and others, Oryx Energy and ARCO
independently developed new completion
techniques utilizing extreme overbal-
ance—perforating with wellbore pressure
significantly above the level required to frac-
ture the formation. The patented Oryx and
ARCO methods differ in their approach, but
each involves a process that may generically
be called extreme overbalance perforating
(EOP) and a related method of forcing an
extreme overbalance pressure into existing
perforations, called extreme overbalance
surging (EOB surging).3
Perforating underbalance or with extreme
overbalance are in many ways opposites,
but they are not mirror images of each other.
In underbalance perforating, the goal is to
create a channel and clean the channel with

18 Oilfield Review
 1865 1910 1948 1975 1980 1993

10 to 30 ft

Arrows illustrate pressure differentials

■The first 130 years of perforating. 1865: Tin torpedoes filled with gunpowder, and later with nitroglycerin, are lowered to the depth of
interest and detonated. 1910: The single-knife casing “ripper” involves a mechanical blade that rotates to puncture a hole in the casing.
1948: The first shaped charges, developed by Welex Jet Perforating Company, are applied to oil wells, generally with a slight overbal-
ance for well control. 1970s: Underbalance, under investigation for 20 years, is tied with perforating by Roy Vann. Continued work by
others through the 1980s accelerates its popularity. 1980s: Propellant fracturing produced fractures from the burst of pressure developed
by rapid burning of propellant. Although still under investigation, the method encounters problems operationally and in reproducibility.
1993: Extreme overbalance perforating, pioneered by Oryx Energy Company, succeeds in commercial wells.

flow from the formation, then stimulate or
gravel pack later as necessary. In extreme
overbalance methods, the idea is to simulta-
neously create and stimulate the channel,
which develops into a small biwing fracture
that obviates the need for cleaning the per-
foration tunnel. Some operators have also
proposed extreme overbalance methods that
simultaneously place resin for sand control
or acid for etching fracture faces.4
Since extreme overbalance methods
became commercial in 1990, their applica-
tion has taken a roller coaster ride. Follow-
ing an initial wave of interest, only a small
but devout core of proponents continues to
carry the torch—last year about half the
extreme overbalance jobs in North America
were performed by only five operating
companies. Nevertheless, limited but per-
sistent curiosity from the industry refuses to
die. In 1994, seventeen operating and ser-
vice companies jointly sponsored experi-
ments on large blocks of sandstone to
investigate EOP fracture mechanics and
ways to optimize pressure requirements
and perforation design.
Field experience has also broadened, with
about 900 EOP jobs performed to date,
mostly in North America. Marathon per-
formed extreme overbalance procedures in
20% of its wells in 1995 and expects that
number to reach 35% in 1997. As more
data accumulate, the case for extreme over-
balance resurfaces, each time with a bit
more ammunition. The technique has
clearly established a niche, yet its applica-
bility remains incompletely defined. Where
are extreme overbalance methods today,
and what promise do they hold?

1. Screenout is the point at which no more proppant can
be pumped into a hydraulic fracture system without
an increase in pump pressure. Early, or premature,
screenout is caused by an impermeable bridging of
material across the fluid pathway that prevents further
extension and propping of the fracture.
2. King GE, Anderson A and Bingham M: “A Field Study
of Underbalance Pressures Necessary to Obtain Clean
Perforations Using Tubing-Conveyed Perforating,”
paper SPE 14321, presented at the 60th SPE Annual
Technical Conference and Exhibition, Las Vegas,
Nevada, USA, September 22-25, 1985.
Tariq SM: “New, Generalized Criteria for Determining
the Level of Underbalance for Obtaining Clean Perfo-
rations,” paper SPE 20636, presented at the 65th SPE
Annual Technical Conference and Exhibition, New
Orleans, Louisiana, USA, September 23-26, 1990.
Behrmann LA: “Underbalance Criteria for Minimum
Perforation Damage,” paper SPE 30081, presented at
the SPE European Formation Damage Control Confer-
ence, The Hague, The Netherlands, May 15-16, 1995.
3. Three early works:
Bundy TE and Elmer MJ: “Perforating a High-Pressure
Gas Well Overbalanced in Mud: Is It Really That
Bad?” paper SPE 16894, presented at the 62nd SPE
Annual Technical Conference and Exhibition, Dallas,
Texas, USA, September 27-30, 1987.
Cipolla CL, Branagan PT and Lee SJ: “Fracture Design
Considerations in Naturally Fractured Reservoirs,”
paper SPE 17607, presented at the 1988 SPE Interna-
tional Meeting on Petroleum Engineering, Tianjin,
China, November 1-4, 1988.
Northrup DA and Frohne K-H: “The Multiwell Experi-
ment—A Field Laboratory in Tight Gas Sandstone
Reservoirs,” Journal of Petroleum Technology 42,
no. 6 (June 1990): 772-779.
The Oryx method:
Handren PJ, Jupp TB and Dees JM: “Overbalance Per-
forating and Stimulation Methods for Wells,” paper
SPE 26515, presented at the 68th SPE Annual Techni-
cal Conference and Exhibition, Houston, Texas, USA,
October 3-6, 1993.
The ARCO method:
Petitjean L, Couët B, Abel JC, Schmidt JH and Fergu-
son KR: “Well Productivity Improvement Using
Extreme Overbalance Perforating and Surging—Case
History,” paper SPE 30527, presented at the 70th SPE
Annual Technical Conference and Exhibition, Dallas,
Texas, USA, October 22-25, 1995.
For a more recent version of the same work:
Couët B, Petitjean L, Abel JC, Schmidt JH and Fergu-
son KR: “Well-Productivity Improvement by Use of
Rapid Overbalance Perforation Extension: Case His-
tory,” Journal of Petroleum Technology 48, no. 2
(February 1996): 154-159.
4. Handren PJ and Dees JM: “A New Method of Over-
balance Perforating and Surging of Resin for Sand
Control,” paper SPE 26545, presented at the 68th SPE
Annual Technical Conference and Exhibition, Hous-
ton, Texas, USA, October 3-6, 1993.

Autumn 1996 19
 6500

6000 Well A Well B

Wellhead pressure, psi

5500

Well A
5000

4500 2600 ft gas

4000 Actual
Modeled 4984 ft liquid
Well B
3500

3000
-40 -20 0 20 40
Time, sec
6753 ft liquid
The EOP Why and How ■The 30-second moment of truth. Modeled
John Dees and Pat Handren, extreme over- (solid line) and actual data (circles) for pres-
sure drop in two ARCO wells (above). The
balance pioneers at the Sun Company (now 5472 ft gas
slower decay in Well A is due to greater
Oryx Energy), began investigating overbal- friction produced by a large volume of
ance methods in the late 1980s when fluid (right). The goal is to minimize friction
underbalance failed to give good results in losses, increasing propagation speed, min-
West Texas fields. The Oryx team found imizing leakoff into the formation and
maximizing fracture length. (From Coüet Surge disk
what others had also observed for some et al, reference 3.)
time: Correctly applied underbalance perfo-
rating can be compromised by fairly com-
mon reservoir and operational conditions. If
reservoir pressure is low or depleted, the
pressure differential may be insufficient to
clean perforations. Likewise, if permeability ■Extreme overbal-
is low—probably less than 10 millidarcies ance perforating.
Most procedures
(md), but the value depends on reservoir follow a variant of 4000 to 8000 psi
pressure and oil viscosity—formation fluid this basic (0.8 to 1.4 psi /ft)
may not flow vigorously enough for clean- approach.
N2 Pump
ing. And if rock strength is low, underbal-
ance pressure differential large enough for
effective cleaning may collapse the forma-
tion and necessitate further intervention to
save the well.
Underbalance perforating can also be hin-
dered by more complex problems. Improper
killing of a well, for example, can replug Casing
perforations with filter cake that may not be Pressurized N2
dislodged during production.5 Sometimes, Hydrostatic or
despite good reservoir pressure and perme- atmospheric
Cement
ability, the damaged zone reaches deep Packer 30 to
enough to limit the effectiveness of under- 90 ft above valve
balance. Also, when permeability varies Fracturing fluid
8000 to 12,000 psi
dramatically—such as a thin, 1-darcy layer Isolation valve:
sandwiched between two thick 10-md lay- either retrievable
SXPV or permanent 50 to 100 ft,
ers—the thicker sections will dominate the SXAR equipment less often to 300 ft
flow properties and can reduce the effec-
tiveness of underbalance.6
Extreme overbalance perforating can
sidestep these problems. In EOP comple- Pay Invaded
tions, tubing pressure is increased before the zone zone
guns are fired and then released into the
wellbore with gun detonation. At this point,

20 Oilfield Review
because wellbore pressure exceeds rock
yield strength, perforating initiates one or
more small fractures. These fractures do not
develop the length or height of conventional
hydraulic fractures, but the event lasts long
enough to push the fractures beyond the
zone damaged by invasion and past the tip
of the perforation ( previous page, top ).
While EOP fractures are shorter in length
and height, they may develop greater width
and so possibly have a higher conductivity
per foot than hydraulic fractures.
Most EOP jobs follow the same basic pro-
cedure (previous page, bottom ). Perforating
guns are lowered to the depth of interest,
then spotted to the top of the guns is a small
amount of liquid selected for the well con-
ditions—brine, lease crude, fracturing fluid,
acid or liquid with proppant. All or most of
the wellbore above the liquid is filled with
compressible gas, usually nitrogen, less
often carbon dioxide or air. The gas col-
umn is then pressured up, like a tightly
squeezed coil spring. Sometimes liquid is
also spotted above the gas to further com-
Time 1
4000 to 8000 psi
(0.8 to 1.4 psi /ft)
N2 Pump
Pressurized N2
Rupture disk
Fracturing fluid
Pay zone
Invaded zone
■Extreme overbalance surging. Surging is performed on wells with existing perforations.
It may follow immediately after perforating or a few hours to a few days later. In general,
the sooner after perforating, the more effective the surge. Comparison of time 1 (left) and
time 2 (right) shows the penetration of fluid into the rock and propagation of the fracture
over time.
Autumn 1996
press it. Rarely does the liquid fall through
the gas because compressed gas, typically
at about 4000 psi [27,500 kPa], develops a
density of 1 to 3 lbm/gal [0.12 to 0.36
g/cm3] and a high surface tension. This cre-
ates an interface which, in the small diame-
ter of tubing, prevents liquid from displacing
gas. Because the surface pressure of gas can
reach 10,000 psi [69 MPa] or more, tubing-
conveyed perforating (TCP) guns are usually
preferred over wireline-conveyed guns
because they are operationally easier to
handle at high pressures.
With detonation of the guns, the liquid is
driven at very high flow rates by the rapidly
expanding gas and rushes into the perfora-
tions. Because the liquid is nearly incom-
pressible, it acts as a wedge that initiates
fractures, extending the effective wellbore
radius. Erosion from the liquid and any
entrained proppant flowing at more than
100 bbl/min [16 m3/min] may scour the for-
mation, creating stable flow channels. In
many EOP jobs, the event is timed to stop
just when the gas reaches the perforations,
Time 2
N2 Pump
Tip of
fracture
since the gas would quickly leak off into the
formation. Some operators, who have wells
with large tubular volumes, continue apply-
ing pressure as the gas enters the perfora-
tions. The gas also acts as an abrasive that
scours the perforation. In either case—stop-
ping as the gas hits the perforations or con-
tinuing—the higher the pressure and larger
the gas volume (a larger “spring”), the
greater the fracturing power.
Pressure generated at the perforations dur-
ing EOP or EOB surging must be high
enough to overcome two obstacles: it must
exceed the minimum in-situ rock stress,
and it must fracture through any imperme-
able debris barrier remaining in the perfora-
tion. The debris barrier often defeats the
conventional process of perforation break-
down and cleanup.7 Modeling shows that
to overwhelm the barrier, the extreme over-
balance pressure gradient usually needs to
reach at least 1.4 psi per foot [31.6 kPa/m]
of well depth.8 This gradient produces a
fracture radius that is on the order of 10 to
20 ft [3 to 6 m] although it may extend up
to 30 ft [9 m].
In the high-energy context of extreme
overbalance, flow restriction due to perfora-
tion damage has such a minor effect on per-
foration conductivity as to become almost
irrelevant. Charge debris has no time to
harden and is thought to be pulverized and
blown far back into the created cracks, like
a mashed up cork pushed into a wine bot-
tle. The low permeability of the shattered
zone is more than compensated for by the
high permeability of the fractures. In addi-
tion, gas jetting into the tunnel at nearly the
speed of sound may erode and scour walls
of the tunnels and fractures.
An extension of this method involves
pumping additional fluid at a high rate
immediately following EOP or EOB surging,
with or without proppant, to drive the frac-
tures farther (left ). Pump rates have to be
5. Mason JN, Behrmann LA, Dees JM and Kessler N:
“Block Tests Model of the Near-Wellbore in a Perfo-
rated Sandstone,” paper SPE 28554, presented at the
69th SPE Annual Technical Conference and Exhibi-
tion, New Orleans, Louisiana, USA, September 25-28,
1994.
6. Hsai T-Y and Behrmann LA: “Perforating Skin as a
Function of Rock Permeability and Underbalance,”
paper SPE 22810, presented at the 66th SPE Annual
Technical Conference and Exhibition, Dallas, Texas,
USA, October 3-6, 1991.
7. Behrmann LA and McDonald B: “Underbalance or
Extreme Overbalance,” paper SPE 31083, presented at
the SPE International Symposium on Formation
Damage Control, Lafayette, Louisiana, USA, February
14-15, 1996.
8. Prudhoe Bay seems to be the exception. ARCO
reports good results there with only 1.1 psi/ft
[24.2 kPa/m].
21
high enough to keep the fluid above the for-
mation fracture pressure. The injection rate
needed for success depends on formation
characteristics and in some cases, up to
15,000 ft3/min [420 m3/min] has been used.
ARCO also developed a gas-surging tech-
nique that enhances hydraulic fracturing in
wells previously perforated. (see “Elements
of EOP Design: Operations,” page 31.)9
What can be expected from an extreme
overbalance operation? Recoverable
reserves may be increased, and under favor-
able conditions production rates can
increase dramatically, due to reduction in
near-wellbore pressure loss and in reservoir
skin—70% of EOP wells show a negative
skin (see “EOP for Skin Reduction,”
below ).10 In one Oryx field, where conven-
tionally completed wells—fracture treated
with 20,000 gallons [3180 m3] of gelled
diesel and 20,000 lbm [9070 kg] of 20/40
sand—produced 500 Mcf/D, EOP wells pro-
EOP for Skin Reduction
In 1994, Marathon stood at a completion cross-
roads in an eastern New Mexico gas field. In the
30-year-old field, Marathon and other operators
produced from 35 wells, each making 4 to 6
MMcf/D. In a typical completion, skin averaged
around 50 but reached as high as 150. The reser-
voir, an upper Pennsylvanian carbonate, averaged
39 md with a reservoir pressure of about 1000 psi
[6890 kPa]. Could skin be reduced?
Studying the success of nearby operators who
use extreme overbalance methods, Marathon
decided to try extreme overbalance surging to
reduce skin. The first candidate well was in the
North Indian basin section, with a perforated
interval of 171 ft [52 m].
A fairly conventional completion design was
used for extreme overbalance surging with acid.
A 60° shot phasing was chosen to evenly distribute
20% HCl acid and encourage development of a
22
duced initially at twice that rate and
depleted in 3 years instead of 7 to 10 years.
EOP also facilitates lower treating pressure
due to creation of a more conductive flow
path. In addition, EOP and EOB surging
allow for placement of a higher percentage
of proppant when followed by a conven-
tional frac job. ARCO, for example, reports
placement of 95% of sand in extended-
reach wells, probably due to higher conduc-
tivity of flow paths into the formation.11 It
formerly placed only 35% of sand.
The economics of extreme overbalance is
not clear-cut, and has contributed to skepti-
cism. One-to-one comparison with conven-
tional completions is sometimes difficult.
Should EOP be compared to underbalance
perforating alone, or to perforating and
hydraulic fracturing? While the latter may
seem logical, in practice EOP does not fully
replace hydraulic fracturing. Marathon, for
example, will later frac more than a third of
biwing fracture. Previous jobs were at two shots
per foot (spf), but this low shot density was
thought to contribute to high skin. Inflow perfor-
mance and NODAL analysis indicated that 4 spf
would probably result in more effective acid place-
ment. Deep penetrators were used to reduce flow
restriction thought to be associated with big-hole
charges. And the pressure gradient was designed
to be 1.4 psi/ft—low by today’s standards, in
which some jobs are designed at 2 psi/ft
[45 kPa/ft].
About 500 ft [152 m] of fluid was spotted at the
bottom of the string, and the tubing was charged
with N2. Fluid pumped after the gas achieved a
flow rate of 200 bbl/min [31.8 m3/min] after the
guns had fired. When the well came in, skin was
computed at 5, a dramatic 10-fold reduction.
Well production was higher than expected—over 5
MMcf/D, compared to 4 MMcf/D for conventionally
completed wells. Although completion costs were
about 8% higher, the initial rate gain quickly paid
back the higher cost. Today, the well continues
producing at a higher rate than conventionally
completed wells of similar age.
Since this job in 1994, Marathon has moved
toward a new completion strategy. During the two
years of production, decline in reservoir pressure
its EOP wells, and about 80% of wells that
were surged after EOP. Marathon finds that
EOP operations average one day longer, but
produce first oil one to three days earlier.
Some operators use EOP as a cost-effective
way to identify hydraulic fracturing candi-
dates, as a means to minimize near-well-
bore tortuosity and thereby reduce hydraulic
fracturing costs (less fluid pumped and
lower surface pressures), or as a low-cost
means to establish a high flow rate early.
The biggest benefit of EOP, however, is the
ability to place more sand and prevent a
near-wellbore screenout during a subse-
quent frac job.
Costs for EOP can vary widely, and depend
mostly on availability of compressed nitro-
gen. With easily accessible nitrogen, tubu-
lars fit for EOP pressures, and a completion
that would normally include TCP guns, EOP
costs slightly more than a small hydraulic
fracture. If nitrogen is not readily available,
from 1000 to 800 psi [5512 kPa] means that skin
plays an increasingly significant role in well pro-
ductivity. Because every well intervention risks
an increase in skin, Marathon is attempting the
least intrusive completion strategy: air drilling in
slight underbalance and completing open hole
(barefoot), without perforations. The carbonate is
fractured enough to produce without perforations,
yet competent enough to withstand production.
Other operators in the region report wells produc-
ing over 5 MMcf/D with this approach.
If reservoir pressures were 2000 to 3000 psi
[13.8 MPa to 21 MPa], according to Ron Folse,
a Marathon engineer who works in the field,
then extreme overbalance might be the method of
choice. But with the lower reservoir pressures,
Marathon is able to drill underbalance with air,
and with the barefoot completion achieve the ben-
efit of minimizing completion skin. Proof of this
new method is pending pressure transient analy-
sis, which is expected to indicate a lower skin and
less fluid invasion than with EOP.
Oilfield Review
EOP can cost more than twice that of a con-
ventional completion. While EOP wells pay
out faster, with higher initial production, it

