This article describes mechanical sand control methods and covers the following topics:

·Gravel packing techniques i.e. Internal Gravel Packing (IGP) and External Gravel Packing (EGP), and the various gravel packing systems presently available.

·Post-gravel pack operations and remedial treatments with special emphasis on through tubing techniques.

·Other mechanical sand control methods i.e. non-gravel packed screens and external casing packers.

·Special sand control applications including horizontal wells, fracpacking, and the Auger system.

Two main objectives were pursued when preparing this chapter:

 . To provide all the required information for designing a mechanical sand control completion,

·To promote a common and consistent approach to gravel packing within the Group.

The latter objective was more difficult to achieve as gravel packing is still being developed through extensive research and field trial programmes.

1 General gravel pack design considerations

1.1 What is a successful gravel pack?

Effective sand control was long considered to be the main measure of success for a gravel pack and current procedures allow this objective to be achieved with confidence for most applications. Some reduction in well productivity as a result of gravel packing has long been recognised, and accepted by many companies as being a price to be paid for sand control. However recent tests have highlighted that this damage, when expressed as Productivity Index reduction, may be of the order of 50-80%. Well damage of this magnitude is often unacceptable in terms of deferred production and energy losses in the production system.  

The requirements for a successful gravel pack are to reduce sand production to a tolerable level whilst minimising productivity impairment.The emphasis today is to optimise gravel packing methods with the prime objective of minimising productivity impairment. Despite all the efforts of the past the industry has been unsuccessful in uniquely identifying optimum gravel packing procedures.

Consequently this article can only:

·Provide a summary of current "best" practices.

·Give acceptable ranges for the main design parameters.

·Identify procedures and methods that offer the best scope for productivity enhancement.

·Encourage operators to challenge established practices and field test in a structured way new approaches to sand control.

1.2 Formation sand analysis

1.2.1 Sampling

Samples should be representative of the interval under consideration. The number of samples required is a function of the formation heterogeneity. The field geologist should be consulted to establish to what extent grain size distributions obtained from samples in a well can be extrapolated to other parts of the field.

Formation samples can be obtained from the following sources:

1.The best samples are obtained by coring the reservoir. Good core recovery can be achieved in an unconsolidated sand with rubber or fiberglass sleeved coring barrels, although some material will be washed away and lost when cutting the core. The number of sample points on a core should be established after a proper core description. Unlike many other core properties the sand grain distribution will not change during the aging of the core, hence old core material can be used.

2.Side wall samples can provide acceptable formation samples although they are small and often contaminated by mud cake. Care must also be taken with the crushing effect of this sampling method which may cause a bias towards smaller grain sizes.

3.Samples from drilling cuttings, bailer runs or separators should be treated with care as they may not be representative. Bailer samples will generally include the larger grain fractions and samples recovered at surface will be biased towards the smaller, more easily transportable sand grains. The origin of these samples in the reservoir is obviously unknown.

Formation samples need to be cleaned, thoroughly dried and separated into individual sand grains. This must be done carefully for partially consolidated sands to avoid breaking the individual sand grains and to ensure that only individual sand grains are present.

1.2.2 Grain size distribution

1.2.2.1 Sieve analysis

The sieve analysis of a sand sample consists in passing the sample through a series of sieves, allowing the sand grains smaller than each sieve's openings to fall to the next sieve.

The results of the sieve analysis are generally plotted as cumulative weight percent versus grain size (sieve opening) on a linear or logarithmic scale. This plot typically yields an S shaped curve.

An alternative plotting technique used by sedimentologists is to plot the cumulative weight percentages against the sieve opening size.

The advantage of this method is that a sample with a Gaussian distribution of sand particles will plot as a straight line (Fig. 681). This makes interpolation between data points easier.

1.2.2.2 Grain size distribution statistics

The main parameters to be derived from the sieve analysis are the median grain size and the uniformity coefficient.

a. Median

The median grain size (D50) is the particle size corresponding to the 50 percentile of the cumulative weight distribution from the sieve analysis. This parameter is directly used to select the gravel used for sand control.

b. Uniformity coefficient

The uniformity coefficient C is defined by the ratio of the 40 percentile grain diameter to the 90 percentile grain diameter.

C = D40/D90

This parameter gives a measure of the spread in the grain size distribution, in other words the sorting of the sample and is commonly used when selecting sand control measures. Schwartz used the following classification:

·C < 3: Uniform sand

·C > 5: Non uniform sand

The more poorly the formation sand is sorted, the more difficult it is to efficiently control sand production. Geologists use different parameters to characterise sorting but the meaning is the same

1.3 Gravel considerations

1.3.1 Gravel sizing

Several gravel sizing criteria have been used in the past and it is interesting to review some of these concepts to see how the modern sizing criteria have evolved.

1.3.1.1 Coberly criterion

Coberly formulated one of the first gravel sizing criteria. He observed that spherical grains could effectively bridge over openings twice as large as the grain diameter. It can be shown that an hexagonal packing of uniform spherical grains leaves a pore diameter 6.5 times smaller than the sphere diameter. Coberly further reasoned that if gravel is sized to bridge the largest 10 % of the formation sand grains, then the remaining formation sand will bridge on these larger grains. Hence Coberly recommended that the gravel should be selected to be 13 (6.5´2) times larger than the 10 percentile formation sand grain size.

In practice this criterion was found unsatisfactory. Experience has shown that the bridging process is unreliable and allows for too much sand invasion into the gravel which severely restricts the permeability of the pack. Sand bridges will collapse under disturbed flow conditions caused by production fluctuations or interruptions.

Several authors elaborated on Coberly's original concept and the so-called 10 % rule was proposed, i.e. the gravel should be 5 to 8 times the 10 percentile formation sand grain size.

1.3.1.2 Schwartz criterion

Schwartz used the formation sand uniformity coefficient and velocity considerations to select the gravel size. He assumed that sand bridges collapse at high flow velocities.

For most formation sands, Schwartz's and Saucier's method will select the same gravel size especially for uniform sands. For poorly sorted sands, Schwartz's criterion will advise very small gravel sizes, which may be less practical and more susceptible to impairment.

1.3.1.3 Saucier criterion

Saucier carried out laboratory experiments to determine the optimal gravel to sand size ratio [924]. His work proved that it was more effective to stop the formation sand at the gravel/sand interface.

Saucier used a linear flow test cell with gravel on one side and formation sand on the other. The pressure drop through the cell was measured under various flow rates and a range of gravel to sand ratios were tested. He showed that under disturbed flow conditions (i.e. by alternating the flow direction) and for certain gravel/sand size ratios, bridges break down and formation sand invades the gravel pack resulting in impairment at the gravel/sand interface. Hence the effective pack permeability will only be a fraction of the initial permeability.

The permeability ratio shown on this generalised curve is the ratio of the effective pack permeability (i.e. after exposure to disturbed flow conditions) to the initial pack permeability.

Above a ratio of 10, high pack permeabilities are obtained but the formation sand can flow through the gravel and sand control is not achieved.

Saucier also calculated the ratio of the effective pack permeability to the formation sand permeability.

Hence the optimal gravel to formation grain size ratio i.e. the ratio which maintains the highest pack to formation permeability ratio and avoids sand production is between 5 and 6.

1.3.1.4 Comments on Saucier's criterion

Saucier's gravel sizing criterion has set the trend towards small gravel sizes and has become the industry standard. The following important comments on Saucier's experiments should be noted:

·For a reasonably uniform sand, Saucier's results are logical as the median grain size is theoretically smaller than the pore throats of the gravel pack (assuming a hexagonal packing mode). Hence the gravel acts as a filter for the formation sand.

·The gravel invasion process by formation sand should be a function of the gravel packing quality or tightness.

·Even under controlled laboratory conditions, small amounts of formation sand were produced continually, independently of the gravel size used. This observation indicates that finite sand production is inherent in gravel packs and this is confirmed by field experience.

·Saucier's criterion leaves room to be conservative if required. Selection of a smaller gravel size than indicated by the 5 to 6 ratio will reduce the ratio of effective pack permeability to formation permeability. However even when selecting one size smaller gravel, the permeability contrast should remain adequate for all practical purposes. It should be further noted that gravels smaller than 40/60 are susceptible to plugging by formation fines clays and wellbore contaminants.

1.3.1.5 Gravel sizing guideline

a. Uniform formation sands

Before proceeding with the selection of the gravel size, it is important to check the quality of the formation sand samples. A wide spread of the data points in the grain size distribution plots indicates poor quality or non representative samples. Such samples should be treated with caution as they may lead to inefficient sand control:

For each sample the permeability should be measured, and it should be determined whether or not the sample is from an interval that is likely to be produced.

For each producing sample, the gravel size should be selected according to Saucier's rule i.e. select a gravel median diameter 5 to 6 times bigger than the formation sand median diameter:

D50, gravel = 5 to 6´D50, sand

The gravel selected for each producing interval sample should be the next smallest API recognised size or commercially available size.

The gravel selected for packing the hole interval should then be the smallest gravel size chosen for all the producing interval samples.

b. Non-uniform formation sands

If most samples obtained for the interval under consideration are poorly sorted then there is a higher risk that the gravel pack will not effectively control formation sand and will become impaired due to fine sand invading the pack. This poor sorting may be due to the sampling quality (e.g. few samples of poor quality) and if possible better samples should be obtained.

In the confirmed case of poorly sorted sands or when better samples cannot be obtained then a more conservative gravel selection criterion should be adopted.

c. Fine formation sands

In practice, gravel sizes smaller than 40/60are rarely used because this fine gravel is more susceptible to impairment and more difficult to retain in place (smaller screen slots required). Field experience indicates that 40/60size gravel is small enough to control most "productive" formation sands. However in some cases 50/70or even 80/100gravel has been used to minimise fines production.

According to Saucier's criterion, the finest formation sand that can be controlled with 40/60gravel ( D50, gravel = 340 mm) has a median grain size in the range of 68 to 57 microns. Referring to Table 70, a sand with a grain diameter less than 60 microns is classified as silt!

1.3.2 API recognised gravel sizes

Gravel can be ordered by specifying two mesh sizes. A selected number of gravel sizes for sand control applications were recognised by the API. Following API specifications, 96 % of the gravel should pass the coarse designated sieve and be retained on the fine designated sieve (see also Section 1.3.4, Gravel quality specifications).

The nominal median diameter is the arithmetic average of the two mesh sizes given by the gravel designation.

It is recommended that the API recognised gravel sizes are used whenever possible. Other gravel sizes are commercially available but may be more difficult and more costly to obtain. Also API gravel quality control specifications are fully defined for the recognised sizes given above.

For logistical reasons, operators with large gravel packing operations will tend to standardize on a minimum number of gravel sizes.

1.3.3 Why not always the smallest gravel ?

Small gravel sizes provide adequate permeability contrast with formation sand. For example, the permeability of 40/60gravel is some 50 Darcies, much larger than the permeability of most reservoirs. Hence one could argue that small gravel sizes could be used in all cases without significant effect on the well deliverability. However small gravel sizes are more prone to impairment through contamination by dirty fluids or poor placement techniques. Smaller screen slots can plug-up more easily and the manufacturing of these smaller screen slots is more difficult. In practice 40/60gravel is the smallest gravel size routinely used as experience indicates that it is small enough to control most "producible" formation sands.

1.3.4 Gravel quality specifications

The use of high quality gravel is an important factor for placing unimpaired gravel packs. Gravel characteristics which can result in reduced pack and formation permeabilities include excessive fines and clay content, excessive oversized grain content, excessive fines generation under load and pumping conditions. High quality gravel should consist almost exclusively of quartz which is a hard mineral very resistant to crushing and attack by acids. High quality gravel normally contains a minimum of 98 % quartz. Impurities are an indication of a weaker gravel and also indicate that the gravel will be more soluble in acid, steam and even in water. These are important considerations for water or steam injectors. Grain multi-crystallinity contributes to fines generation when gravel is subjected to load. Well rounded gravel is preferred as it is less subject to grain breakage and gives more permeable packs.