ss l
streizonta
hor imum
Preferred
remains unclear under what conditions the fracture plane
long-term payout from EOP is comparable to

Min
that of hydraulic fracturing.
Economic uncertainty aside, the lack of Maximum Maximum
early enthusiasm for these methods also horizontal stress horizontal stress
resulted from unclear explanations for their
success and benefits, and inconsistent Created
fracture
results probably related to misapplication.
Today, with the analysis of more data, the
proper role of EOP is coming into focus.
Central to this understanding is an apprecia-
tion of EOP mechanics.

Are EOP Fractures Different?

The most effective fracture, regardless of the
generating mechanism, paves an autobahn
between the reservoir and wellbore: a sin-
gle, straight parting of significant width with
few smaller, competing fractures. This ideal
■The ideal relationship between perforations and the in-situ stress field.
Fractures aligned with the direction of maximum horizontal stress will open
in the plane of least resistance, against the minimum horizontal stress.
The closer the fracture and maximum stress planes align, the less tortuosity
fractures develop, resulting in less pressure drop. Fractures generally initiate
at the sandface, the base of the perforation where it meets the formation.

■Timing of down-
hole events. A Perforation
Detonation
Whereas hydraulic
Overbalance Created
fracturing can last
Propellant fracture
tens of minutes to
Hydraulic hours, extreme
Pressure

overbalance perfo-
rating is finished in
20 or 30 seconds.

1000 1000 1000

µsec msec
Logarithmic time
would be achieved with perforations 180°
apart and aligned with the maximum in-situ
stress (top ). But this direction, called the
preferred fracture plane (PFP), is often
unknown, and many operators cannot yet
routinely control gun orientation. Still, it is
possible to approach this ideal with both
hydraulic fracturing and EOP. They work by
different mechanisms, however, and pro-
duce fractures with different characteristics.
In hydraulic fracturing, pressure at the rock
face rises gradually from a slight overbal-
ance to the point of failure, after which the
9. Petitjean et al, reference 3.
10. Skin is a dimensionless value that refers to the pres-
sure drop near the wellbore during production or
intervention. The pressure drop indicates a resis-
tance to flow attributable to reduction in permeabil-
ity near the wellbore occurring during drilling, com-
pletion or operations. A positive skin value denotes
formation damage; a negative skin indicates an
increased ease with which fluid can flow between
the wellbore and formation.
sec
fracture propagates as long as the treatment
continues—and it doesn’t screen out or
dehydrate. This gradual buildup of pressure
is equivalent to opening a door by pushing
it slowly (above ). Two events probably take
place. First, multiple fractures may develop
from the perforation base or tip, depending
on proximity of perforations to the PFP and
on shot phasing.12 If perforations are more
than 30° from the PFP, fractures may take
circuitous routes around the cement-forma-
tion interface before turning to align in the
PFP (right ).
11. Couët et al, reference 3.
12. Phasing is the angle between shots. Phasing of 180°
means two shots in opposite directions, whereas a
120° phasing distributes three shots around the well-
bore circumference, one every 120° degrees.
30°
Preferred fracture
plane
■A long way to Tipperary. Hydraulic frac-
tures may have to march around the bore-
hole circumference before extending into
the formation, depending on the distance
between the perforation and the preferred
fracture plane (PFP). Here are two scenar-
ios for a 120° phased gun with conven-
tional hydraulic fracturing. If the perfora-
tion lies in the PFP (A), one wing of the
fracture will initiate from the perforation,
and the other winds around the borehole
from the perforation base until it turns into
the PFP. If the perforation is 30° or more
from the PFP (B), multiple parallel fractures
may develop from the perforation.
(Adapted from Behrmann and McDonald,
reference 7.)

Autumn 1996 23

A A
B extreme overbalance pressure overwhelms
30°
■The long and short of EOP. If a perforation
is more than 45° from the PFP, EOP frac-
tures tend to propagate immediately from
a point on the wellbore wall aligned with
the PFP, rather than wind around the well-
bore from another initiation point (A).
The channel shown between the perfora-
tion and fracture develops after the frac-
ture initiates and is eroded during the EOP
process. Where perforations are within 45°
of the PFP (B), fractures initiate from the
perforation. Only small fractures initiate
from perforations more than 45° from the
PFP. These fractures do not grow. Note that
fractures initially ignore the in-situ stress
field and probably extend straight from
the wellbore. Within a few wellbore diame-
ters they turn into the PFP. The absence
of multiple parallel fractures is thought to
be related to the sudden pressure load
on the cement-by-formation microannulus,
sealing off a path for the development of
secondary fractures.
The second event is development of a sin-
gle, dominant biwing fracture initiating at
the borehole wall. Within about two well-
bore diameters, this major fracture curves
until it aligns with the PFP, and becomes
responsible for most communication
between the wellbore and formation. Com-
peting with it, however, may be the many
small, curving fractures near the wellbore.
These fractures act as chokes that increase
the near-wellbore pressure drop.
24
By contrast, an EOP fracture is produced
Created
by a sudden burst of pressure. This high-rate Perforation fracture
pressurization of the rock results in a rate-
dependent fracture mechanism that
<15°
approaches the ideal fracture system more
Preferred fracture plane
closely than hydraulic fracturing. Instead of
opening the door by pushing gradually, an
EOP operation is analogous to breaking the
door down with a sledgehammer. Because B
the fracture breakdown pressure, EOP frac- 15 to 30°
tures initially overwhelm the in-situ stress
field and probably extend straight from the
wellbore, like spokes from the hub of a
wheel, then turn gradually into the PFP
(left ). Fractures may form at all perforations,
but extend only from those nearest the PFP, C
creating a biwing fracture.
Multiple parallel fractures are not seen in 30 to 60°
EOP studies, possibly because the sudden
pressure load closes the microannulus
between the cement and formation, shutting
off a path for secondary fractures. Another
possible explanation for the absence of mul-
tiple fractures is that rate-dependent fracture
initiation favors only weakest portions of the > 60°
D
rock. This observation is borne out in con-
ventional fracturing, where high-rate injec-
tion is known to minimize creation of multi-
ple fractures. EOP also allows a larger angle
between the PFP and perforations before
fractures ignore the perforations and initiate
at a site on the wellbore aligned with the
PFP (right ).13 ■Rate of reorientation of an extreme over-
balance fracture into the PFP. Angles of
EOP Candidates less than 15° (A) result in rapid reorienta-
tion. From 15 to 30° (B), reorientation is
Widely accepted candidate criteria for EOP more gradual. At 30 to 60° (C), the fracture
are low permeability (below about 10 md), initiates at the base of the perforation and
reservoir pressure insufficient to achieve runs parallel to the tunnel before turning.
cleaning with underbalance, a highly Beyond 60° (D), the fracture ignores the
perforation and initiates at the sandface.
mobile clay content, and the need to estab-
[From Salsman A, Behrmann L and Brown-
lish a fracture in multiple layers, even those ing G: “Extreme Overbalance Perforating,”
with different mechanical or flow properties. The Perforating and Testing Review 8,
Other reasons to use EOP are reduction in no. 1 (May 1995): 1-8.]
near-wellbore damage and pressure drop
associated with poor linkage of fractures fol- Dees lists two leading reasons for EOP:
lowing underbalance perforating, to avoid immediate indication of well producibility
several days of swabbing before the results (which can be delayed with underbalance
of underbalance perforating are known, and perforating) and skin reduction. Behrmann
to eliminate near-wellbore tortuosity during and McDonald also list as applications
hydraulic fracture stimulation of extended- diversion of acid in carbonates and intersec-
reach wells.14 tion of natural fractures.15 Other authors
Oilfield Review
 50
0.1 md

40
1.0 md

Fracture extension, m
30

10 md
20

10
100 md

1 darcy
0
0 5 10 15 20 25 30 35
Time, sec

■Theoretical modeling of fracture extension versus time for various permeabilities.

The zone of fracture extension corresponds to the sloping section of the curves. Where the
curves flatten indicates the time after which only fracture erosion takes place. At 100 md,
fracture extension reaches its maximum of about 10 m in 15 seconds. The remaining time
probably represents only erosion in the fracture. Fracture extension in a 0.1-md setting
continues for about 30 seconds and reaches 45 m [148 ft]. (From Couët et al, reference 3.)

present several scenarios for application of
EOP in high-permeability settings, often for
placement of acid or proppant. Marathon
engineers, on the other hand, use EOP in
the Rocky Mountains of the USA mainly in
hard rocks, where they assume perforations
penetrate only 10 in. [25 cm] and so remain
in the damaged zone. In hard rocks, EOP
drives the effective wellbore radius well
beyond the damaged zone.
The majority of extreme overbalance pro-
ponents view it as complementary to con-
ventional underbalance perforating and as a
precursor to conventional hydraulic fractur-
ing. Bryan McDonald, who studies extreme
overbalance methods for Schlumberger, has
ranked variables that influence the choice
between EOP and underbalance perforating.
In order of importance, the leading variables
are as follows:
Well depth—The limiting influence with
depth is friction between the fluid and tubu-
lars, which reduces ability to deliver suffi-
cient pressure at the perforations. The mini-
mum pressure gradient was originally
placed at 1.2 psi/ft [27 kPa/m], but this has
inched up as insufficient pressure was
thought responsible for early failures. Some
operators today use gradients as high as 4
psi/ft [90 kPa/m]. In general, deeper wells
require a lower pressure gradient because
the energy available to propagate a fracture
is proportional to:
(pressure gradient × depth) – in-situ stress
in which the in-situ stress is given by the
fracture gradient × depth, so the expression
becomes:
(pressure gradient – fracture gradient) ×
depth.
The fracture gradient may vary from 0.4 to
1 psi/ft [9 to 22 kPa/m], with a typical value
of 0.7 psi/ft [15.8 kPa/m], so that with
increasing depth, the gradient difference
can be reduced. Even with the lower pres-
sure gradients required in deep wells, depth
can become a limiting factor at 10,000 to
15,000 ft [3000 to 4500 m].
Technique, however, sometimes compen-
sates for physical limitations. Oryx has suc-
cessfully treated an interval at 19,000 ft
[5790 m] by placing calcium bromide water
above nitrogen to deliver 24,000 psi [165
MPa] at the perforations. “We always place
less than 1000 ft [300 m] of fluid on bot-
tom,” said Pat Handren of Oryx. “This way,
energy goes into fracturing, not into over-
coming fluid friction.” Still, most operators
are more comfortable performing EOP
above 10,000 ft true vertical depth.
Permeability—There may be no more con-
and leakoff result in significantly shorter frac-
tures (above ).
Work by Petitjean and Couët indicates that
at 100 md, up to 80% of fluid can leak off in
less than 10 seconds, limiting an already
short fracturing event. 16 Some operators
place the cutoff at 100 md, although others
have claimed success at 1 darcy. John Dees,
who has a patent on continuous pumping
immediately after the extreme overbalance
event, maintains that EOP followed by surg-
ing with resin can succeed even with perme-
ability in the 1-darcy range.
Pat Handren considers permeability-length
a more useful parameter than permeability
alone. This takes into consideration vertically
variable permeability, which can affect flow
properties. Handren’s breakpoint is about 20
md-ft. If permeability hits this value, Han-
dren would choose underbalance perforat-
ing—unless reservoir pore pressure is below
0.35 psi/ft [8 kPa/m] in which case he prefers
EOP to be sure perforations are cleaned. At
permeability around 1 darcy, some forma-
tions may be friable enough to require EOP
with injection of resin to prevent sand from
13. Behrmann and McDonald, reference 7.
14. Petitjean et al, reference 3, ARCO method.
15. Behrmann and McDonald, reference 7.
16. Petitjean et al, reference 3.

troversial an EOP topic than how permeable

is too permeable. The concern is that leakoff
will outpace the flow rate needed to main-
tain a pressure that exceeds rock strength
and extends the fractures. High rates of spurt

Autumn 1996 25
flowing out the tunnel or to prevent the tun-
nel from collapsing.
Tubular and wellhead ratings— Tubing
diameter and pressure ratings limit gas pres-
sure and volume, which determine horse-
power deliverable at the perforations. Bigger
is always better, and biggest and strongest is
best. At a minimum, tubing needs to endure
1.4 psi/ft. ARCO uses up to 7-in. tubing on
the North Slope in Alaska, USA, and
Marathon has moved from 31/2-in. to 41/2-in.
tubing wherever possible.
Likewise, wellhead pressure control
equipment must at least match tubular rat-
ing. The objective is to have tubulars that
can safely withstand the pressures neces-
sary to deliver a fracturing pressure at the
perforations. ARCO’s rapid overpressured
perforation extension, or ROPE, method
owes much of its development to the
opportunity at Prudhoe Bay, Alaska, of
working on closely spaced wells with large
tubing (more than 2 7/8 in.).
Perforated interval length—Dissipation of
pressure over distance limits the interval
length that can be effectively treated with
extreme overbalance. Treatments on inter-
vals of up to 1000 ft have been performed
in a few wells, but most operators are confi-
dent that uniform, effective treatments can
be carried out over only 70 to 100 ft [21 to
30 m]. Shot density and permeability also
influence treatable interval length. As a
rule, Oryx finds that 1 shot every 2 ft [60
cm] will maintain sufficient pressure over
several hundred feet. Marathon prefers to
limit intervals to about 50 ft [15 m].
Elements of EOP Design: Hardware Before After
To one degree or another, the first genera-
tion of EOP jobs was constrained by precon-
ditions of the well completions. Now, oper-
ators recognize that success of the
procedure often relies on planning comple-
tions to optimize EOP jobs. The main con-
straints are surface-control equipment and
tubular ratings. ARCO makes routine use of
a wellhead isolation tool, or tree saver. This
device fits on top of the wellhead and has a
mandrel that extends through the wellhead,
sealing in the tubular. Downhole, Oryx will
use casing with a higher pressure rating, and Shock
run cement bond logs to determine whether absorber
the interval to be perforated is fully
cemented. When cement bond cannot be
confirmed, Oryx prefers to keep pressure
only on the tubing, using an isolation valve Shear
to avoid exceeding the casing burst rating.17 pins
In addition, Oryx uses the largest possible
tubing diameter to deliver the largest possi-
ble volume of gas, and always pressure tests
tubing to be sure it meets its rating.
After the completion configuration, the Air
next most important variable is wellbore chamber
fluid composition. All operators agree that
Operating
whatever type and quantity of fluid are piston
used, friction reduction polymers are a
must. In many operations, they are already a
given. A review of fluid design criteria and
their application has been prepared by John Pressure
Dees (below ). ports
EOP calls for new thinking about down- Firing pin
hole equipment, from the firing head to the
guns. A new kind of firing head, for exam-
ple, has been designed to better accommo-

Reservoir Conditions Fluid Type

Sandstone (some carbonates)
Sandstone with water retention
Carbonate
Sandstone with severe
water sensitivity
Carbonate and sandstone
Shaly low-pressure gas,
or <1000 psi
Heavy oil, paraffin present or well
drilled with oil-base mud
Gas sandstone <1000 psi
2% KCl water
2% KCl water with alcohol
15 to 20% HCl
Less crude
10% acetic acid
10% acetic acid or 1.5% HF ■EOF firing head before and after firing.
The firing head nests inside either a pro-
duction valve or a gun release system,
Xylene and is designed for no-delay firing.
Diesel

■Rules of thumb in fluid selection.