Efforts to raise the quality of gravel pack supplies were started by the publication of proposed gravel pack sand specifications. Some problems were recognised with the original Shell gravel specifications, i.e. roundness and sphericity tests were probably subjective, the "visual appearance" test was unsatisfactory and no strength or abrasion resistance tests were available. The latter test should measure the gravel's ability to resist fines production during storage, transport and pumping into the well.

Discussions on this subject have been carried out throughout the industry under the auspices of the API subcommittee on the evaluation of gravel packing materials.

Full quality control of gravel is generally made at the production chemistry laboratory. It must also be ensured that the proper gravel is delivered at the wellsite and that it has not been contaminated with foreign material.

If gravel is used with a silt and clay content near the API 1% limit then formation impairment is a substantial risk. It is recommended to check this parameter on batches of gravel. A new specification has not been established but as a working guide, a level of 0.1 % silt and clay content should be maximum acceptable.

1.3.5 Effects of pumping equipment on gravel

Gravel slurries are normally pumped with positive displacement pumps. Roll has found that an insignificant amount of fines is generated when pumping gravel with viscous or low-viscous carrier fluids. He observed that a general reduction in the gravel grain size due to grain breakage can occur if poor quality gravel is used.

1.3.6 Miscellaneous data

1.3.6.1 Gravel permeability

Fig. 685 shows gravel permeability data for various sizes of commonly used gravel as reported by Sparlin, and Gurley. Note that grain shape and gravel compaction are factors that also affect pack permeability.

1.3.6.2 Effect of mixing of gravel and formation sand

Sparlin has shown that the permeability of gravel can be significantly reduced if mixing with formation sand 

1.3.6.3 General data

Absolute sand density: 2650 kg/m3 (164 lb/ft3)

Bulk gravel density (loose pack): 1620 kg/m3 (100 lb/ft3)

One gravel sack 100 lb (50 kg): 1 ft3(0.028 m3)

1.3.6.4 Low density gravel

In many cases low productivity from internal gravel packs, particularly in highly deviated completions, has been attributed to poor perforation tunnel fill. Many researchers have associated this with poor rheological properties and leak off characteristics of the gravel pack carrier fluid. Much of the research and development work in this area has therefore concentrated on the development of improved viscosifiers.

1.4 Screen considerations

Slotted liners and screens come in many forms and their use demands a correct evaluation of the slot width necessary to retain the gravel. This equipment should also satisfy a number of quality specifications.

1.4.1 Slot sizing

The slot width of a screen is sized in accordance with the gravel size used to control the formation sand. A bridging criterion is totally inadequate here as experience shows that bridges collapse under disturbed flow conditions. Any production of gravel will jeopardise the success of the treatment by creating voids in the pack or erosion of the slots. An absolute stoppage criterion has to be used i.e. the slot width or wire spacing must be smaller than the smallest gravel grain size used.

The term gauge refers to the slot width or the spacing between wires as measured in thousands of an inch. For example a "20 gauge" screen has a slot width of 0.020 inch. This can easily be checked with a feeler gauge.

Because of the relationship between formation sand grain size, gravel grain size and slot width, it is obvious that formation sand can flow through the screen openings if it is not completely covered by gravel. In critical gravel pack applications (e.g. long, highly deviated intervals) where it is difficult to ensure that the screen is fully covered with a gravel sheath, a prepacked screen can be used as a further insurance against sand production

1.4.2 Screen types

Four basic types of screens are currently available i.e. slotted liners (SL), wire wrapped screens (WWS), prepacked screens (PPS), and sintered pack screens (SPS). This section briefly discusses the basic characteristics of each. Other recent developments in screen technology are also covered.

The configuration of the openings in all screens is very important. If the sides of the slots are parallel, plugging will occur as small particles can bridge inside the slot. Openings with non parallel sides and with the narrowest width on the outside are less susceptible to plugging 

1.4.2.1 Slotted liners (SL)

Slotted liners can be manufactured from oil field tubulars which are slotted with a precision saw or mill. Slots should be cut longitudinally so as not to weaken the pipe in tension. Various vertical slotting patterns can be used depending on the desired inflow area and retained strength of the base pipe. Vertical staggered slots as shown in Fig. 690 are most commonly used. The base pipe material should be selected according to the downhole conditions. In a corrosive environment, erosion/corrosion of the slots may lead to plugging of the screen.

Undercut slots are not recommended for slotted liners as they are difficult to saw and in practice still tend to plug because of the thickness of the pipe and the poor undercuts generally obtained. Undercut slots may also present feathered edges which are susceptible to much more rapid erosion/corrosion than square cut slots. It is difficult (i.e. expensive) to cut small slots with the necessary degree of quality control. The smallest practical slot size is about 0.3 mm (0.012 in) straightcut or 0.51 mm (0.020 in) undercut.

Slotted pipe screens have relatively low initial cost but have their limitations, i.e. limited inlet flow area and susceptibility to plugging and corrison/erosion. Manufacturers have been able to create undercut slots using laser cutting technology. Slot sizes as small as 0.006" have been cut. Rapid development of this technique is expected if costs drop significantly below the cost of a comparable wire wrapped screen. 

1.4.2.2 Wire wrapped screens (WWS)

Wire wrapped screens are constructed by wrapping a wire around a perforated or slotted base pipe. The wire is spaced to give the required slot width and a self cleaning slot is obtained by using a "keystone" shaped wire. The "all welded" wire wrapped screen type (Fig. 691) has superseded all other wire wrapped screen types used in the past (wrapped-on pipe, grooved type, ribbed type.

All welded screens are generally manufactured as follows. The jacket is first manufactured by wrapping a "keystone" shaped wire on longitudinal ribs. The wire is welded at each contact point with the ribs by induction welding and this process is continuous and automated, giving good manufacturing control of the slot width. The longitudinal ribs are required to manufacture the jacket but they also increase the flow capacity by providing an annular space between the wire wrapping and the pipe base. This jacket is then slipped over and welded to a perforated base pipe.