26 Oilfield Review
date the special needs of EOP jobs. Conven- SXAR Equipment Detonating Cord Initiated, Guns Dropped,
as Assembled Guns Shot Pressure Applied
tional TCP firing heads are activated by posi-
tive pressure, but have a hydraulic time-
delay mechanism that stalls firing for up to
15 minutes until activating pressure is bled
off to achieve underbalance. This delay is
Release
unnecessary with EOP—good safety practice housing
minimizes exposure of surface and down-
hole equipment to high pressure. For this
purpose, Schlumberger introduced the EOF-
BA extreme overbalance firing head, which
has no time delay (previous page, right ). The
firing head starts the train of events leading
to gun detonation when pressure exceeds
Release
the predetermined strength of shear pins. pin
This pressure drives an operating piston
upward to release a firing pin. Because the
piston must move against gravity, the firing
head is unaffected by vertical drops.
The EOF firing head nests inside either a
gun-release or an isolation valve system,
depending on completion type. The Explo-
sively Initiated Automatic Release (SXAR) Release
system is used for permanent completions, piston
in which the guns are fired and immediately
dropped into the rathole, allowing the well Break
to come on line immediately after perforat- plugs
ing (right ). The Explosively Initiated Produc-
tion Valve (SXPV) is used for “shoot-and-
pull” operations, in which the well is often
killed and the guns removed to provide an
unobstructed flow path, to run the comple-
tion string or to perform other work such as
a frac job.
In both completion types, safety and well
performance depend on quick and precisely
timed release of pressure. The valve and
gun-release system must assure that release
Pressure
of pressure from the tubing is synchronized
with detonation of the guns, thereby keep- Applied Hydrostatic Atmospheric
ing excess pressure off the casing. If unper-
forated casing is subjected to extreme over-
balance pressure, it can damage mechanical ■SXAR system for extreme overbalance operations. A member of the X-Tool equipment
components in the well, cause a packer family for permanent completions, in which guns are dropped into the wellbore rathole
leak, blow packers uphole (packers are after firing, this completion tool makes use of a break-plug mechanism for rapid release
of guns and allows production immediately after perforating. Applied pressure (red),
often the weakest link), burst casing or col- before and after activation, is typically 10,000 psi.
lapse perforating guns.
For this reason, Schlumberger developed a
unique mechanism that assures rapid

17. For a safety checklist prepared by Oryx, see the

appendix in Handren et al, reference 3.

Autumn 1996 27
 ■Heart of the X-Tool release of pressure only when perforation is
system, before and
after detonation. certain. The heart of this device is a stack of
Whole break plugs break plugs that have high compressive
(beside the golf ball) strength, but low lateral strength (left ). The
support the trip plugs can support the compressive force of
mechanism, but
the pressure applied to the valve, but not
shatter when the
detonating cord, high lateral stress. The detonating cord
passing through passes through the plugs and if it undergoes
their center, under- a high-order detonation—which virtually
goes a high-order assures firing of the guns—the lateral shock
detonation. This vir-
tually assures that of the cord detonation shatters the plugs and
guns release or the trips the mechanism. Because the SXPV and
production valve SXAR are activated by the detonation train
opens at the that fires the shaped charges, there is little
moment the guns
chance of loading the casing prematurely.
fire, safely keeping
high pressure off The SXPV and SXAR belong to a family of
casing, packers and five new completion tools, the X-Tools, that
surface equipment. make use of this gun-activation method.
The SXPV design enhances EOP perfor-
SXPV Equipment Detonating Cord mance by opening the flow ports faster than
as Assembled Initiated, Guns Shot Pressure Applied
any other valve (below left ). Rapid transfer
of pressure to the perforations is essential,
Flow ports since a typical EOP event lasts 15 seconds
Flow ports open at most, and must take advantage of each
closed
second to maximize the amount of work
applied to the formation. Loss of time con-
veying fluid from the tubing to the perfora-
tion translates into reduction in fracture
length and possibly in fracture width.
The SXPV valve begins opening 8 to 20
Firing head
milliseconds (msec) after gun detonation,
and is fully open after another 4 msec (next
page, top ). This high-speed operation means
the valve opens fully before fluid pressure
starts transferring from the tubing to the per-
Detonating
cord tube foration tunnels. With a conventional,
adapter hydraulically activated valve, detonation to
full opening of the valve can take a full sec-
ond, or 10% of a typical 10-second EOB
Operating event. This time delay means that less energy
piston goes into creation of fractures, resulting in
less fracture length (next page, bottom ). In
addition, the SXPV flow ports provide a flow
area 11/2 times that of the tubing diameter,
which minimizes friction and maximizes
transfer of pressure from the tubing.
In permanent completions, where guns are
Release pins dropped into the rathole after detonation,
Support the release system must overcome two engi-
sleeve
neering challenges: It must assure isolation
Break plug of completion hardware from gun shock,
support
which can unseat packers, and it must over-
Break plugs come friction in deviated wells that prevents
the guns from falling. Unassisted, guns usu-
ally will not drop until well deviation
reaches 40 to 45°. Excessive sanding and
debris can also prevent gun release.

■SXPV equipment for EOP. This X-Tool is used when wells are killed to remove guns—
shoot-and-pull—for unobstructed flow, running completions or to perform other work.
Break plugs ensure rapid opening of the production valve. Red is applied pressure, blue
shows hydrostatic and green is atmospheric pressure.
28 Oilfield Review
 ■A simplified view
7000 Pulse lasts 15 to 20 seconds of the first millisec-
Gun shock after this point
onds of perforating.
6000 The SXPV valve
begins opening as

Extreme overbalance pulse

5000 early as 8 msec
after gun detona-
4000 tion, and is fully

Pressure, psi
open after another
3000 4 msec. Rapid
opening means the
2000
valve opens fully
X-valve before fluid pres-
opens sure starts transfer-
1000 Gun
firing ring from tubing to
Hydrostatic the perforation tun-
0 nels, maximizing
energy that goes
-1000 into creation of
fractures.
-2000
0 1 2 3 4 5 6
Time, msec

wellbore fluid inflow

Jet interaction with

Maximum annulus
Jet tail formed

Slug exits gun

pressure from
wellbore fluid

Gun open to
penetration
gun swell

End of
0 1000
Time, µsec

20

16
C
E
Fracture radius, m

12
D

8
B

A
4

0
0 5 10 15 20 25 30 35 40
Time, sec

■How valve speed makes a difference. The more energy that goes into creating
fractures, the greater potential to extend fracture length. In these five ARCO wells,
examples A, B, D and E had enough nitrogen to cause significant erosion, whereas
example C ran out of gas, literally, before the curve plateaued and fractures
reached full extension. (From Petitjean et al, reference 3.)

Autumn 1996 29
 Accelerometer Hydrophone
Guns fired
and dropped
3.07 sec (66.68 ft)
Guns hit
bridge plug
■Fastest gun this side of the Pecos. The shot detection
system, which records accelerometer and acoustic
(hydrophone) signals during gun firing, documents
the rapid gun release of the SXAR system. In this well,
a 150-ft gun deviated 76° dropped in 3.07 seconds into
a 66-ft [20-m] rathole.
The SXAR addresses both shock and fric- The guns themselves provide the final link
tion problems by timing the firing of guns in the perforating chain. Charge type can
shortly after disintegration of the last break significantly influence results. Failure of
plug. At the instant the guns fire, they have some early EOP jobs was traced to a kind of
already been released, so the shock wave is screenout caused by plugging of perfora-
expended overcoming friction rather than tions with debris. Evidence for this was sur-
traveling up the completion string. Tests and face pressure that did not decline after the
field trials show the system works in a range guns went off. Investigators discovered that
of settings. In vertical wells, guns drop at 20 perforator debris behaves differently under
ft/sec [6 m/sec] and reach terminal speed a different settings. It can be permeable under
few milliseconds after release.18 The detona- the slowly rising pressures of hydraulic frac-
tion shock also effectively moves guns in turing, which can pump through some
highly deviated wells. In one well, the SXAR debris. But to the high-speed extreme over-
system successfully released a 150-ft [46-m] balance pressure pulse, debris can act as an
gun that was deviated 76° at the top and impermeable barrier. Behrmann and col-
86° at bottom (above ). leagues demonstrated that in EOP jobs,
accumulation of this debris, and subsequent
increase in pressure required to blast
through it, is minimized by use of deep pen-
etrators instead of big-hole charges.
30
Shot phasing also plays a role. In vertical
wells, a minimum of 120° phasing will typi-
cally result in two thirds of the perforations
lying within 45° of the PFP, the maximum
distance before EOP fractures start forming
away from perforations. Phasing choice is
also affected by procedures following EOP.
If EOP is a precursor to hydraulic fracturing,
a single biwing fracture is most desirable. In
this case, 60, 90 or 120° phasing is optimal.
However, if matrix acidizing is to follow,
then a higher, 45° phasing will help dis-
tribute acid around the wellbore.
While most operators use 60 to 120°
phasing, ARCO prefers 180°. Based on
large near-wellbore pressure losses in Prud-
hoe Bay wells deviated 30 to 50°, ARCO
determined that multiple fractures were
developing from 60° phase shots. By
changing shot phasing to 180°, the pressure
drop fell from 2000 psi [13.8 MPa] to under
500 psi [3445 kPa], which contributed to
placement of a larger amount of proppant.
“We think this suggests a single, biwing frac-
ture,” said Joe Schmidt, the ARCO engineer
who helped develop the ROPE program. In
gun systems with phasing other than 60°,
maximum benefit is obtained by keeping the
same alignment for all gun segments.
Shot density—the number of shots per ver-
tical foot—becomes increasingly important
in longer perforation intervals. Too high a
density can result in excessive leakoff. Too
low a density can result in longer exposure
of tubulars to high pressure and increased
risk of unseating a packer. In general,
because extreme overbalance procedures
produce instantaneous flow at 100 to 200
barrels per minute, a rule of thumb is a shot
density two to four times normal. ARCO,
for example, typically shoots EOP jobs at 4
shots per foot using 120° phasing over a
20-ft interval.
Oilfield Review
Elements of EOP Design: Operations Comparison of Extreme Overbalance Methodologies
Today’s extreme overbalance operations fol- ARCO Marathon Oryx
low one of two basic routes: perforating as a
Reservoir Properties
stand-alone event followed by pumping, or
perforating and surging at the same time Interval, ft 20 to 60 20 to 120 4 to 1501
(right ). In the Oryx method, EOP is followed Depth, ft 8000 to 15,000 4000 to 9000 4000 to 15,000
by high-rate pumping. In wells that are Bottomhole pressure, psi 10,000 to 11,000 500 to 4000 500 to 11,000
already perforated, Oryx finds that continued
Permeability, md 10 to 300 10 to 150 <100 or 100 to 300 if
pumping after overbalance surge can reduce pressure insufficient
skin, as long as the pumping precedes flow for underbalance
from the reservoir.19 The surge method, using perforating
a frangible disk or expendable plug, has Job Design
proved to reduce pressure requirements of
Gradient,2 psi 1.2 to 1.6 1.8 to 2.1 1.4 to 3.0
subsequent hydraulic fracturing.
ARCO, working on the Alaskan North Fluid/gas <300 ft fluid; N2 at <1000 ft fluid; N2 at <1000 ft fluid; N2 at
>50,000 scf, with >50,000 scf, with 50,000 to 100,000
Slope, reports similar results with its ROPE slick water on top slick water on top scf, with slick water
method and, since 1994, a high-energy ver- and 15% acid on top
sion nicknamed HE ROPE. The main differ- Guns TCP guns; wireline TCP guns, usually TCP guns, 60° or
ence between the Oryx and ARCO methods guns with big-hole 60° phasing, 120° phasing, 4 or
is that ARCO uses nearby wells as holding charges for new 4 to 6 spf more spf
tanks for high-pressure nitrogen. Instead of completions;
pumping, ARCO pressures up nitrogen in 180° phasing
nearby wells, connected by hard line to the Other Surge disk for previ- Proppant carrier over Resin gel typical for
treated well. A plug is set in the tubing tail of ously perfed wells TCP carrier sand control
the storage well, and both the target well Large-diameter tubu- Surge disk for previ- Clean fluids
lars to reduce friction ously perfed wells essential to success
and storage well are simultaneously pres- loss and increase N2
sured. At the appropriate time, the surge disk Pump after all jobs,
volume same day or
ruptures with firing of the perforating guns, High-Energy ROPE next day
releasing all the gas into the target well. method, in which N2
The HE ROPE method uses nearly 100% in a nearby well is
nitrogen, with a small amount—10 to 20 bbl pressured and piped
to the well being
[1.6 to 3.2 m3]—of liquid over the guns. Bot- treated
tomhole pressure exceeds 10,000 psi, with
Results
an overbalance of about 6500 psi [45 MPa].
The high pressure and high volume of nitro- 40% of wells with Increased production Negative skin on 88%
production increase rates and oil cuts of wells
gen lead to injection rates close to 270 bar-
of 10 to 20% in highly saturated No increase in
rels per minute [40 m3/min]. This rate results intervals
80% of wells requiring recoverable reserves,
in greater fracture width, erosion of the for- low frac pressure Completion time just in recovery rate
mation by gas and reduced near-wellbore when frac job follows reduced 30% Fractures likely to
pressure losses.20 The high-energy version Additional 15 to 20% Completion cost stay in zone due to
doubles fracture radius over conventional of frac sand placed reduced 11% low volume and short
ROPE, and because fractures tend to propa- 35% of TCP jobs period of fluid/gas
gate in a straight line, also allows for better fraced later inflow
Higher percentage of EOP as screening for
18. Huber K and Egey J: “SDET-B Confirms SXAR Fast frac sand placed other treatments: if
Release and High Speed of Dropping Guns,” The no response to EOP,
Perforating and Testing Review 6, no. 3 (October no further intervention
1993): 19. planned
19. Dees JM and Handren PJ: “Extreme Overbalance 1 With significant wellbore storage effect, the interval maximum rises to 500 ft.
Perforating Improves Well Performance,” World Oil
2 Achieved gradients sometimes fall slightly below the designed value, mainly because crews may feel more comfortable with lower
215, no. 1 (January 1994): 96-98. pressures. At the 1.4-psi gradient, however, the pressure on tubulars is usually no greater than that encountered during screenout
20. Couët et al, reference 3. of a conventional frac job, although EOP pressure may be applied for a longer period.