A major advantage of the all welded construction technique is that the wire is well secured to the jacket. A common problem with earlier screen designs is that the wire wrapping could unwrap from the base pipe when attempting to run or retrieve the screen. Induction welding also allows the use of narrow wrapping wire thereby maximising the slot flow area.

All welded screens with a slot size of 0.002 inch have been successfully manufactured for special applications. However small slot sizes result in reduced inlet flow area and increased plugging risks. For gravel packing applications, the smallest slot size used in practice is 0.006 in with 50-70 gravel.

Stainless steel wire (304 or 316 grade) is generally used because of its superior abrasion and corrosion resistance. The wire is usually wrapped on normal grade carbon steel tubulars for non- corrosive environments.

The manufacturing of wire wrapped screens is a semi-automated process. Hence screen design can be adapted to suit particular needs.

1.4.2.3 Comparison of wire wrapped screens and SL

The slot size that can be economically cut for a SL is an obvious limitation. For the same slot size, a SL has a smaller inlet flow area compared to a wire wrapped screen. This may create an excessive pressure loss depending on flow rates and fluids produced although an adequate flow area is still provided for most applications. However fines production and high flow velocities may lead to rapid erosion of the slots.

It is claimed that SL are more susceptible to plugging because of the slot configuration and the smaller inlet flow area. Slot plugging may be caused by hydrocarbon or scale precipitation or even corrosion. SL is stronger and cheaper than wire wrapped screen but is more susceptible to erosion/corrosion of the slots which may lead to a sand failure. SL can be manufactured in corrosion resistant alloys.

Because of their lower costs, slotted liners are the preferred option for marginal, low rate wells. However, where corrosion resistant alloys are required the incremental cost of wire wrapped screen compared to slotted liner becomes marginal, hence the use of wire wrapped screen in such cases is generally preferred.

1.4.2.4 Prepacked screens (PPS)

The primary use for prepacked screens is generally as a substitute for wire wrapped screen when:

1.gravel packing highly deviated zones, or

2.as a primary form of sand control in horizontal wells.

Prepacked screens are basically an assembly of concentric screens packed with gravel. The gravel should always be consolidated to prevent or losing gravel when handling the liner. The design criteria is basically the same as for gravel packing i.e. the gravel should be sized to stop the formation sand (Saucier). The 40/60and 20/40U.S. mesh gravels commonly used in these screens have average pore throats of 50 and 100mm, much smaller than wire spacing of conventional screens, perhaps explaining why they plug more easily.

Furthermore, mechanical damage to prepacked screens has been widely reported. This is thought to be attributed to:

1.manufacturing defects,

2.flexing of prepacked screens during pick up and make up, or

3.while installing as this may cause the brittle thermo-setting plastics used to crack.

Prepacked screens are more expensive and are usually physically larger than conventional screens. By design, they are less likely to fail because of erosion. However, plugging of the gravel sheath during installation due to dirty completion fluids or by formation fines during production is a major problem. To help minimise plugging some operators run the prepacked screen open ended.

a. Types of prepacked screens

Fig. 692 illustrates the main screen types which are marketed by the major sand control service companies and screen manufacturers.

Single screen prepack - This screen consists of an inner perforated base pipe surrounded by an inner screen. Only resin coated gravel is used as the filtration medium placed between the inner screen and outer perforated case. This is probably the most commonly used prepacked screen, and is the most rugged design.

Dual screen prepack - This screen consists of an inner perforated base pipe, surrounded by screens. The gravel may or may not be resin-coated, and is normally contained by an outer screen. This is a heavy duty premium grade assembly designed primarily for maximum erosion resistance. Hence it found most of its application in high rate gas wells.

Thin body prepacked screen - Recently developed, thin bodied prepacked (with gravel or resin coated gravel) screens are now being used extensively (mainly in the US) as replacements for wire wrapped screen. Because of the reduced OD and/or increased ID, it allows better optimisation of the radial clearance outside the screen and washpipe, thereby enhancing gravel placement. A reduction in overall screen OD was made possible by using a wire mesh as the inner screen to retain the gravel. There are no known guide-lines for the minimum gravel sheath thickness to be used.

A number of variation are available from the major screen manufacturers:

1.Baker Sand Control's Slim-Pack,

2.Johnston's Thin-Pak,

3.Howard Smith's LOW-PROFILE screen marketed by Otis Sand Control.

Further advantages and disadvantages of the "new" thin bodied prepacked screen are briefly discussed below.

Advantages:

·These types of prepacked screen have the same OD as regular wire wrapped screen, hence increasing potential applications.

·Less expensive than larger diameter prepacked screen.

·It has a higher flow capacity per unit length of screen than conventional prepacked screens with thicker gravel bodies.

Disadvantages:

·They are more susceptible to physical damage than conventional prepacked screens.

·The thin gravel layer is less resistant to erosion.

·The even placement of a thin gravel layer is difficult to achieve during fabrication.

In summary, highly permeable slim pre-packed screens seem to present an insurmountable series of manufacturing and quality control problems, thus making their use difficult to justify as a stand alone primary means of sand control.

b. Quality control of prepacked screen

Wellsite checks - Since the permeability of gravel inside the screen cannot be directly checked it is recommended that each joint is thoroughly inspected to ensure the resin coated gravel is not cracked or missing. Apart from the obvious visual checks, screens can be tested by flowing filtered water through the screen and monitoring the pressure drop. Squirting jets of water emanating from a localised point usually indicate a screen defect.

Screen manufactures should in all cases be consulted on recommended quality control checks. Additionally, service companies will help establish the maximum dog-leg which prepacked screen can endure without causing irreversible damage or deformation.

1.4.2.5 Sintered metal screens (SMS)

One major screen manufacturers has developed a unique sintered stainless steel gravel pack screen. This screen consists of an outer sintered sleeve welded onto a perforated base pipe. It is claimed that the mechanical characteristics of such a screen make them ideally suited for high dogleg, deviated or horizontal wells where, the performance of prepacked screen is questionable.

One size screen is compatible with all gravel sizes used in gravel packing and is available in different alloys. For example the 100mm sintered pack screen has a median pore diameter size of 42mm and is suited to control sand in the 40-100mm range.