Autumn 1996 31
 Extreme Overbalance Perforating Underbalance Hydraulic Fracturing

M
a
ho xim
riz um Time 1 Time 1
■The many become on
ta
one. Extreme over- ls
tre
balance perforating ss
in a horizontal well
causes a single frac-
ture to develop from
multiple perfora-
tions, since the
applied pressure
exceeds in-situ m ss
imu stre
stress. By time 2 Min ontal
r i z
(bottom left), separate ho
fractures from sepa-
rate perforations Time 2 Time 2
have merged and
rotated into the
plane of maximum
horiztonal stress.
With hydraulic frac-
turing in the same
stress scenario (right),
separate fractures
develop.

height control during fracturing of thin inter- ■Hot proppant.

vals (above ). A gamma ray
Several operators have looked into sus- tracer log in a
Marathon well
pending proppant in extreme overbalance shows a concentra-
fluids, in hopes of blowing it into the perfo- tion of tagged
rations and propping the fractures. Most proppant in perfo-
efforts have failed because of the high shear rations (yellow)
rate of fluid-proppant mixtures. At pump More proppant
appears to have
rates used in treatments, the fluid becomes been injected into
a “plug” and will not flow. Another prob- the upper two inter-
lem is suspending the proppant uniformly vals than into the
in the fluid while it is being pumped down bottom interval.
Subsequent produc-
the tubing. tion testing con-
Nevertheless, progress has been made. firmed that there
Marathon, for example, along with Owen was no communi-
Oil Tools has developed a proppant deliv- cation behind pipe.
ery system positioned just above the guns (From Snider et al,
reference 21.)
that minimizes difficulties of mixing prop-
pant with completion fluids (such as uncer-
tain suspension time and increased viscos-
ity). Marathon has run tests with 200 to 400
lbm [90 to 180 kg] of radioactively tagged
proppant. Gamma ray scanning logs run Work is just starting on EOP in horizontal overbalance methods in this setting may
afterward indicate no proppant in the wells, which present a new range of chal- require a higher pressure gradient, well
rathole and high radioactivity at perforation lenges related to gravity. An underbalance above 1.4 psi/ft, to pulverize and move
entrances (right ).21 “We know it’s going into perforating experiment at TerraTek shows debris that tends to fall to the low side of the
the holes,” said Phil Snider of Marathon. that detonation products could not be hole. Mobil and Halliburton have used EOP
“But we still don’t know if we’re running flushed from down-side perforations in a successfully in a horizontal well at 15,693 ft
enough proppant volume, or for enough wellbore deviated 29°, even with a 400-psi [4783 m] true vertical depth to tap a tight
time, to get effective propping.” [2756 kPa] underbalance differential. Based gas reservoir. Perforating guns were con-
on this result, some operators are using ori- veyed on drillpipe.22
ented guns and perforating horizontal drains
on the top and sides only. Applying extreme

32 Oilfield Review
Unanswered Questions
“Most of the time, we know when extreme
overbalance will work,” said Phil Snider of
Marathon. “But the few jobs that go wrong
are the ones driving our research.”
This sentiment informs much of today’s
investigations into extreme overbalance
methods. But as knowledge increases, so do
the apparent number of unknowns. Most
workers in EOP have a wish list of what
would make their lives easier and more pro-
ductive. Highlights of this list include the
following questions:
What is the pressure-rise time? Knowing
the shape and exact length of the pressure-
time curve from the moment of gun dis-
charge will allow calculation of the number
of fractures and fracture width. In addition,
watching what happens in the first few sec-
onds can shed light on friction levels to opti-
mize job design. To achieve this, Snider is
using a downhole high-speed pressure gauge
that measures 20,000 data points per sec-
ond. Early results indicate that perforation
and fluid influx may behave more as a single
event than as two distinct ones. In addition,
results indicate that optimization requires
limiting flow restrictions in tubing and
reducing the height of liquid in tubulars.23
How does gun system design affect frac-
tures—big hole or deep penetrators? 24
Marathon’s experience indicates that big-
hole charges may lay down filter cake that
prevents injection of the surge. Except for
gravel packing, deep penetrators are gener-
ally the charge of choice.
How conductive are EOP fractures, and
how long do they work? Experience shows
that short, unpropped fracs have a useful life
of 6 to 12 months. As Marathon has
observed, proppant may pack EOP perfora-
tion tunnels, but may function more as an
abrasive than a proppant. And the abrasive
action decreases near the fracture tip, as
flow velocity decreases. In support of this
claim, large gains in well productivity with
the addition of a small volume of prop-
pant—100 to 200 lbm [45 to 90 kg]—sug-
gest scouring as the mode of action. Bauxite
has proved to provide better scouring than
conventional frac sand.
Over what distance do EOP fractures turn?
Experiments in large blocks of rock indicate
turning is complete in two or three wellbore
diameters, but there is evidence that in situ,
turning may take 5 to 10 ft [0.6 to 3 m].
Autumn 1996
Does EOP mean that subsequent frac
jobs require more or less horsepower?
Again, experience is ambiguous. Some
operators report always needing less horse-
power, while others say sometimes less,
sometimes more. Oryx finds that regardless
of horsepower required, EOP jobs always
result in placement of a higher proppant
concentration during subsequent frac jobs.
The treatment screens out, for example, at
8 lbm/gal [958 kg/m3] instead of 4 lbm/gal
[479 kg/m3].
Where Are You Going, EOP?
There are parallels between the develop-
ment of extreme overbalance perforating
and tubing-conveyed perforating. In the
1970s, TCP burst on the scene with Roy
Vann’s innovative designs for downhole
equipment. Hailed by proponents as a per-
forating panacea, and by others as an aber-
ration that would soon disappear, it was ini-
tially adopted by a few operators. Trial and
error smoothed out the rough edges, con-
vincing the industry at large of its technical
and economic benefits. Eventually, TCP set-
tled in as a niche service, which today
accounts for more than 25% of the perforat-
ing business.
EOP emerged in a slightly different con-
text. After 1986, low and stable oil prices
instigated a flurry of engineering creativity,
moving completion designs toward further
specialization and producing a change in
perspective. The idea of “completion equals
plumbing”—seals, tubulars, valves and
packers—began to give way to the view that
“completion equals well optimization.” By
the beginning of the 1990s, the completions
world was no longer neatly divided into
wireline, slickline, tubing and coiled tubing
techniques. Now the approach is to find the
best mix of solutions to optimize recovery.
Extreme overbalance is one such develop-
ment that blurs old boundaries. Usually per-
formed on tubing, EOP can be achieved
with wireline; it also marries perforating and
pumping, and inches toward stimulating,
but stops short at near-wellbore enhance-
ment. In the continuum of well treatments,
it more closely resembles matrix acidizing
than hydraulic fracturing, although its tech-
niques borrow more from the latter.
Whether EOP will join the mainstream in
the style of TCP or remain a small niche ser-
vice depends mainly on proof of its eco-
nomic viability. This proof requires further
technical refinements so that results can be
clearly related to technique.
One trend that may prove fruitful is the
recent move away from the pressure gradi-
ent rule and toward the tubular limit
rule—selecting the highest pressure safely
allowed by the tubular rating. ARCO, for
example, keeps pressure to within 80% of
tubular burst rating. However, it is still
unclear whether more pressure is always
better. By contrast, there is a move toward
larger diameter tubing, which typically has a
lower pressure rating but allows a higher
volume of gas. Also under investigation is
the use of proppant carriers, which have
been tried in only a few dozen wells.
ARCO has been looking to test the ROPE
method, with its precise control of fracture
height, on coal degasification projects,
where water intrusion often limits develop-
ment. Another possible use of this method is
in remote wells, where highly mobile nitro-
gen generation equipment could be readily
moved on site.
“We think of EOP not as a replacement for
the hydraulic frac,” said Joe Schmidt of
ARCO, “but as a pretreatment for fast return.
If we have trouble with a conventional frac
because of earth stresses or well deviation,
this works.”
Schmidt’s view represents that of many
workers trying to define the limits of the
EOP frontier. True, it is known to work in
some settings, and yes, the mechanics may
not be clear, but that is simply a problem
that will wash away with more study.
“We know just about all we can about
underbalance perforating,” said Phil Snider
of Marathon Oil Company. “Extreme over-
balance still presents the possibility of dis-
covering new benefits.” —JMK
21. Oriold F and Snider PM: “TCP Proppant Carrier Sys-
tems for Extreme Overbalance Perforating: Experi-
ence to Date,”paper CIM 96-82, presented at the
47th Annual Technical Meeting of the Petroleum
Society, Calgary, Alberta, Canada, June 10-12, 1996.
Snider PM, Hall FR and Whisonant RJ: “Experiences
with High Energy Stimulations for Enhancing Near-
Wellbore Conductivity,” paper SPE 35321, presented
at the International Petroleum Conference and Exhi-
bition, Villahermosa, Mexico, March 5-7, 1996.
22. Chambers MR, Mueller MW and Grossmann A:
“Well Completion Design and Operations for a
Deep Horizontal Well with Multiple Fractures,”
paper SPE 30417, presented at SPE Offshore Europe
1995, Aberdeen, Scotland, September 5-8, 1995.
23. Snider et al, reference 21.
24. A big-hole charge increases hole diameter at the
expense of penetration depth. Hole diameter in cas-
ing is two to three times greater than with a deep
penetrator, but penetration is substantially reduced.
33
 Multipurpose Service Vessels:
Versatile Toolkits for Well Intervention

The oil industry’s push toward greater efficiency and more integrated services is profoundly impacting

marine support activities. New concepts and designs that offer an expanded range of capabilities from

a single vessel rather than multiple boats and barges are rapidly transforming the face of well workover

and remediation operations. This article describes how innovative approaches are solving logistics and

performance problems that have challenged offshore operators and service providers for decades.

Sam Adamson In today’s oil field, delivering cost-effective advances is ushering in a new era of perfor-
Francisco Cupello solutions to difficult problems is para- mance from offshore barges and lift boats.
John Hicks mount. As oil companies continue their Former single-purpose or limited-use ves-
Malcolm Keenleyside intense drive to lower finding and produc- sels are being retrofitted to enhance effi-
Las Morochas, Venezuela ing costs and increase efficiency in every ciency and upgraded to add capabilities.
facet of operations, service companies are Purpose-built vessels, which can swiftly
Dave Formas working closely with them as proactive deliver a wide spectrum of services, with
Claude Gabillard
solution-providers. Success hinges on fewer constraints and with a smaller cadre
Montrouge, France
exploiting the new business relationships of highly skilled personnel, are being
that have sprouted and thrived during the designed and commissioned in key oil-
Francisco Gamarra
past few years, typified by a proliferation of producing regions worldwide.
Alexis Sanchez
alliances and integrated services contracts. To put it simply, today you can have a
Lagoven SA
Tia Juana, Venezuela At the core of these initiatives are align- Swiss Army knife at your disposal instead
ment of fundamental goals and an unparal- of a common penknife.
leled application of cutting-edge technol-
ogy targeted at productivity enhancement.1 Progress in the Lake
Cooperative operator-service company Lake Maracaibo, Venezuela remains one of
programs are closely examining the cost- the most prolific oil-producing areas of the
For help in preparation of this article, thanks to Larry effectiveness of oilfield equipment used world—contributing over 1.5 million
Hibbard and Yves Lemoign, Sedco Forex, Montrouge,
France.
throughout the field development cycle. BOPD [240,000 m3/d] to help satisfy the
WASP is a mark of Schlumberger. Swiss Army is a mark Radically different approaches are being ever-increasing global demand for energy.
of Victorinox. adopted to reduce logistical complexity and This represents more than 50% of the 2.9
1. Chafcouloff S, Michel G, Trice M, Clark G, Cosad C
and Forbes K: “Integrated Services,” Oilfield Review 7,
nonproductive time, deficiencies associated million BOPD [460,800 m3/d] produced
no. 2 (Summer 1995): 11-25. with many long-standing practices. This is by Venezuela. Lake Maracaibo is also
Austin CB, Dole S, Chmilowski W, Vernon G, Heidt particularly true for well construction, inter- home to a unique field development
JH, Lewis R, Thompson J, Vinson M and Watson T:
“Alliances in the Oil Field,” Oilfield Review 7, no. 2 vention and remediation where a combina- scheme of over 11,000 wells and a unique
(Summer 1995): 26-39. tion of innovative thinking and technical set of well maintenance problems.
2. Offshore 55, no. 5 (May 1995): 36.
State-owned Petroleos de Venezuela SA
3. “Venezuela Awards Exploration Risk Contracts,” Oil &
Gas Journal 94, no. 30 (July 22, 1996): 28-29. (PdVSA) has committed to more than dou-

34 Oilfield Review

ble oil production capacity by the year
2005 with an impressive assembly of
drilling and workover programs. Three-
dimensional seismic mapping of the entire
Lake, completed in 1994, paved the way
for a new, intensive exploration campaign
that has already led to several important
light-oil discoveries.2
PdVSA’s programs are being comple-
mented by aggressive efforts from interna-
tional oil companies, including Mobil,
Enron, British Petroleum, Shell, Amoco,
Maxus, Conoco and Elf Aquitaine, as they
develop acreage acquired in recent offer-
ings. Earlier this year, Venezuela leased
acreage in the first bidding round since it
reopened to foreign investment in 1993.
Profit-sharing contracts covering explo-
ration on eight tracts with a combined area
of 18,000 km2 [6950 square miles] were
awarded. Undiscovered oil reserves in the
Autumn 1996
Maracaibo
Tia Juana
Lagunillas
newly assigned areas are projected at 7 to
23 billion bbl [1 to 3.6 billion m3].3 Lake Maracaibo
Lagoven, the company that along with sis-
La Ceiba
ter affiliates Maraven and Corpoven consti-
tute PdVSA, has seized the initiative with a
vengeance. In 1994, for example, 77 mil- Venezuela
lion bbl [12 million m3] of light crude were
added to reserves with three major discover- 54%
29%
ies in the Lake that helped boost Lagoven’s Heavy Medium
production. Currently, Lagoven maintains Crude production
over 8000 wells and produces about 14%
3%
730,000 BOPD [116,000 m3/d] from 4940 Light
Condensate
active wells mainly in the eastern sector of
the Lake (right). Water depths in the sector ■ Lagoven operations in Lake Maracaibo.
average about 60 ft [18 m] and well depths Lagoven produces from over 4900 active
average around 4000 ft [1220 m]. wells, mainly in the Campo-Costanero-
Bolivar region of the Lake (light blue).
Of the total number of wells, nearly 1000 Eighty-three percent of the wells produce
are presently shut-in, and as many as 4000 medium to heavy crude. Nearly 4000 out
require some type of workover operation. of a total inventory of over 8000 wells cur-
rently require some type of workover.
As the fields in this area of the Lake have
matured, the need for more frequent and
35
 Current Future

Conventional drilling and workover barges; support

barges (cementing, coiled tubing, gravel packing, logging) Integrated multipurpose units

Major scheduling and Streamlined scheduling

logistics problems and logistics

Excessive Stable platform,

weather-related delays enhanced mooring capabilities

Organization and Efficient, multiskilled

communications deficiencies organization

Insufficient, poor-quality data Rapid, high-quality data generation

■ Evolution in well workover practices. The logistical and coordination problems and excessive downtime currently experienced using
single-service barges can be solved by integrated, multipurpose vessels capable of more rapid, efficient, less costly operations.

complex workover and remediation proce-
dures has grown accordingly.
Although wells in Lake Maracaibo are
technically “offshore,” many operations are
reminiscent of land development. Flat-bot-
tom barges with cantilevered derrick sets
are used for drilling. Historically, Lagoven
built drilling barges up to 220 ft [67 m] in
length and then adapted them for a range
of workover functions. An array of separate
barges, outfitted to provide one or two spe-
cialty services each, such as cementing
and coiled tubing, supports the drilling and
workover barges. Currently, Lagoven owns
11 out of 28 barges and three jackups ser-
vicing its sector of the Lake.
When well construction or intervention
tasks call for multiple barges in a pre-
arranged sequence, the logistics of barge
scheduling and positioning are compli-
cated and inefficient. Time spent waiting
for the next barge to arrive is often exces-
sive. Lake Maracaibo is prone to rapid
weather fluctuations, further impeding
operations since most barges have minimal
tolerance to waves. Delays waiting for
weather to improve increase nonproduc-
tive time. The scene evokes visions of a
queue of rain-delayed jets awaiting takeoff
at a major airport. These constraints, and
having a limited fleet of service vessels,
restrict rig-based workovers to about 400
per year, far fewer than needed.
Lagoven’s efforts to sustain production
from depleting zones and increase overall
hydrocarbon recovery rely heavily on well
reentries and workovers. Lagoven views
this as the most economical way to
enhance hydrocarbon production and stop
its primary nemesis—water production.
This, in turn, has prompted a rethinking
about how workovers should be per-
formed. An evaluation of existing deficien-
cies and benefits from an integrated ser-
vices approach convinced Lagoven to
propose multipurpose barges for delivering
a suite of services from a single vessel.
Smaller, more flexible and cost-effective,
these vessels solve scheduling problems
and eliminate the excessive downtime
plaguing current operations (above).
The changeover will, of necessity, be evo-
lutionary. The progression from dedicated,
single-use barges to custom-designed, mul-
tipurpose vessels is currently passing
through an interim stage—barge retrofitting.
More Than a Cosmetic Makeover
For several years, service companies have
sporadically reconfigured—or retrofitted—
drilling and specialty service barges to
increase functionality by adding dedicated
equipment. These attempts represented the
first valid steps toward more efficient well
workover services. Today, efforts are more
directed and less experimental.
A notable effort by Dowell, based on
work in the late 1980s when barges were
first outfitted with coiled tubing units, has
helped bolster confidence in coiled tubing
drilling on the Lake. A 140-ft [43-m] barge
was specially configured to support either
conventional or coiled tubing drilling. To
date, nearly 50 wells have been drilled
from the barge with coiled tubing. Initially,

36 Oilfield Review
 ■ The Lagoven
406. This can-
tilevered drilling
barge has been
retrofitted to pro-
vide integrated
services by
adding coiled
tubing and fluid
mixing, blending
and pumping
capabilities.

hole sizes were 37⁄8-in. diameter. Now, wells ■ Reconfigured

are being drilled through shallow gas areas Lagoven 406
deck layout with
with diameters as great as 123⁄4 in. Well cas- new equipment.
ing, up to 10 3⁄4 in. in diameter, is run and Coiled The addition of
Coiled tubing tubing coiled tubing
cemented with the same unit. The success of reel power pack and WASP sand
this approach paved the way for the even Drill floor
mixer modules
more ambitious efforts that followed. on the deck level
Recently, two barges in the Lagoven provides addi-
Engine Engine tional flexibility
fleet—Lagoven 405 and Lagoven 406— for well workover
and and Tanks
were retrofitted. Lagoven 406 (LV-406), a WASP pump pump and remediation
cantilevered drilling barge with on-board sand programs.
mixer
derrick set and wireline logging unit, was
the first vessel with hoisting and rotating
capabilities to be retrofitted (above). Work-
ing alongside Lagoven, Schlumberger Oil- Deck level
field Services, with Sedco Forex as project
manager, reconfigured the barge in order
to provide integrated services.
The barge was delivered for retrofitting in
mid-February 1996 and equipped with a
1.5-in. diameter coiled tubing unit (control
cabin, power pack, reel and injector head),
WASP water and sand proportioning mixer
for blending gravel pack slurries, fluid fil-
tering system, silo, displacement tanks and
recirculating pump—replacing the bare
pipe racks on the deck of LV-406 (right).

Autumn 1996 37
 Upgraded piping and new systems for
air, water, fuel and electricity were added.
During the construction phase, a premium
was placed on efficient use of deck space.
Since the reconfiguration primarily
involved coiled tubing and pumping
equipment, a team of experienced Dowell
engineers was appointed to define an opti-
mal deck arrangement (left).
Although over 100,000 lbm [45,360 kg]
were added to the vessel, the excellent load-
bearing capacity of LV-406 ensured that vari-
able deck load limits were not exceeded. A
comprehensive stability analysis conducted
by a naval architect confirmed that vessel
stability would not be compromised.
The reconfigured barge was placed in
operation in mid-April 1996 (below left). In
its first few months of operation, the retrofit-
ted LV-406 has mainly performed horizontal
well cleanouts and cement plug place-
ments with coiled tubing, as well as wire-
line operations, including perforating and
■ Deck and mez-
setting bridge plugs or permanent packers.
zanine levels
showing individ- Previously, cleanouts of produced sand
ual equipment required use of larger diameter drillpipe or
modules. Optimal
contracting for a separate coiled tubing
module place-
ment was deter- barge to work in tandem with LV-406. The
mined by a team former was expensive and time-consuming;
of experienced
the latter meant dealing with ever-present
Dowell engineers.
The placement of scheduling and positioning problems. With
each unit was two barges, the well would often sand up
based on control,
again after cleanout, before the rig could be
interface and pip-
ing criteria. repositioned to install a prepacked sand-
control screen.
With the retrofitted LV-406, however,
cleanout of a 3000-ft [914-m] horizontal
section typically requires only 17 hours
compared to 36 hours with multiple barge
operations, a savings of 53%. Worked-over
wells now produce sand-free.
According to Francisco Gamarra, Drilling
and Workover Manager for Lagoven’s Tia
Juana District and Alexis Sanchez,
■ Looking across Workover Engineer in charge of LV-406,
the deck toward
the cantilevered $220,000 and 15 days were saved working
derrick. One goal over the first 10 wells.
of the retrofitted “With everything now on the same barge,
design was a
functional, safe multiple barge moves have been eliminated
and uncluttered and logistics have improved significantly,”
deck. Equipment says Gamarra. “We have taken a major step
modules are
arranged for forward in the quality and efficiency of our
easy change-out operations. Cross-training between coiled
and efficient tubing and conventional workover experts
delivery of the
widest possible
range of services.

38 Oilfield Review
has reduced rig-up time. Based on this suc-
cess, we plan to convert our remaining
barges over the next three years.” 1 Coiled tubing reel
2 Injector head
Designing the Ultimate Toolkit 3 Jib, 8.5 metric ton capability
Retrofitting has its place as a logical, 3
4 Injector head (working position)
intermediate measure. The long-term goal, 5 Riser
however, is a fleet of versatile, purpose- 6 BOP stack
built vessels. “We don’t have a single
mindset on how to accomplish this. We 2

are willing to listen to the innovative ideas

service companies are proposing,”
4
Gamarra indicates.
1
Designing vessels for the stringent 5 Helideck
workover and intervention activities
e x p e c te d in th e fu ture requires an 6
exhaustive analysis of the parameters
Main deck
affecting practices on the Lake. What
basic vessel structure should be used?
How can deck equipment best be inte-
grated to maximize functionality and effi-
ciency? Reconfiguration of LV-406 pro-
v id e d inva lu a b le e x perience in the
■ Barge configuration. Starting with a clean slate, Schlumberger engineers designed a
definition and selection process—high- customized vessel—PRISA—to fulfill the well workover and intervention needs of
lighting what worked and what didn’t. Lagoven. The efficient, compact design affords modular flexibility and adheres to
The assessment required three iterations, stringent quality, health, safety and environmental guidelines. The barge version of
PRISA is shown with a modified cantilevered derrick, supplemented by a coiled tubing
with each revision giving a tighter, more unit for reentry and underbalanced drilling.
cost-effective design.
The result is a new vessel concept called
PRISA—a Spanish term which translates ”to
do things quickly”—symbolizing the objec-
tive established at the outset. PRISA is envi-
1 Crane, 100-ft boom length
sioned as the solution to Lagoven’s
2 Folding mast
workover and intervention needs for the
3 Drill floor
future—a multipurpose unit that reduces
4 Stern thrusters
logistical complexity, improves operational
5 Bow thruster
flexibility and efficiency, and decreases
time spent waiting on weather. 1
Two design variations are possible. The 2
first is a barge style, 180 ft [55 m] in length
with a 4000-ft2 [370-m2] deck, capable of 3
Helideck
operating in water depths to 150 ft [46 m],
Main deck
wave heights to 6 ft [1.8 m] and surface
currents to 2 knots (above right). Bottom deck
4 5
The second is a self-propelled, lift-boat
style, 180 ft long with a 6000-ft2 [558-m2]
deck, able to operate in water depths to
100 ft [30 m] (right). This configuration is
bottom-supported for better stability over a
wider range of environmental conditions.

■ Lift-boat configuration. This self-propelled version of PRISA is equipped with bow and
stern thrusters to permit easy mobilization and precise positioning.

Autumn 1996 39

Maximum
well depth,
8000 ft
Cement
plug
Maximum
existing wellbore,
7-in. casing
■ Well reentry and drilling of laterals with drillpipe or coiled tubing. The ambitious workover and production enhancement pro-
grams of Lagoven emphasize reentry of existing wells to drill deviated or horizontal sections to reach previously untapped areas of
the reservoir.
Services in both configurations include:
• conventional well repair and workover,
including workover fluid services
• conventional reentry drilling and snub-
bing (barge configuration only)
• coiled tubing drilling, including under-
balanced air or foam drilling with
quick changeover to conventional
drilling and pulling mode in the barge
configuration
• measurements- and logging-while-
drilling
• wireline logging and perforating
• pumping mud, cement, acid, gravel
packs and completion fluids
• slick-line operations.
But the design goes further, providing
capabilities for reentry and production
improvement drilling of short-radius hori-
zontal sections and multilaterals (see “Reen-
try Drilling Gives New Life to Aging Fields,”
page 4) from existing wellbores with casing
sizes up to 7 in. [18 cm] (above). These tech-
40
Maximum
horizontal extension,
2000 ft
6 1/4-in. sidetrack with drillpipe
or coiled tubing
niques will proliferate rapidly as crucial ele-
ments in the long-term production improve-
ment strategy for Lake Maracaibo. Late in
1996, Lagoven plans to drill two wells using
multilateral technology. For 1997, this num-
ber jumps to 25.
PRISA includes an integral, cantilevered
derrick and major support facilities—moor-
ing system, crane, living quarters, power
generation, and fluid pumping and storage
equipment. The deck space is readily
accessible by the main crane, allowing effi-
cient placement of service modules.
Hydraulic and power units support multi-
ple applications.
Automation and process control maxi-
mize productivity during running, tripping
and pulling of completion equipment. A
central control cabin monitors functions in
the derrick and on the drill floor, as well as
electric and slick-line operations.
In stark contrast to existing vessels, PRISA
simplifies logistics and enhances onboard
operator-service company coordination,
allowing rapid, sequenced operations. The
stable platform and an advanced position-
Target
reservoir
ing system decrease waiting-on-weather
time. Better equipment layout, modular
construction and tighter organization of
deck personnel reduce nonproductive time.
Quality, health, safety and environmental
performance improves, combined with
lower operating and maintenance costs—
overall, an impressive list of benefits.
Lagoven currently envisions construction of
up to six such vessels to support its produc-
tion enhancement initiative.
Indonesia and West Africa:
Reflecting Similar Needs
The well workover and intervention market
in Indonesia is sizable. As primary fields
mature, they require increased attention.
Estimates indicate that this market will
grow in the near term, with yearly
workover targets of 500 wells or more.
Today, there simply aren’t enough service
vessels. Oil companies plan workovers
using the available, diverse assortment of
(continued on page 43)
Oilfield Review

Performance and reliability in compact packag-
ing have epitomized the last decade of computer
and electronics breakthroughs. In similar fash-
ion, technical innovations are streamlining oil-
field equipment, decreasing failure rates and
delivering increasingly accurate data more
rapidly. Whether it’s onboard processing for
seismic surveys, revolutionary downhole wire-
line logging tools, or integrated fluid mixing and
blending systems, an insistence on speed, quali-
ty and versatility is permeating the industry.
Offshore well construction, maintenance and
intervention services that require vessel support
are being closely scrutinized. These tasks can be
performed more efficiently and cost-effectively
with multipurpose service vessels—MPSVs—
than with an assembly of traditional, single-ser-
vice units. With the number of deviated, horizon-
tal and multilateral wells mushrooming, and
revisited reservoirs requiring more complex well
completion schemes, custom-designed vessels
are increasingly viewed as the preferred equip-
ment of choice.
Multipurpose service vessels provide the ser-
vice company with an arsenal of capabilities and
integrated platforms for delivering a veritable
smorgasbord of services. At the same time, they
offer oil companies more options for improving
hydrocarbon recovery and extending the life
expectancies of offshore wells. The MPSV is a
win-win solution for both parties.
In late 1995, a self-propelled MPSV, Irish Sea
Pioneer, was commissioned by Halliburton Energy
Services for well workover and intervention activ-
ities in the BHP Petroleum Liverpool Bay field off
the west coast of England. It is the largest such
vessel built to date and underscores the benefits
that operators and service companies alike per-
ceive in the MPSV concept.1
The MPSV initiative within Schlumberger is a
natural outgrowth of the drive toward integrating
services and supplying value-added solutions in
direct response to client needs. Defining specifi-
cations for and designing the first-generation
1. O’Callaghan MJ: “Jackup Workover Platform Begins
New Era in Europe,” Offshore 55, no. 12 (December
1995): 42-43.
Autumn 1996
A New Generation of Capabilities
Radio room/
wheel house
Stern
controls
Crane with
120-ft boom
Jack house
Flare boom
14 ft
7 ft
Stern
thruster
300 ft
■Multipurpose Service Vessel. Outboard profile of the MPSV showing the primary control, lifting and propul-
sion systems.
MPSVs required the combined strengths and
expertise of the Schlumberger Oilfield Services
companies—led by Sedco Forex for vessel man-
agement and drilling expertise, Anadrill for direc-
tional drilling and measurements-while-drilling
services, Wireline & Testing for logging and well
testing services, and Dowell for pumping, fluids
engineering and coiled tubing services.
The resulting design adheres to the fundamen-
tal quality, health, safety and environmental
tenets being applied uniformly by Schlumberger
worldwide and conforms to standards estab-
lished by regulating organizations, including the
American Bureau of Shipping and International
Maritime Organization, and initiatives such as
Safety Of Life At Sea.
The chosen blueprint is a jackup-style vessel—
properly termed a lift boat—with four, 300-ft
[91-m] legs (above). The vessel is self-propelled
at speeds up to 6 knots, compared to jackups that
are towed at a sluggish 1 to 2 knots by tugboats.
Navigation mast
Helideck
Bow
thruster
Anchor
rack
It is smaller than a normal jackup, requires less
steel and is sleeker than the flat-bottom barges
typically used in areas like Lake Maracaibo.
The MPSV is equipped with four legs. This
configuration saves time during vessel moves,
improves safety, reduces maintenance costs and
provides greater space for equipment on and
below the main deck.
The MPSV has a streamlined bow and stern to
minimize hull resistance, and utilizes state-of-the-
art azimuthal thrusters for precise station keeping
and maneuverability. It can be used safely in
water depths as shallow as 10 ft [3 m] or as deep
as 170 ft [52 m] and is stable in 60-knot winds
and 29-ft [9-m] waves. Fully-retractable spud cans
are integrated into the hull design, minimizing
drag during transit. Top and bottom jetting capa-
bilities aid removal when leaving location.
The MPSV is self-elevating with leg-jacking
speeds of 6 to 8 ft/min [1.8 to 2.4 m/min], five
times faster than a conventional jackup. A
power-generation system that isolates propulsion
functions from jacking functions ensures that the
41
vessel can maintain accurate position, even in high currents, during jack-
ing operations and when adjacent to wellheads.
The 6000-ft2 deck is open, usable and supports a 1.5 million-lbm
[0.7 million-kg] deck load with the legs elevated and a 1.0 million-lbm 1 Container
[0.5 million-kg] load at other times. Twin 100-ton cranes with 120-ft
2 Surge tank
[37-m] booms are capable of moving equipment modules onto and around or knock-out
the vessel, and are qualified for major lifts directly to and from attended drum

platforms or wellheads. The cranes revolve a full 360° without interference 3 Crude oil
export pump
from the legs, increasing flexibility and reducing deck space requirements.
To permit extended, 24-hour operation, accommodations and water 1 Twin pump 4 Control skid
production facilities are provided for up to 60 people, as operations skid
5 Gauge tank
dictate. Additional personnel can be housed in temporary modules on 2 Nitrogen 1 6 Air
storage tank 1 2 2
the main deck, if necessary. Quarters are located for ready access to the compressor
helideck and lifeboats or life rafts. 3 Nitrogen
3 3 5 6
7 Oil manifold
MPSV services cover the spectrum from routine, daily repair and mainte- pump 1 4
2 8 Production
nance on wells and platforms to short or extended well tests, horizontal 7 separator
production logging and acid stimulation with coiled tubing, and drilling mud 1 Burner head 8
9 Lab cabin
treatment. The MPSV can be equipped as an early production system during 2 Crude oil 8
the initial stages of field development, allowing operators to fully evaluate reinjection 9 8 10 Logging
pump 2 cabin
reservoir potential while generating immediate cash flow. 10
3
A single process-control system, ergonomically designed primary con-
trol cabin and modular, integrated equipment layout simplify operations,
improve safety and permit staffing by a smaller cadre of multiskilled ■ Equipment configuration for extended well testing. The large, accessible deck
personnel than with a random assortment of single-purpose vessels. provides ample room for the variety of equipment modules needed to support an
extended well test.
Modularity allows rapid, easy customization. In the extended-well-test-
ing mode, the vessel is outfitted with production separators, surge tanks
or knock-out drums, export and reinjection pumps, a special logging
cabin, a twin, high-pressure pumping skid, flare boom and burner (top).
For horizontal production logging, a coiled tubing unit—including
power pack, reel and injector head—blender, a twin, high-pressure
pumping skid, production pump, and surge tank or knock-out drum are 1 Logging
used (bottom). For acid stimulation, a pressure-swing absorption nitrogen cabin
unit and acid storage tank are added. Other coiled tubing applications— 2 Container
nitrogen lift, logging and cementing—can be accommodated easily.
3 Surge tank
Electric or hydraulic high-volume, high-pressure pumps can be added to or knock-out
extend the operating range of coiled tubing and snubbing units. drum
For well repair activities, a cantilever, mast and hoisting facilities, mud 4 Gauge tank
pumps, fluid storage tanks and mud treatment skids are provided. The
5 Air
cantilever offers versatility by permitting direct positioning over the well- compressor
head. The MPSV can also serve as a diving-support vessel or as a tempo- 3
2 5
1 Twin pump 6 Crude oil
rary logistics and supply vessel for platforms or other offshore facilities. 1
skid reinjection
4 pump
Currently, MPSVs based on this design are being evaluated for
2 Blender
diverse areas, such as the Gulf of Mexico, west coast of Africa, 2 7 Oil manifold
3 Power pack 6
Malaysia and Indonesia.
1 8 Production
4 Reel 3 separator
7
5 Injector 4 8 9 Choke
head manifold
9
5

■Equipment configuration for horizontal production logging. The addition of a

coiled tubing unit and auxiliary equipment supports logging of horizontal wells, a
key duty expected for MPSVs as the emphasis on horizontal and multilateral wells
expands worldwide.