Vendors claim that this type of screen has maximum erosion resistance. They are certainly more rugged, and therefore suitable to pipe handling procedures. The susceptibility to plugging is thought to be greater than that of prepacked screens. Howard Smith Screen Co. the manufacturers, claim that sintered screens are cleared more readily by back flushing.

The manufacture of sintered metal screen is a three part process. The first phase is known as isostatic compaction. A polished steel mandrel forms the ID of the screen, with an outer shell forming the screen OD. Powdered metal is placed in the "jig", which is sealed and subjected to radial pressure, which directly controls screen thickness.

The second part of the process is sintering, resulting in a re-crystalisation of the micro-structure, and in inter-particle bonding. Screens are placed in vacuum ovens and cycled over a range of temperatures for a 24 hour period.

The third part of the process involves electro-polishing of the screen jacket which increases corrosion resistance and tends to slightly open pore throats. The jackets are then installed on a perforated pipe and welded into position.

The maximum OD sintered pack screen that can be manufactured is currently limited by the size of the existing sintering ovens which allows 11" to be manufactured.

Potential advantages:

·Sintered pack screen design is tougher and more damage resistant than wire wrapped screen and prepacked screen.

Less flow restriction compared with wire wrapped screen and prepacked screen.

·Flexibility/elasticity to pass high doglegs without causing irreversible damage. Tests carried out by the manufacturers have demonstrated the ability to pass through tight radius bends with no damage to the screen jacket.

·There is scope to further reduce screen costs by omitting the base pipe. Manufacturers are investigating alternative coupling techniques.

Disadvantages:

·Susceptible to fines, scale and wax plugging.

·Certain high nickel alloys are not suited to the manufacturing process, hence their use is presently restricted to non-hostile environments.

·Sintered screens are more costly than wire wrapped and prepacked screens.

1.4.2.6 Selective Isolation Screens (SIS)

This screen type is a variation of the normal wire wrapped screen. Its development was driven by the need to improve the success ratio of remedial treatments through gravel packed completions.

The basic design includes inner annulus seals and only a few holes or slots machined in the pipe base. The inner annulus is restricted by using wire wrapped screens fabricated with seals approximately every 1.5m. The seals do not need to totally stop fluid flow, but only restrict the movement of fluid in the inner annulus. In all cased hole completions the screen inflow area (maximum number and size of the holes) should not exceed the number and sizes of the perforations in the casing.

A number of potential advantages of this type of screen are:

·If for whatever reason the gravel pack has to be washed after placement (with acid, or under saturated brine) each short section of screen can be somewhat isolated with a wash tool to promote positive injection into each interval.

·Diversion of acid or solvent can be achieved by means of conventional ball sealers.

·Allows the possibility of isolating bottom water once it breaks through

The main disadvantage with this type of screen is the higher cost as most screen manufacturers make them to order.

1.4.2.7 Shunt screen

This screen is also a variation of the normal wire wrapped screen.

Poor gravel pack integrity as a result of premature sand bridging is a classical problem associated with the completion of long, highly deviated intervals. In such applications the risk of bridges forming adjacent to zones of higher permeability due to preferential leak off is pronounced.

a. System description

The shunt or alternative path gravel pack system was developed and licensed by Mobil Research and Development, and is presently being marketed by Baker Sand Control (BSC) as the ALLPAKTM gravel pack system.

Alternative flow paths, called shunts or secondary conduits are attached to the side of a gravel pack screen. The shunts have small holes drilled every few feet allowing communication to the screen/casing (or hole) annulus across the entire completion zone. Four rectangular shunts are set a right angles to ensure that at least two are at the top of the liner. The rectangular geometry has been selected to maximise clearance in the annulus while providing sufficient flow area. The tubes are open at top of the screen only to facilitate alternate by-pass.

The number and size of holes drilled in the shunt is a critical parameter in order to avoid excessive leak off, hence low shunt velocity and hole plugging. They are sized and configured with respect to gravel size, completion configuration and fluid rheology. Since no design guidelines have been published, readers are advised to consult BSC directly for more details on shunt design criteria.

b. Test results and field trials

The screen has been tested [987] in BSC's large wellbore model which allows visual observations to be made. The results on a 10m long, highly deviated (> 70°) interval indicated gravel pack efficiencies of 95-100% compared to 65-80% for conventional procedures. Furthermore over 20 shunt screens have now been installed by a number of companies operating in the Gulf of Mexico and the North Sea. Results to date demonstrate that long (>100m) highly deviated (> 70°) wells can be successfully packed.

Opcos interested in evaluating the shunt screen for "difficult" completions are advised to consult their local BSC representative for more information.

1.4.3 Screen dimensions

The selection of the screen size is mainly a function of the type of gravel pack. In cased hole, screen/casing annular clearance is critical to ensure proper gravel placement and to allow for fishing operations if the need to retrieve the screen arises. In multiple completions the engineer must take into account the size of tubulars needed to drain reservoirs below the current gravel pack. From a productivity point of view, little will be gained by selecting a screen ID larger than the tubing ID. 

1.4.4 Screen material selection

Screen material specifications should be derived from the expected service conditions. The main concern for a screen is to ensure slot integrity over the desired lifetime. In practice no slot corrosion can be allowed as tolerances on the slot width is typically a thousandth of an inch. Screens made of special alloys are obviously more expensive and quality control specifications are more stringent.

Slotted liners can be manufactured from carbon steel or stainless steel. It should be noted that a carbon steel configuration is very susceptible to corrosion/erosion in the slot area. Wire wrapped screens are generally made of carbon steel for the base pipe and AISI 304L or AISI 316L stainless steel for the jacket as a minimum specification. Although corrosion of the base pipe is less critical than for the jacket, this aspect should not be overlooked. Some general guidelines for materials selection are given below

304L and 316L alloys are good for CO2 service (when < 0.1 bar partial pressure H2S) in producing wells (no free oxygen present) and up to at least 120°C. Experience suggests there is no upper limit on the CO2 level that can be tolerated. These austenitic alloys are however susceptible to pitting, crevice and stress corrosion cracking (SCC) in the presence of chlorides and oxygen (AISI 316L has a higher corrosion resistance than AISI 304L). To minimise the risk of corrosion, the pH of the workover brine should be kept above nine, also when extended exposure (more than a day) to brine is unavoidable, an oxygen scavenger should be used.