42 Oilfield Review
boats and barges, and must live with the
technical and operational restrictions inher-
ent in their designs. High-efficiency, flexible
units—at least four just to satisfy the demands
of major international players—are desper-
ately needed both to increase workover fre-
quency and to optimize scheduling.
The majority of today’s fleet are floaters,
mainly flat-top barges. Historically, these
were the cheapest to build and operate
while offering modest flexibility. But they
have severe limitations, require substantial
marine support—up to three auxiliary boats
for moving and positioning—and large
crews, 50% greater staffing than for effi-
cient vessels. They are also highly sensitive
to weather, often unusable if swells exceed
a few feet. Nonproductive time rigging up,
rigging down and waiting on weather runs
as high as 40 to 50%. These barges can
interface only with platforms equipped with
a derrick set or designed to accommodate a
rig or coiled tubing unit. This lack of con-
formance restricts certain operations and
promotes substandard quality and safety.
With a purpose-built intervention unit,
capabilities to run completions, perforate
and clean up wells—common practices in
many parts of world, but currently difficult
to impossible in Indonesia—can be added.
Today, importing specialized units for these
purposes is an option, but usually prohibi-
tive due to mobilization costs and incom-
patibilities with platform designs.
A Schlumberger study outlines the ideal
solution: a versatile, self-propelled, multi-
purpose service vessel—or MPSV—to min-
imize transport time and the amount of
marine support needed (see “A New Gen-
eration of Capabilities,” page 41). Modular
construction allows fit-for-purpose equip-
ment sets to be placed and changed out
rapidly and efficiently on deck, reducing
rig-up and rig-down time. The vessel has a
working deck at the same level as those on
production platforms to minimize lost time
and eliminate the need to transfer equip-
ment from barge to platform. The vessel’s
deck accommodates most service equip-
ment for greater efficiency, safety and
improved communications. Staffing is
Autumn 1996
reduced through multitasking and cross-
training in several services.
The vessel is capable of running and
pulling completions and can incorporate a
derrick with rotating capabilities to drill out
cement, mill windows and washover
downhole equipment. Onboard living
quarters and support facilities allow 24-
hour operation.
A generalized strategy, based on MPSVs,
is emerging to satisfy the burgeoning needs
of the Indonesian workover market. Even-
tually, when service vessel capabilities
expand sufficiently, there will be an addi-
tional benefit—the size, complexity and
cost of production platforms will decrease
substantially since fewer, lighter facilities
will be needed.
Similar well servicing challenges and
opportunities exist off the west coast of
Africa. Here, MPSVs will have to meet the
rigorous demands of a wide range of
clients and incorporate servicing modules
for well testing, early production facilities,
wireline logging and coiled tubing.
This means more deck space, variable
load capacity, hefty cranes, permanent
onboard equipment and a large workshop
to carry out maintenance and repair work
on platforms and wellheads. Many such
operations are currently performed by con-
ventional drilling jackups at high day rates.
Like operations in Indonesia, a nearly
identical MPSV concept provides the
answer. Since it is designed to be moved fre-
quently, mobilization is rapid and easy.
Transit times and costs are reduced, and
towing vessels are not required. To be finan-
cially attractive to operators, costs must be
lower than for conventional drilling jackups.
The MPSV is expected to bring greater func-
tionality at considerably lower day rates.
Managing a Successful Changeover
The drive to eliminate core inefficiencies in
equipment and operations is spreading
globally as a result of open communica-
tions and teamwork between operators and
service companies. The results are innova-
tive concepts to solve industry problems
that have lingered for decades.
Advances in offshore support vessel
design and functionality are opening up
new avenues for greater productivity and
efficiency. The evolution from single-pur-
pose to retrofitted designs and from inte-
grated service vessels to custom-designed,
multipurpose vessels has been a phased,
but relatively rapid, one. While each oil-
producing area has specialized needs that
must be fully addressed during design,
operators and service companies are real-
izing that a common strategy, MPSVs, can
be the ultimate answer—a versatile toolkit
for well intervention.
Implementation, however, is the real
challenge. It will require investing capital in
new equipment, instilling an across-the-
board commitment to quality and efficiency
in offshore servicing, and training a large
number of multiskilled teams to staff these
vessels. The potential rewards in terms of
productivity enhancement and improved
field economics, however, are enormous.
—DEO
43
 Environmental Applications of
Oilfield Technology

Darwin Ellis Investigation of subsurface sites before and after disposal of

Ridgefield, Connecticut, USA
hazardous waste is a new and growing field for oilfield technology.
Bob Engelman Whether the problem is identifying and monitoring contaminated
John Fruchter
Bill Shipp layers in the subsurface or characterizing a potential waste
Battelle Pacific Northwest
repository, techniques designed for hydrocarbon exploration and
National Laboratory
Richland, Washington, USA production are finding applications in a new environment.

Robert Jensen
Rick Lewis
Englewood, Colorado, USA
Hugh Scott
Sugar Land, Texas, USA
Steve Trent
Bechtel Hanford Inc.
Richland, Washington
For help in preparation of this article, thanks to Andrew
Bowden and Robert Chaplow, United Kingdom Nirex
Limited, Harwell, Oxford, England; Kevin Dodds,
Schlumberger Wireline & Testing, Aberdeen, Scotland;
Philippe Guerendel, Rachel Kornberg, Richard Parker,
Simon Robson and Pieter van der Groen, GeoQuest,
Gatwick, England; Michael Kane, Schlumberger-Doll
Research, Ridgefield, Connecticut, USA; Greg Kubala,
Schlumberger Limited, Sugar Land, Texas, USA; and Tim
Scheibe and Steve Yabusaki, Battelle Pacific Northwest
National Laboratory, Richland, Washington, USA.
FMI (Fullbore Formation MicroImager), Formation
MicroScanner, Litho-Density and UBI (Ultrasonic Bore-
hole Imager) are marks of Schlumberger.
Keeping the Earth’s ecosystem a safe and
healthy place to live and work is a chal-
lenge today and will remain so into the 21st
century. For thousands of years, humanity
has sought protection from natural hazards
and defense against predators and foes.
Ironically, some activities designed to pro-
vide such protection—the energy, defense
and medical industries—also threaten us
with another danger: hazardous and
radioactive waste. To safeguard against
unnecessary exposure to these wastes, gov-
ernments now regulate the treatment, dis-
posal and storage of industrial leftovers. A
whole new industry, for management of
environmental protection, has sprung up to
defy this new danger with technologies tai-
lored to the specific substances and hydro-
geological settings in question.
The goals of environmental management
are many: to minimize generation of and
exposure to hazardous waste; to dispose of
waste in a manner that meets government
regulations and community standards; and
to assess, monitor and remediate damage
caused by disposal gone awry. Several disci-
plines, including agriculture, soil and
groundwater engineering, and hydrocarbon
exploration and production (E&P), are con-
tributing measurement methods, modeling,
and treatment and containment technolo-
gies to achieve these goals. This article
examines how oilfield technologies are
helping to identify and characterize zones
where waste or other substances have accu-
mulated after leakage or disposal and to
assess potential subsurface repositories for
hazardous and radioactive waste. (For defi-
nitions of types of waste see “Hazardous
Waste and Radiation Basics,” page 48.)
Successful application of oilfield technol-
ogy for disposal site characterization and
monitoring is based on understanding the
differences and similarities between the E&P
industry and the environmental manage-
ment industry. The most significant differ-
ence is in the economic incentives in the
two fields. In the oil industry, the driving
force behind use of technology is increased
oil recovery, leading to increased profits. In
contrast, in environmental management the
driver is cost-effective protection of people
and the environment in compliance with
government regulations. In the latter case,
simple, inexpensive technologies are usually
chosen over sophisticated, expensive ones.
Soil and groundwater samples are more rou-
tine data sources than borehole logs; in
some environmental management firms
“logging” is a term generally associated with
forestry, not with boreholes.
In spite of the emphasis on low-cost solu-
tions, the environmental management
industry worldwide has annual revenues
estimated at $250 billion. Demonstrating
the value of E&P technologies is a first step
to ensuring that site evaluation and restora-
tion efforts harness the most effective tech-
nologies available.
The main similarity between the oilfield
and environmental management industries
is the need to describe subsurface fluid
behavior through cost-effective technologies
( next page ). This implies accurate

44 Oilfield Review
■ Characterization of soil, rock and fluids for managing
environmental protection. As in the oil field, description
of subsurface layers and prediction of fluid movement
require cost-effective technologies.
45
measurement of fluid constituents and pre- CANADA
diction of their movement with time. Other UNITED STATES
similarities—such as the actions taken to
mobilize, extract or contain fluids—depend
Spokane
on the disposition of fluids and the desired Seattle
results. In the oil field, the goal is to extract WASHINGTON
the maximum percentage of hydrocarbons
while keeping injected or connate water ■ Hanford Site on
away from producing wells. In the arena of Richland the banks of the
N
waste disposal, characterization is required Columbia River in
for design of repositories, so that hazardous the state of Wash-
materials are kept out of aquifers and reser- ington, USA, chosen
in the 1940s for its
voirs that may communicate with the sur- remote location and
face ecosystem. Whenever water supplies ample supply of
are threatened to unacceptable levels, N Reactor cooling water.
action is necessary to extract, treat or iso-
late the contaminants.
In both industries, characterization of the

Co
lu m
subsurface is essential. The following case

bi
studies show how technologies designed for Ri

a
ve
understanding hydrocarbon reservoirs are r
being advanced to address this need.

Hanford Site, Washington State, USA

An example of successful application of E&P
technology for environmental management
is the delineation and monitoring of radioac-
tive and hazardous wastes in the sedimen-
tary layers below the Hanford Site in Wash-
ington, managed by the US Department of
Energy (DOE) in the northwestern USA.
Built on the banks of the Columbia River in
the 1940s, Hanford is the site of the world’s
first full-scale nuclear reactor and spent-fuel
reprocessing plant, designed to produce plu-
tonium (239Pu) for atomic weapons (above
right ). The location was chosen for its
remoteness, ease of defense, mild climate,
ample supply of electricity and availability of
water for cooling the reaction process.
Hanford Site
Boundary
er
a Riv
kim Richland
Ya
0 km 10
0 miles 10
Plutonium was produced at Hanford from cal separation process, or reprocessing, was
1944 to 1987. At one time up to nine reac- placed in single- and double-shelled car-
tors and five reprocessing plants were active bon-steel tanks ( below ). The tanks were
at the site. All are still present, though inac- encased in cement and buried in “tank
tive and scheduled for decontamination and farms” located on the Hanford site central
decommissioning. High-level radioactive plateau, approximately 10 miles [16 km]
waste from the plutonium extraction chemi- west of the Columbia River and about 200 ft

Processed Filtered
Liquid level gauge
air intake air exhaust
Leak
detection pit

Primary
tank
Secondary
Liquid tank
Reinforced
concrete

Sludge Carbon Annular space

steel for leak detection
■ Tanks for storing waste from nuclear reactors. Of the 177 tanks
at Hanford, 28 are double-shelled and 149 are single-shelled.
The carbon-steel tanks (left) are encased in cement and buried
(right). Sixty-seven of the single-shelled tanks have leaked.

46 Oilfield Review
 Depth, ft
HNGS Cobalt-60-in. pCi/g HNGS Cesium-137-in. pCi/g
0.1 10,000 0.1 10,000

10

20

30
<

40
< <
< <
50
< <
Cobalt-60 Activity from 222-S Laboratory Cs-137 Activity from 222-S Laboratory
60 < <
< <

■ Comparisons between laboratory (dots) and HNGS-logged values (solid curve) for cobalt, 60Co, (left) and cesium, 137Cs (right). At
radioactivity concentrations less than about 2 pCi/g, laboratory measurements are not sensitive enough to give quantitative read-
ings, while the HNGS maintains sensitivity.

[60 m] above the water table. Tank waste
includes about 250,000 metric tons of
mostly sodium nitrite and nitrate plus 215
million Curies of mostly cesium (137Cs) and
strontium (90Sr) and smaller quantities of
various metal hydroxides. With time, how-
ever, 67 of the 149 single-shelled tanks
(none of the 28 double-shelled tanks) have
leaked or are suspected to have leaked. It is
now the responsibility of the US DOE to
clean up the site, and the DOE has awarded
the environmental restoration contract to
Bechtel Hanford Inc. This contract includes
cleanup of groundwater and soil and
decommissioning of surplus facilities.
If the exact locations of leaked contami-
nants are not known, they must be identi-
fied cost-effectively before restoration is
undertaken. Though several large plumes of
contaminated groundwater exist beneath
Hanford, most of the waste intentionally or
unintentionally released into the subsurface
remains in the vadose zone—the partially
saturated layers above the water table.
Observation boreholes drilled in the tank
farms and surrounding areas are air-filled
and steel-cased with no cement and no
rathole. They occasionally contain radioac-
tive contamination. Ordinary cased-hole
logging tools and interpretation techniques
are ineffective in this environment. Under a
cooperative research and development
agreement (CRADA)—initiated and funded
by the Department of Energy—with groups
at the Hanford site, Schlumberger modified
existing wireline logging tools and con-
ducted extensive computer and laboratory
modeling to calibrate tool responses to the
unique operating conditions.1
■ Unsaturated matrix. Air in the pore
space requires new models for deriving
porosity from density measurements.
The initial effort focused on modification
of natural gamma ray spectroscopy and
neutron porosity tools. The natural gamma
ray system was aimed at efficient identifica-
tion of gamma-emitting waste and litho-
logic correlation. Neutron porosity logs
were to be used for characterization of
hydrogeologic properties. Specifically, they
would help determine moisture content.
Unlike in fully saturated environments,
moisture content in unsaturated sediments
is not equal to porosity (above ).
Data quality objectives (DQOs) were
established for each of the measurements,
based on requirements set by a CRADA
committee consisting of data users, technol-
ogy providers, state and federal regulatory
agency representatives and service com-
pany representatives. For natural gamma
measurements, the DQOs included opera-
tion in high-level radioactive fields and
identification of man-made radioisotopes of
cobalt (60Co) and cesium (137Cs) in addition
to naturally occurring radioisotopes tho-
rium (Th), uranium (U) and potassium (K)
measured during oilwell logging. All con-
centrations were to be reported in pico-
Curies per gram (pCi/g).
To achieve these objectives, the Hostile
Natural Gamma Ray Sonde (HNGS), which
measures the energy of incident gamma
rays with two scintillation detectors, was
adapted to allow interpretation of high lev-
els of radioactivity and man-made
isotopes.2 This required modifications to the
tool gain regulation system and new meth-
ods for constructing elemental standards for
calibration. No 60Co- or 137Cs-rich test for-
mations or API test pits were available, as
are used for Th, U and K. Calibration was
achieved through measurements in scaled
models of field conditions built in a labora-
tory and combined with Monte Carlo simu-
lations for normalization. The sonde was
also turned upside down so that the detec-
tors—now 4 ft [1.3 m] from the bottom
instead of 40 ft [13 m]—could log more of
the borehole. Acceptable comparisons are
found between HNGS-logged values and
radioactivity of samples evaluated in the
laboratory (top).
An additional objective of the CRADA was
to measure formation moisture. Without
human intervention or natural disturbance,
such as earthquakes, waste is assumed to
stay where it is placed, discharged or leaked
as long as it is not dispersed by fluid flow. If
1. Ellis DV, Perchonok RA, Scott HD and Stoller C:
“Adapting Wireline Logging Tools for Environmental
Logging Applications,” Transactions of the SPWLA
36th Annual Logging Symposium, June 26-29, 1995,
Paris, France, paper C.
2. For background on the HNGS: Flanagan WD, Bram-
blett RL, Galford JE, Hertzog RC, Plasek RE and Ole-
sen JR: “A New Generation Nuclear Logging System,”
Transactions of the SPWLA 32nd Annual Logging
Symposium, Midland, Texas, USA, June 16-19, 1991,
paper Y.