Alloys such as Sanicro 28 or Incoloy 825 are adequate for higher H2S levels (up to 5 bar partial pressure), with or without CO2 and for the same temperature range. 

1.4.5 Mechanical properties

API grade J 55 provides enough mechanical strength for screen base pipes of slotted liner or wire wrapped screen as these are only subject to limited tensile loads and pressure differentials. The connection type is also not critical and normal flush API connections are generally specified.

1.4.6 Centralisers

Screen centralisation is critical for good gravel placement, especially for cased hole applications when there is low clearance between the screen and the casing. Welded, blade type centralisers should be used as positive centralisation is achieved and they are more reliable than clamp-on types. Spacing will be a function of the specific wellbore conditions. In open hole completions, bow-type centralisers are normally used. However a number of companies have used rigid aluminium centralisers in barefoot wells to provide sufficient stand off to minimise screen damage.

When wire wrapped screen are used in gravel pack operations the length of blank section should be reduced as much as possible to minimise the risk of bridging.

Local sand control service companies will help design a centralisation programme

1.4.7 Screen quality specifications

Screen quality is as important as gravel quality. Quality control of the slot width is the most critical aspect. Oversize slots will allow some gravel to be produced through the screen and may jeopardise the success of the treatment by creation of voids in the pack.

The API Std 11D on Miscellaneous Production Equipment was withdrawn from publication in 1976. The requirements for screen pipe and slotted pipe openings contained in Std 11D were dropped and are no longer covered by an API standard.

These tolerances normally do not present a problem except for manufacturing the full weld on type of screen. The problem is caused by slight irregularities in the dimensions commonly found in the base type. These irregularities are enough to cause the wire spacing to deviate during manufacture. For this screen type the tolerance figures shown above may be unrealistic. In such cases +2 and -3 gauge specifications may be appropriate.

It should also be noted that certain doubts remain within the industry as to whether or not the above tolerances are realistic in view of base pipe distortion during the manufacturing process

1.5 Gravel packing systems

Gravel placement requires the use of specialised downhole tools. A wide variety of gravel placement techniques and tools exist and only the basic tools and methods are reviewed in this section. Historically, the first gravel placement methods used were:

·The wash-down method: gravel is first deposited in the well and the screen assembly is subsequently washed down through the gravel. The gravel is then left to settle around the screen. Disadvantages of this method are the difficulty of placing a dense pack around the screen, the stirring of gravel inside the wellbore and the risk of sticking the work string when washing out the gravel.

·The reverse circulation method: gravel laden fluid is circulated down the annulus, gravel is retained by the slotted pipe and the fluid returns through the work string. The equipment and procedures are fairly simple but the main disadvantage is that the gravel is subject to contamination as the slurry is pumped down the annulus.

These methods have been completely superseded by the crossover circulation method.

1.5.1 Crossover circulation method

The standard technique for gravel packing either in cased hole or open hole relies on a packer (or a simple annulus pack-of device in the simpler systems) and cross-over tool combination for routing the gravel slurry from inside the work string to the screen-casing annulus.

After the screen and liner are run and positioned, the packer is set and the gravel pack port is opened to allow communication behind the screen and below the packer. A washpipe is used to force the slurry to the bottom of the screen/casing annulus. By choking the returns or by the use of a multi position gravel pack tool, the slurry can be circulated behind the screen and partially squeezed into the formation. After the gravel pack has been placed, the work string with the crossover is pulled and the well is completed by running a production tubing with a seal extension, landed in the packer. The main advantages of the crossover technique are:

·The slurry is pumped through a work string which is kept clean to avoid contamination of the slurry.

·Treatment pressures are confined to the completion interval below the gravel pack packer.

·Multi-position gravel pack tools allow for accurate control of the different phases of a gravel packing job i.e. circulating behind the screen, squeezing into the formation and reversing out excess gravel. Additionally, all gravel packing operations can be carried out in one trip.

There are many different types of gravel packing tools, each developed for specific applications. A thorough discussion of all these variants would be beyond the scope of this manual and detailed technical information can be provided by vendors upon request. 

1.5.2 Multi-position crossover tool

Gravel pack tools which allow the gravel pack liner assembly to be run and all gravel placement operations to be carried out in one trip have been developed by all reputable gravel packing contractors. It is the recommended method for gravel packing in cased or open hole as it allows to work clean, is faster, gives better control over the job and provides better capabilities for good gravel placement.

The heart of the system is the multi-position crossover tool. Different gravel placement modes are obtained by manipulation of the work string.

·Running position: when running the gravel pack liner assembly into the well, circulation can be achieved through the crossover port and up the work string annulus.

·In the squeeze position, no fluid returns are possible and the fluids are pumped into the formation. This position is used either to carry out injectivity tests prior to pumping the slurry or to dehydrate the slurry into the formation exclusively.

·In the upper-circulating position the slurry is circulated through the gravel pack port and behind the main screen. Fluid can leak-off into the formation or circulate through the screen and return through the washpipe.

·A lower telltale screen can be added at the bottom of the completion string. A seal bore is then incorporated between the two screens. The end of the washpipe seals into the sealbore in the lower circulating position and fluid is forced to circulate through the lower telltale screen.

·In the reverse circulation position fluid can be circulated above the packer from the annulus to the work string. This position is used after screen-out to reverse circulate excess slurry out of the work string. It can be also used to circulate slurry batches to the crossover port when attempting a top-up.

The different positions of the crossover tool have to be located and tested prior to carrying out gravel packing operations.

All the capabilities discussed above may not be available for a particular tool. It is important to be aware of the design and capabilities of the specific tool which is used. Vendors will provide detailed documentation upon request.

1.5.3 Gravel packing systems for special applications

Many variants of the one trip gravel packing tools have been developed for use in special applications e.g.:

·Systems enabling several zones to be gravel packed in one trip, i.e. Baker Sand Control "Beta" system for Shell Oil's Beta field in the U.S. [932], [933].

·Systems tailored for deep water offshore applications i.e. the four position long stroke gravel pack tool for deep water offshore operations [934].

1.5.4 Low cost gravel packing system

Low cost gravel packing systems generally feature a crossover tool fitted with a temporary annular pack-off device i.e. rubber cups. After gravel packing, the crossover and liner section above a back-off sub are pulled from the well. If required, the liner extension can be sealed with a low cost lead drive-over seal. This gravel packing system is applicable to low pressure, economically marginal wells. One disadvantage is that many round trips are required in order to complete the well. Many variants on this completion technique are possible. Detailed information can be obtained from vendors.

1.5.5 Through tubing gravel packing

There are systems available for gravel packing through tubing which can be attractive for remedial sand control work in economically marginal operations. Very little experience is available within the Group with this system.

Coiled tubing is used to clean out the hole and to spot gravel to cover the perforations. A small diameter screen and liner is then run and washed down to the desired depth. After the liner is placed, the work string is removed and a wireline pack-off may be run to seal the top of the liner.

1.5.6 One trip perforate and gravel pack technique

Integrated one trip perforating-gravel packing systems have been available to the industry since the early 1970's. These systems are often preferred by many operators to minimise exposure to losses, and also to reduce operating costs. The original concept was first developed by Otis, and since then all major sand control service companies have developed equivalent one trip systems (eg Baker's SCP system, Dowell's PERFPAC). The basic design is similar and based on TCPs being positioned below a gravel packing assembly, separated by a retrievable "perforation" packer. The installation procedures are also comparable for each of the marketed systems, and are summarised as follows:

1.TCPs are positioned on depth and the "perforation" packer set.

2.The well is perforated and allowed to clean up.

3.The well is killed by reverse circulating. Note the TCPs are normally dropped into the well sump, but are sometimes left hanging below the 'perforation packer'.

4.The "perforation packer" is unset and the assembly moved down until the screens straddle the completion interval.

5.The 'perforating packer' is then re-set. The gravel pack packer is then set and gravel pack operations commence.

A ceramic flapper-knock out isolation valve is normally incorporated into the gravel pack string which will help reduce losses when the inner wash pipe is withdrawn, hence losses during tripping are minimised. Once the final completion string has been run and tested, the flapper valve can be shattered using the completion assembly tail, or by the application of pressure.

Although this technique is field proven it is mechanically more complex than conventional systems, thereby increasing the chance of operational problems. The incremental cost of drilling a well sump should also be taken into consideration.

It should be noted that prepacking the perforation tunnels prior lowering the assembly is considered a risky operation and is not generally carried out. This technique is not therefore applicable to long intervals (> 10 m) or intervals with significant variations in reservoir permeability, as the risk of premature bridging is considered too high.

1.6 Gravel pack liner assemblies

1.6.1 Work string

The work string can be either tubing or drill pipe. Drill pipe is commonly used and has the advantage of a higher load rating. The work string should be rigorously cleaned prior to the gravel packing operations. The work string ID should be small enough to allow proper placement of the slurry (i.e. to prevent U-tubing). Pumping should be carried out under turbulent flow conditions to minimize mixing of the various fluids.

1.6.2 Gravel pack packers

1.6.2.1 Requirements

A packer is required during gravel packing operations to isolate the treated interval from the wellbore. It is also required when the well is on production to avoid pack fluidisation by upward flow through the annulus and in many cases, to seal the annulus as a mandatory completion requirement. However calculations have shown that fluidisation is theoretically possible only for very high flow rates or small gravel reserves.

In some cases a temporary gravel pack packer is used for the gravel pack operation. The retrievable packer is subsequently replaced by a cheaper annulus pack-off assembly.

1.6.2.2 Permanent gravel pack packers

Permanent gravel pack packers fulfil the dual role of a gravel pack and production packer. They form part of the gravel pack liner assembly which is run together with the crossover tool. Some important considerations when selecting a packer are the setting and crossover release mechanisms.

1.6.2.3 Permanent-retrievable type

Probably the most commonly used gravel pack packer type is the "permanent-retrievable" type. They offer nearly the same capabilities as a permanent type packer but have a retrievable feature. The disadvantage is that they are mechanically more complex implying a higher cost and potentially lower reliability than permanent packers. In a sand control context, retrievability is no longer a primary requirement as gravel pack installation is generally successful and failure rates are low.

Permanent-retrievable packers can be set mechanically or hydraulically. In a gravel packing context, these packers are run together with the crossover tool which can also be released either mechanically or hydraulically. The choice of the setting and crossover tool release mechanism will be a function of the particular wellbore conditions. Hydraulically set packers are generally preferred in difficult holes where mechanically set packers are more difficult to run and operate.

The retrieval procedure requires the use of a specialised retrieval tool run on drill pipe. The packer is released by a straight pull or by rotation. Wireline retrievability is probably effective only in shallow applications. If required permanent-retrievable packers may also be retrieved using packer milling tools.

a. Permanent type

Permanent type packers may be preferred in critical applications (high pressure and corrosive environment) because of the higher reliability of their sealing capability. The same considerations as for retrievable packers apply when selecting a permanent packer.

1.6.2.4 Temporary gravel pack packers

In some areas it is common to use a rented temporary gravel pack packer and crossover tool assembly for the gravel packing operations. The packer can be either a cup type packer or squeeze type (RTTS) packer. Cup type packers are more susceptible to damage when running in the well.

After the gravel pack operations are complete, the temporary packer and crossover tool are retrieved and a cheap (usually lead seal) annulus pack-off is set above the top of the liner, if required. Several trips may be required to complete the well. This type of equipment is used for low-cost, low pressure completions.

1.6.3 Blank liner sections

A reserve gravel volume, which is placed in the annular volume between the gravel pack port and the top screen, is required for the following reasons:

·To allow for pack settling (filling of voids, compaction, gravel dissolution in water or steam).

·As a safety margin to cover uncertainties when determining the total gravel volume required i.e. the gravel volume placed behind casing for IGPs and wellbore volume for EGPs.