Autumn 1996 47
liquid, or moisture, is not present, contami- Cutaway View of Injection Experiment
nants are immobile, and so less likely to
■ Tracer-injection
move away from the site. The DQO set experiment at Han-
accuracy standards for measuring moisture ford. Cesium and
in formations of up to 40% porosity with strontium injected
12.5%, 30% and 50% of the porosity filled in the central hole
(left) in 1980 were
with water. Standards were also set for mea- monitored using
surements in beds of different thicknesses Schlumberger
and through different casing size. The nuclear probes in
Accelerator Porosity Sonde (APS), a neutron observation holes
porosity tool that emits neutrons and detects (red) 15 years later
(below).
them again after they have interacted with
hydrogen in the formation, was chosen.3
No calibration standards existed for log- Well Numbering Scheme
ging moisture content, so these had to be Injector well
1m
designed and built. By carefully mixing dry
pure-quartz sand, SiO2, and aluminum tri-
hydrate, Al(OH3)—a dry material containing
a known amount of equivalent bound
water—and presettling the mixes on vibrat-
ing tables, well understood moisture models
were created.4 Precision and accuracy of
the moisture measurements were demon- E-7 E-5 E-3 E-1 A-1 A-3 A-5 A-7
strated to meet or exceed required specifica-
tions in tests on the specially designed cali-
bration standards.
The natural gamma and moisture measure-
ments were successful, so the CRADA was
extended to develop a through-casing den-
sity measurement for porosity determination.
The Litho-Density Sonde, an openhole den-
sity tool, was altered for this purpose. This

Hazardous Waste and Radiation Basics

Every country uses different terminology to cate-
gorize hazardous waste, but the two main types
contrasted here are nonradioactive and radioac-
tive. Nonradioactive hazardous waste means
chemicals and materials that are toxic, corrosive,
reactive or ignitable. These may include hydro-
carbons, explosives, asbestos, metals, solvents,
medical wastes, pesticides and polychlorinated
biphenyls—PCBs. In the groundwater, many of
these contaminants are considered hazardous
at levels several parts per billion (ppb). In the
US, for example, groundwater limits in ppb
are 0.5 for PCBs, 5 for benzene, 5 for carbon
tetrachloride and 15 for lead. Limits in soil
may be different.
1. “Committed to Results: An Introduction to DOE’s Environ-
mental Management Program,” Office of Environmental
Management, US Department of Energy, Washington DC,
USA, Document DOE/EM-0152P (April 1994).
2. For more on radiation effects and oilfield practice: Bram-
blett D, Kurkoski P and Racster C: “Advancing Wellsite
Radiation Safety,” Oilfield Review 2, no. 4 (October 1990):
13-23.
Within the radioactive waste category, there
may be further distinctions depending on the con-
centration levels of radioactive material and the
half-life of the radionuclides. Low-level radioac-
tive waste, such as clothing, equipment and soil
contaminated with minute concentrations of
radioactivity, or medical and oilfield tracers that
decay rapidly with time, constitute the largest
volume of radioactive waste. These wastes are
generally considered safe if stored in the shallow
subsurface. High-level radioactive waste, such as
spent fuel from nuclear reactors, must be iso-
lated from the surface for thousands of years.
Standards of isolation differ from country to
country. In the United States, the Environmental
Protection Agency standards state that a high-
level waste repository may pose no greater risk
than unmined uranium ore from which the high-
level waste was produced.1
Radiation Basics
The four principal types of nuclear radiation are
alpha particles, beta particles, gamma rays and
neutrons. Alpha and beta particles are both
charged, so they are easily stopped by electro-
static forces in matter. An alpha particle is a
helium atom nucleus, and can be stopped by
paper, clothing or skin. Beta particles—high-
energy electrons—can be stopped by a thin sheet
of metal or wood, but will penetrate water or
skin. Gamma rays and neutrons are highly pene-
trating and require dense shields, such as lead or
thick concrete. Natural radiation from the sun and
earth makes up about 82% of the average per-
son’s exposure to radiation. The rest typically
comes from medical sources, such as X-rays.2

48 Oilfield Review
 ■ Nuclear log data
density logging sonde irradiates formations
1800 interpolated along
with medium-energy gamma rays that col- a transect of the
lide with electrons in the formation. With injection test vol-
1600
each collision, a gamma ray loses some of its ume. The matrix
energy and continues. The reduced-energy bulk density logs
1400 (top) from the Litho-
INJ
gamma rays that reach the detector are Density tool chart
counted as an indication of the electron den- 1200 the natural sedi-
sity of the formation. mentary layering of
The Litho-Density algorithm for measuring 1000 the Hanford forma-

cm
density was calibrated to compensate for tion. The gross
800 gamma ray logs
casing and the air gap between casing and Matrix Bulk Density
(middle) from the
g/cm3
formation, since there is no annular fill mate- 600 HNGS show the
rial in the Hanford holes. Laboratory experi- same layering, as
2.03 400 well as the anoma-
ments were conducted at the Schlumberger 1.9
1.8 E-1 A -1
A-3 A -5 A -7
1.7 E-7 E-5 E-3 lously high levels of
Environmental Effects Calibration Facility in 1.6
1.5
200 134Cs radioactivity
Houston, Texas, USA to verify that the new 1.4
1.3 near the injection
1.2 00
algorithm performed within the data quality 1.12 N 00 1200 14 well. Moisture logs
0 800 10
400 60 m
objectives set by the CRADA committee. c (bottom) from the
The modified measurement techniques APS show no
anomaly associated
have been run commercially in 75 bore- with the injection,
holes at Hanford, including an injection 1800 indicating that the
experiment designed to predict subsurface formation has
transport models (previous page ).5 Radioac- reequilibrated to its
1600
preinjection mois-
tive tracers 134Cs and strontium (85Sr) were ture profile. This
injected in a central well in 1980. These are INJ
1400 means the injected
short-lived isotopes of the same elements radionuclides are
whose long-lived isotopes, 137Cs and 90Sr, 1200 immobile.
have contaminated portions of the Hanford
1000

cm
subsurface.6 In 1995, the three modified
nuclear probes were run in eight holes 800
Gross Gamma
along a transect of the injection test volume
AEI 600
(right ). Density logs from the Litho-Density
tool chart the natural sedimentary layering 76.4 400
A-5 A-7
of the shallow unconsolidated formation. E-1 A-1 A-3
72.0
E-7 E-5 E-3
68.0
64.0
200
Gamma ray logs from the HNGS show the 60.0
56.0
52.0
same layering, as well as noticeably high 48.0
44.0 1400
N 00 1200
0 800 10
40.7
levels of 134Cs radioactivity near the injec- 400 60 m c
tion well. Moisture logs from the APS show
no anomaly associated with tracer injection.
The logging was able to locate the 15-
year-old cesium plume and verify that the 1800
only detectable 134Cs remaining from the
injection was located near the injection 1600
point. No strontium was detected. During
the 15 years since the injection, 85Sr had 1400
INJ

3. For background on the APS: Flanagan et al, reference

1200
2; and Scott HD, Wraight PD, Thornton JL, Olesen J-R,
cm

Hertzog RC, McKeon DC, DasGupta T and Albertin IJ: 1000

“Response of a Multidetector Pulsed Neutron Porosity
Tool,” Transactions of the SPWLA 35th Annual Log- Volumetric Fraction 800
ging Symposium, Tulsa, Oklahoma, USA, June 19-22, of Water
1994, paper J. % 600
4. Engelman RE, Lewis RE, Stromswold DC and Hearst
JR: “Calibration Models for Measuring Moisture in 22.5 400
Undersaturated Formations by Neutron Logging,” 20.0 A -3 A -5 A -7
E-1 A -1
Pacific Northwest National Laboratory Report PNL-
18.0
16.0 E-7 E-5 E-3
14.0 200
10801 (1995). 12.0
10.0
8.0
5. Fayer MJ, Lewis RE, Engelman RE, Pearson AL, Murray 1400
00 1200
6.0
N
CJ, Smoot JL, Randall RR, Wegener WH and Lu AH:
4.53
0 800 10
400 60
“Reevaluation of a Subsurface Injection Experiment cm
for Testing Flow and Transport Models,” Pacific North-
west National Laboratory Report PNL-10860 (Decem-
ber 1995).
6. The half-life of 134Cs is 2.06 years and that of 85Sr is
64.8 days—short compared to 30.2 years for 137Cs
and 27.7 years for 90Sr.

Autumn 1996 49
 LDS Short-Spaced Detector Counts (Source Removed)
1/s
Depth, ft
100
HNGS Gross Gamma
100 API
10
20
30
40
50
60
■ Natural gamma and Litho-Density logs revealing a highly
radioactive, contaminated zone less than a foot thick at the Han-
ford N reactor site. Natural gamma readings taken by the HNGS
ran off scale at 200,000 API units at a depth of 10 ft [3m], but the
Litho-Density tool delineated the thin contaminated layer.
decayed to less than 0.01% of the original
amount, and 134Cs to less than 1%. The fact
that after 15 years Cs was detected only
near the injection point is an indication of
the high sorption potential in the sediments.
In another part of the Hanford facility, at the
N reactor site near the river, Bechtel Hanford
has recently completed a limited field investi-
gation of two radioactive liquid waste dis-
posal facilities (LWDFs). An LWDF is an engi-
neered soil-column waste disposal system
designed for reactor effluent disposal. Starting
in 1963, demineralized Columbia River cool-
ing water was run through the N reactor.
Sometimes the cooling water became con-
taminated with radioactive waste products.
This water was discharged into LWDFs up to
millions of gallons per day. From there it per-
colated through surface layers, and most of
the contaminants were retained in the soil
column. Tritium and strontium were among
those contaminants that reached the ground-
water and eventually contaminated some
river bank seeps—known as the N-springs. A
second facility was constructed farther from
the river in 1985, but contamination in the
groundwater and N-springs remained. Cur-
rently a “pump and treat” process has been
implemented to protect the Columbia River
from contaminant migration. Contaminants
are pumped, or extracted, for treatment at the
surface. Bechtel and Battelle Pacific North-
west National Laboratory (PNNL) are investi-
gating permeable barriers for long-term pro-
tection of the river (see “Containing
Contaminants,” next page).
50
10,000,000
10,000,000
■ Hanford worker on
duty in the highly
restricted zone
around a tank farm.
To design a remediation program, Bechtel
must locate and quantify the volume of con-
taminated soil and groundwater in a cost-
effective way. Usually, subsurface samples
are collected and analyzed to help charac-
terize the contaminated volume. The use of
well logs enhances the characterization by
giving continuous readings with depth,
reduces worker exposure to contaminants,
and helps keep costs down. In one area, the
cost of a multimillion-dollar characteriza-
tion program was cut in half by optimizing
the drilling and sampling plan and supple-
menting it with logs.
The logs revealed a highly contaminated
zone less than a foot thick (above left). Nat-
ural gamma readings taken by the HNGS
ran off scale at 200,000 API units and were
unable to resolve the bed thickness, but the
Litho-Density tool without the logging
source maintained gain and delineated the
thin contaminated layer.
The integration of sample analysis and log
data is providing a quantitative measure of
the concentration and thickness of the con-
taminated layer. The results will be used by
Bechtel engineers to assess the feasibility of
restoring the environment near the N reactor
to a condition to be determined by the DOE
and state and federal regulatory agencies.
Safe Logging at Hanford
Working near hazardous waste requires pre-
cautions and special training to assure
worker health and safety. At contaminated
sites, access to all areas is highly restricted
and monitored (above).
Training required for workers at Hanford
and other US environmental management
sites includes a 40-hour Occupational
Safety and Health Administration course,
consisting of instruction on hazardous
materials, cardiopulmonary resuscitation
and emergency first aid, and training on
use of respirators and special personal pro-
tective equipment (PPE).
There are four levels of PPE, from A for the
most stringent requirements to D for the
least. Level D, reminiscent of oilfield opera-
tions, requires steel-toed shoes, coveralls,
eye protection, gloves and hard-hat. Level C
requires chemical-resistant clothing and
boots, inner and outer chemical-resistant
gloves and a full-face air-purifying mask with
specially designed filters to remove antici-
pated contaminants. For level B, the highest
PPE level required at Hanford, impermeable
suits and supplied-air respirators are
required (next page, top left). The respira-
tors— connected to portable air tanks or to
stationary air tanks with umbilicals—are
Oilfield Review
 zzzzzz







yyyyy
,,,,,
|zzzzzz







yyyyy
{{{{
||||
,,,,,


{ |
■ Dressed for work. Personal protective
equipment (PPE) level B is required for this
worker lowering a video camera into a
hole at Hanford. The worker must be
qualified through special safety training
to wear a respirator and to operate equip-
ment near hazardous materials.
necessary for work at Hanford where the
exact combination of contaminants to be
encountered is unknown. Level A, an abso-
lutely impermeable “moon-suit,” is required
when there is risk of contaminants that can
be absorbed through the skin.
On-site health and radiation protection
technologists caution workers on all aspects
of job safety. Typically the biggest health risk
to Hanford workers is overheating brought
on by the cumbersome protective equipment
and the high summer temperatures—in
excess of 100°F [38°C]. The technologists’
role ranges from reminders about bringing
water to drink when it’s hot out and where
workers can sit during lunch breaks, to mon-
itoring tools for contamination as they come
out of the hole. Any tool found with the
smallest amount of contamination is wiped
clean, and the contaminated cloth disposed
of in an impermeable barrel stored on site.
Field crew at the Schlumberger Wireline &
Testing division in Bakersfield, California,
USA who work at the Hanford site, and
those at the division in Charleston, West Vir-
ginia, USA—in preparation for work at the
Savannah River site in South Carolina,
USA—have completed training required for
logging hazardous waste sites.7
(continued on page 54)
7. Savannah River, near Aiken, South Carolina, is a
DOE-managed site where tritium was produced.
Autumn 1996
Containing Contaminants
Reagent injection
Liquid waste disposal facility
Mixed waste
Vadose zone
Uranium Technetium
Saturated zone
Chlorinated solvents
Confining layer
■ Injectable permeable barrier concept. Contaminants flow through a volume of injected treatment material
that acts as a subsurface filter.
Current practice for addressing a subsurface vol-
ume of liquid hazardous waste, or plume, is to
pump and treat the liquid. Once the contaminated
liquid is located, it is pumped out and treated
to remove or neutralize the contamination. In
some cases the treated water is reinjected.
Pumping continues until the incoming liquid is
contaminant free, at which point the site may be
considered restored—but subject to further
monitoring. This process may take years.
When contaminants attain parts per billion con-
centrations, often the process of treatment
reaches an asymptotic level.
This technique may have some merit in homo-
geneous formations, but inhomogeneities—
the bane of the oil and gas reservoir—also
plague environmental management sites. In an
inhomogeneous formation, pumping draws liquid
only from high-permeability zones. In cases of
large permeability contrast, it can be easier to
pump liquid from a high-permeability zone miles
away than from a low-permeability zone only a
few feet away. Pumping may continue until water
comes out clean, but a few months later, moni-
toring may indicate that the water is contami-
nated again due to the desorption of contami-
nants from the soil.
Rather than pumping contaminated liquids to
the surface, scientists at Battelle Pacific North-
Borehole
Permeable
Chromate treatment
Treated waste
barrier
west National Laboratory (PNNL) are capitalizing
on the dispersive nature of the contaminants to do
part of the work. Treatment is accomplished by
creation of a permeable barrier—also called a
permeable reactive wall—to treat the contami-
nated liquids as they pass through.
Permeable barriers can be constructed by at
least two different methods. The first, and sim-
plest, is a ditch filled with gravel or absorbent or
chemically reducing material through which
waste liquids are drained. Gravel ditches are
known as trenches, and typically are used to treat
solvent-contaminated shallow groundwater by
aeration. Absorbent minerals such as zeolites,
which act like “kitty litter,” can trap some spe-
cific contaminants. Reducing agents, such as
finely ground particles of iron, are sometimes
effective for treating some contaminants, espe-
cially chromate, uranium and chlorinated sol-
vents. Chemically reducing treatments work by
reducing the valence state of the contaminant.
This reduction in valence destroys organic con-
taminants, such as chlorinated solvents, and
causes certain metallic contaminants to precipi-
tate, trapping them in the filter.
If the contaminants are already deep below the
ground surface, conventional trenching methods
will not be able to reach them. Researchers at
PNNL are investigating a second category—the
injectable permeable barrier—that acts as a sub-
surface filter to treat spreading contaminated liq-
uids (above). Injectable permeable barriers have
51
 1000 ■Laboratory experiment testing removal of chro- been tested in the laboratory and in pilot-scale
Influent mate contamination. Soil reduced, or treated, with field experiments at Hanford. In addition to
sodium dithionite was then flushed with contami-
nated liquid containing chromate. Concentration of absorbent filters and chemical-reducing materi-
800
chromate in the contaminated water (influent) was als, iron-reducing bacterial treatments—
held at more than 800 parts per billion (ppb). More
than 120 pore volumes of liquid had passed before similar to those applied in the oil field to treat
Chromate, ppb

600
there was evidence of chromate in the liquid filtering sour wells—have been tested.
through the treated soil (effluent).
Laboratory tests conducted on soils reduced
400
by one chemical agent, sodium dithionite,
indicate a strong ability to treat chromate-con-
200
taminated liquid (left). Soils treated with dithion-
Effluent
ite were able to filter up to 120 contaminated
0
0 60 120 180 240
pore volumes.
Pore volumes A field experiment monitored the same reac-
tion, but at full scale (left). Dithionite was
■Hanford chromate- injected to reduce the iron in a volume of the sub-
H5-13 removal field experi- surface, then chromate was injected. Chromate
ment. Sodium dithionite
was injected in a layer in levels were sampled in monitor wells surround-
the central well to treat ing the injector out to a radius of 90 ft [27 m].
H5-12 H5-14 the shaded volume.
Over the course of more than ten months, in wells
Chromate was then
Groundwater injected in the central monitored within 30 ft [9 m] of the injector, no
flow direction
well and chromate con- chromate was detected (below).
H5-3p H5-4p H5-1B
H5-5p H5-11 centrations were moni-
tored in 10 nearby wells. Similar tests are under way for examining the
ability of the technique to handle greater time
H5-g H5-10
Injection well and length scales. A current experiment is using
Transect Treated volume a line of injector wells to treat a 2000-ft [608-m]
wide chromate plume, already identified on the
0 10 20 30 Hanford site.
Scale, ft N
H5-15
Modeling
Modeling the distribution of fluids and contami-
80 80 nants is a vital part of planning any containment
80
February 27, 1996 or treatment effort. As in the case of oil and gas

70 March 26, 1996 reservoirs, small-scale variations in formation

70
May 7, 1996 properties can have large-scale effects on ulti-
Detection limit = 8 ppb mate fluid distribution.
60 60
60 Scientists at PNNL are simulating fluid and con-
Chromium concentration, ppb

taminant flow through small but realistically

50 50
50 complicated 3D volumes in hopes of understand-
ing flow in larger volumes. A 3D cube represent-
40 ing part of a complexly layered river sand has
been constructed based on field observations on
30 a meandering river (next page, bottom). Migration
30
of a nonreactive tracer through the cube has been

20 20 20 modeled at a series of five time steps (next page,

20
top).1 Tracer migration is constrained by low-per-
meability features in the model volume.
10
10

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0
H5-15 H5-3p H5-4p H5-5p H5-9 H5-10 H5-11 H5-1B H5-12 H5-13

■Chromate levels monitored in wells surrounding the injector. Wells within 30 ft of the injector showed no 1. For a video rendition see the PNNL web site:
chromate for more than ten months after injection. http://etd.pnl.gov:2080/EESC_public/vis_examples/
scaling/part.mpg

Oilfield Review
Time Step 1 Time Step 2 Time Step 3

Time Step 4 Time Step 5

■Simulated nonreactive tracer transport through the
3D model at five different time steps. Tracer particles
were released as an instantanous pulse in a plane at
the rear of the cube (white line), and their movement
is constrained by the low-permeability layers. Using
massively parallel computers and particle-tracking
methods, transport was simulated through the model
composed of nearly 17 million grid cells. Further
description of these studies—including an animation
of the particle plume—is available on the world-wide
web at:
http://etd.pnl.gov:2080/EESC_public/vis_examples/
scaling/scaling.html.

■Three-dimensional
representation of per-
meability distribution in
a complexly layered
river sand. Yellow-green
represents high perme-
ability and blue repre-
sents low permeability.
The small-scale sedi-
mentary features are
based on field observa-
tions on a meandering
river in Indiana, USA.
Development of this
model is described in
Scheibe TD and Frey-
berg DL: “Use of Sedi-
mentological Informa-
tion for Geometric
Simulation of Natural
Porous Media Struc-
ture,” Water Resources
Research 31, no. 12
(1995): 3259-3270.

Autumn 1996 53

■ Sellafield—concept for a UK national facility for solid intermediate-level and low-
level radioactive waste. Geological and hydrologic investigations are under way to
determine whether the site is suitable as a deep repository.
Characterization Before Storage
A primary lesson learned from the restora-
tion of sites contaminated with hazardous
materials is that future waste storage facili-
ties must be engineered with care. Countries
that rely on nuclear energy are taking a hard
look at how to manage the inevitable waste
products. Most countries plan deep under-
ground repositories, and the search contin-
ues for the best locations.
The goal is to isolate hazardous material
from the surface for tens of thousands of
54
years—no small task. The subsurface vol-
ume selected should be an effective shield,
remain tectonically stable, reduce contact of
waste with groundwater and air, and con-
tain no useful mineral resources that would
tempt human intervention. The storage facil-
ity should allow for retrieval of the waste at
such time that technologies become avail-
able for treatment. Some countries, includ-
ing Belgium, Finland, France, Germany,
Sweden, Switzerland, the UK and the USA
have identified locations that appear to have
many of the desired qualities.8
In the UK, United Kingdom Nirex Limited
(Nirex) is responsible for providing and man-
aging a national facility for underground
storage of solid intermediate- and some low-
level radioactive waste. Geologic and hydro-
logic investigations have been in progress
around Sellafield in West Cumbria, England
since 1989, to determine whether a site
adjacent to the existing nuclear establish-
ment is suitable for a deep repository (left).9
Groundwater is the most likely medium by
which radioactive waste from a repository
could return to the surface. Controls on
groundwater flow include driving forces,
such as gravity and salinity, and also the
geometry of rock units and properties of the
rock mass, such as permeability and fracture
characteristics.10
The building blocks for the geological site
characterization consist of 27 deep bore-
holes, some down to 1900 m [6200 ft] and
the majority with continuous core; 2000 km
[1250 miles] of 2D seismic data and 8000
km [5000 miles] of airborne geophysical
data. Hydrogeological testing and ground-
water sampling provide additional informa-
tion on flow properties of the volcanic for-
mation targeted for study.
A volume of the Borrowdale Volcanic
Group at about 650 m [2133 ft] below sea
level, is under consideration as an under-
ground laboratory, or rock characterization
facility (RCF). The RCF is being constructed
to permit detailed examination and testing
of rock at depth before selecting a final
repository design and location. Investiga-
tions have shown that groundwater flow
within the volcanic formations occurs
through a limited number of fractures, and
current work focuses on characterizing
hydrogeologically significant fractures.
Many techniques developed to identify
and characterize fractures in the oil field are
playing the same role at Sellafield. Integra-
tion of several measurements at many scales
is essential for determining which fractures
are open and connected. At the finest scale,
electrical conductivity logs of the borehole
8. “Going Underground,” United Kingdom Nirex Lim-
ited, Harwell, Oxford, England (November 1993).
9. Chaplow R: “The Geology and Hydrogeology of
Sellafield: An Overview,” Quarterly Journal of Engi-
neering Geology 29 (1996): S1- S12.
10. Michie U: “The Geological Framework of the
Sellafield Area and Its Relationship to Hydrogeol-
ogy,” Quarterly Journal of Engineering Geology 29
(1996): S13-S27.
11. For a description of the technique: Hornby BE, Luthi
SM and Plumb RA: “Comparison of Fracture Aper-
tures Computed from Electrical Borehole Scans and
Reflected Stoneley Waves: An Integrated Interpreta-
tion,” The Log Analyst 23, no. 1 (January-February
1992): 50-66.
Oilfield Review
wall, such Formation MicroScanner and FMI ■ Fracture orienta-
Fullbore Formation MicroImager scans, tion and aperture
show where fractures intersect boreholes. visualized with
the UBI Ultrasonic
These borehole images help identify and Borehole Imager
orient fractures, but may also be combined tool. Like thread
with other data, such as sonic waveforms, to winding around
calculate fracture aperture.11 a spool, ultrasonic
scans of the bore-
Fracture orientation and aperture may also hole track hole
5320
be interpreted on borehole scans made by rugosity in detail.
the UBI Ultrasonic Borehole Imager tool. In this image,
The UBI tool takes 180 transit-time samples breakouts—stress-
around the borehole circumference for related damage in
5321
the plane of least
every 0.2 in. [0.5 cm] the tool moves horizontal stress—
uphole. Transit times converted to distance are seen on oppo-
yield an image that looks like a mold of the site sides of the
5322 borehole.
borehole. Where fractures intersect the

Depth, ft
borehole, extra rugosity may occur, and the
3D UBI image allows visualization of these
geometries. Stress-related borehole damage 5323
may also occur in the direction of least hori-
zontal stress and manifest itself as hole
enlargement, or breakouts, on opposite
5324
sides of the borehole (right). The breakout
direction is an important clue to the stress 5
regime present in the Sellafield area.
Looking beyond the wellbore, borehole 5325 0
acoustic reflection surveying (BARS) offers a
-5
greater range of penetration for fracture
0 -5
tracking. Similar to surface seismic surveys 5
but recorded in the borehole, this new tech-
nique can image fractures extending as far
as 30 ft [9 m] away from the borehole
(below). In other situations, BARS images

P Waves Converted Waves

3700 3700 ■ Borehole acoustic
reflection surveys
(BARS) for tracking
fractures far from
the borehole. In
this case, sonic-fre-
quency data from
3750 3750 a research proto-
type tool were
migrated to obtain
reflections of pri-
mary (P) waves
(left) and converted
Depth, ft

waves (right).
3800 3800 The center of each
image corresponds
to the wellbore
position (yellow).
Fractures identified
on borehole tele-
viewer images are
3850 3850 plotted in black
along the borehole.

3900 3900
-30 -20 -10 0 10 20 30 -20 -10 0 10 20

Distance from borehole, ft Distance from borehole, ft

Autumn 1996 55
may be obtained of fractures and other fea-
tures near but not necessarily intersecting
the borehole. These sonic waveform data
are acquired with multiple source and
receiver arrays on a single tool. For Nirex,
BARS data were recorded with a research
prototype tool; commercial introduction of
the service is scheduled for early 1997.
P-Wave Converted-Wave
Reflection Reflections
Receivers
P S
P yield an image of velocities in the inter-
P P
S grated interpretation (next page). For
Transmitter
■ Reflections recorded in a borehole
acoustic reflection survey. Primary (P)
waves (left) and shear (S) waves (right)
reflect at different angles and can be
used to image different features or por-
tions of the same feature.
■ Network of boreholes linked with crosswell seismic images. The
shaded “curtains” between pairs of holes signify the areas that
have been imaged with a seismic source in one hole and an array
of hydrophone receivers in the other.
56
The signals of interest—those from reflec-
tions—arrive after the primary (P) and
before the shear (S) waves that are typically
analyzed for borehole sonic slowness val-
ues (left). Borehole acoustic reflection sur-
veys are processed using seismic data-pro-
cessing algorithms, and yield both P and S
reflections. Shear waves help image fea-
tures at high angles to the borehole, and are
better than P waves at resolving some near-
wellbore features.
At a larger length scale, seismic images
have been obtained for several pairs of bore-
holes (below left). Seismic travel-time data
are acquired with a downhole seismic
source in one borehole and the Cross Well
Seismic Imager—an array of 16 hydrophone
receivers spaced at 4 m [13 ft]—in the
second one. Travel times are processed to
borehole region. Abrupt changes in veloc-
ity, represented by changes in color, indi-
cate discontinuities. Combined with other
log data, borehole-to-borehole correla-
tions become clear, reinforcing the inte-
example, fractures evident in one borehole
can be tracked across seismic images, and
linked with fracture indicators in neigh-
boring boreholes.
Incorporating flow information is vital to
an integrated interpretation. To identify sig-
400
Depth, m
800
1200
1600
■ Multipacker systems for long-term hydrogeologic monitoring
and testing. These systems have been deployed to depths of
1500 m [4920 ft] with up to 30 zones isolated in any one borehole.
nificant fractures, flow tests have been car-
ried out. Testing is conducted while the
drilling rig is on location, and with a
workover rig after drilling. Long-term moni-
toring and testing with permanent multi-
packer systems is continuing in several
boreholes (below right).12 These systems
allow liquid samples and pressures to be
analyzed at multiple levels in any one bore-
hole with only one tool position.
Hydraulic conductivity of the volcanic for-
mation is typically low, even measured over
fractured intervals. Most fractures intersect-
ing boreholes have no detectable flow, but
those that do are characterized and related
to structural features such as faults. As more
data are acquired and interpreted to help
build and test conceptual models, scientists
at Nirex will improve their understanding of
the controls on groundwater flow through
the subsurface volume.
Future Efforts
Many challenges await the environmental
management industry. Parties responsible
for hazardous waste sites and repositories
need to be able to make informed deci-
sions. Service providers help by acquiring
data and turning them into usable informa-
tion. The application of wireline logging to
site characterization and monitoring pro-
12. Sutton JS: “Hydrogeological Testing in the Sellafield
Area,” Quarterly Journal of Engineering Geology 29
(1996): S29-S38.
Pump
well
Oilfield Review
vides a comprehensive body of information
and early warning of possible problems. If Borehole 2 Cross-well Borehole 4
migration of contaminants is discovered, Formation
Sonic Velocity Tomogram Sonic FMI Images
then there is time for an informed decision Depth, m
Apparent
MicroScanner
Apparent
Depth, m
Dips Dips
on whether to remediate. If the choice is to Orientation BH 2 Offset, m BH 4 Orientation

N S N VDL 0 50 100 150 VDL N S N

do nothing—either because the problem 400
400
isn’t severe enough or more time is required
to develop the proper technology—the site
still must be monitored for regulatory pur- 450
450
2
poses and for continued decision making.
If action is required, adequate characteri-
500
zation of the site is necessary to build con- 500

ceptual models for simulating treatment and

optimizing treatment strategies. During 550
remediation, monitoring can provide valu- 550

able feedback that ensures success. And 2 3

once remediation has been achieved, fur- 600

600
ther monitoring is required. Geophysical 3

logging provides a means for permanent 4

650
and remote monitoring. 650 4
The environmental management industry
is growing at different rates in different 700
countries, and changing as it grows. Manag- 700

ing the vast amount of data requires special 5

5
skills and tools. The rapid pace at which 6
7
750
750
new technologies are introduced leads
some environmental managers to predict
800
that the industry is moving toward deferred 800
6
8
remediation. Full site characterization—
with sensitive detection of contaminant 7 850
migration before it exceeds regulatory 850

guidelines—allows time for solutions to

tough problems to be developed. Regula- 900
900
tions also change, usually becoming more
stringent and all-embracing. Information
8 950
from geophysical logging allows the design 950
9

of environmental management schemes 10

flexible enough to accommodate changes. 1000
Also predicted is a move toward perfor- 1000

mance-based compliance, in which govern- 11

ments regulate what the end results of com- 1050

1050
pliance and remediation should be, but not 12

how to achieve them. As the industry

evolves, environmental management firms 1100
1100

will continue to require technologies that Apparent

Dips of: VDL Flow Zone Flow Zone VDL
Apparent
Dips of:
solve problems in a cost-effective manner so Fault Waveform 1 Waveform 1 Fault
4200 5600 7000
that decisions can be made in a framework Pseudo-
Layering
15 35
0 to 10000 us 0 to 10000 us
.30 150.0 Pseudo-
Layering
of knowledge. —LS Conductivity
Velocity, m/sec
Conductivity
mho mho

■ Complete set of image data for one pair of boreholes—B2 (left) and B4 (right). Dipmeter
stick plots (track 1) indicate apparent dip of identifiable fractures (red sticks). A Forma-
tion MicroScanner image (track 2) shows disruptions at depths corresponding to the
stick plot. Sonic waveform displays (track 3) plot fracture indicators as breaks in the
continuity of vertical color stripes. The event at 650 m in B2 can be tracked across the
crosshole seismic image (center) to a similar signal in the sonic display of B4 at 840 m.
High-velocity (orange) and low-velocity (blue) zones indicate different lithologies. Lin-
ear discontinuities in color may be interpreted as fractures. Intervals in which flow is
detected in B2 (blue circles) and B4 (purple circles) are plotted on either side of the
cross-well image. Logs for B4 are in the reverse order of those for B2, with the exception
of an FMI image in B4 instead of a Formation MicroScanner image.

Autumn 1996 57