In highly deviated wells ( > 60 degrees ), gravity no longer assists the filling of voids in the pack by settling and compaction and limits the effectiveness of the gravel reserve.

In the absence of an upper tell-tale, the final screen out occurs when a certain length of compacted gravel covers the top of the wire wrapped screen. The height of this compacted column is a function of the carrier fluid viscosity, geometry of the completion etc. The amount of gravel contained in the annulus between the top of this column and the gravel pack port at the time of the screen-out is a function of the gravel mix ratio.

With a 15 lb/gal slurry, nearly 65% of the annular volume will be filled with gravel. Conversely, at low gravel concentrations only a small gravel volume can be placed after screen out. This is the reason why upper tell tales are used in water packs. The upper tell tale allows reserve gravel to be placed by providing an additional circulating path above the main screen. This will only be effective if the deviation is not excessive (< 60 degrees). The reserve gravel volume will therefore differ when an upper tell tale is used (refer also to Section 1.6.5.2: Upper tell-tales).

There is no definite set of rules to determine the blank liner length required and practices vary to fit each operator's need. For example BSP uses the following design rule for slurry packs: the gravel pack port to top screen length is 40% of the gravel pack port to sump packer length. When using high gravel concentrations (12 - 15 lb/gal), 60 feet of blank liner is generally adequate for most cases, except very long intervals.

There are cases where the blank liner length has to be restricted e.g. multiple completions where the separation between intervals is small. Smaller reserve gravel volumes can be tolerated when multiple slurry batches can be placed with the placement method used. Also the reserve volume can be reduced when the operator is confident that uncertainties can be minimised i.e. that proper fill at screen-out can be achieved.

1.6.4 Washpipes

A large OD washpipe is the best way to ensure good gravel placement, especially in difficult wells. The favorable effects of running large OD washpipe in the gravel pack assembly have been demonstrated. A large washpipe minimises the flow area between the screen and washpipe thereby forcing the fluid to flow preferentially in the casing/screen annulus. This promotes packing from bottom upwards, prevents bridging in the annulus before complete gravel fill-up, especially for deviated wells, and generally improves the packing efficiency. The perceived disadvantage of using relatively large diameter washpipe is the somewhat increased risk of washpipe sticking inside the screen assembly.

Field experience suggests the following guidelines:

·A washpipe OD to screen ID ratio of at least 0.6 should be used in all cases.

·Perferably and certainly in difficult applications, this ratio should be at least 0.8.

·The washpipe is not required for a single stage squeeze pack (IGP). However this type of gravel pack is generally not recommended, except for very short intervals.

Difficult applications are defined by one or a combination of the following features:

·Highly deviated zones (> 60 degrees).

·Long intervals (> 50 feet).

·Zones with a high permeability contrast.

1.6.5 Tell-tales

Tell-tales are short (5-10 ft) sections of screen which are included in the liner assembly to assist with the gravel placement and/or to give a positive indication over the placement of the gravel.

1.6.5.1 Lower tell-tales (LTT)

The lower tell-tale assembly consists of a seal bore sub and a short piece of screen located below the main screen. The washpipe is sealed into the sealbore and fluid returns are taken through the wash pipe via the lower tell-tale screen.

Historically, lower tell-tales have been incorporated into gravel pack strings to facilitate good gravel placement across the entire completion interval. It is also claimed that lower tell-tales allow a positive indication (lower screen out) when gravel slurry covers the lower screen. However, in practice it is felt that this is not the case.

Although lower tell-tales are in general not recommended they have found applications in specific areas of well completion design. For example, several Opcos have used graded salt (NaCl) mud systems as an under-reaming fluid prior to carrying out an EGP. In such cases the incorporation of a lower tell-tale has proven useful in displacing the wellbore (to remove the filtercake) with an under saturated brine prior to placing gravel conventionally.

Furthermore, when designing long highly deviated wells, the use of a lower tell-tale has assisted in efficiently transporting gravel to the bottom of the completion interval.

In summary, the use of lower tell-tales is one of several areas where there appears to be a lack of consensus.

1.6.5.2 Upper tell-tales (UTT)

Upper tell-tales are used exclusively when using low gravel mix ratios. They are positioned immediately below the gravel pack port with at least one blank joint between the tell-tale and top screen. The upper tell-tale allows to maintain circulation after the main screen is packed-off. Gravel is thus separated against the screen and settles along the blank joint to constitute a reserve volume. This approach is cheaper than to increase the gravel reserve by installing extra joints of wire wrapped screens.

In highly deviated wells (> 60 degrees) the upper telltale is of no use as gravity no longer assists in filling the annulus above the main screen.

Upper tell-tales should not be used with slurry packs (high gravel concentration, high viscosity slurries) where enough gravel reserve is provided by settling of the slurry between the gravel pack port and the top of screen when screen out occurs. The high gravel concentration obviates the need for a tell-tale above the screen. An upper tell-tale in conjunction with high viscosity slurries may cause bridging because of layering of gravel on the screen.

1.6.6 Miscellaneous equipment

1.6.6.1 Gravel pack bases

In cased hole, permanent type gravel pack bases are generally preferred as they provide a reliable fixed platform. Either bridge plugs or sump packers can be used. The advantage of a sump packer is that clean-out trips can be avoided and the possibility for future logging below the pack is maintained. Bridge plugs are however generally cheaper.

In multiple completions, the packer of the lower zone can constitute the base for the next zone. In open hole, a cement plug generally serves as the gravel pack base.

1.6.6.2 Shear safety joint

It is common practice to run a shear safety joint between the gravel pack packer assembly and the blank liner to facilitate fishing operations. The shear rating of the safety joint should be sufficient to withstand the loads associated with installation of the pack.

1.6.6.3 Gravel pack port

The gravel pack port provides communication between the work string and the annulus below the packer. It is part of the gravel pack extension and has to be closed or sealed after placement of the gravel. Different systems are available:

·Gravel pack extension with sliding sleeve: The housing of the sliding sleeve contains the gravel pack port. The sleeve is activated by a shifting tool.

·Perforated gravel pack extension: This consists of a perforated extension with a seal bore sub. The length of the perforated extension is controlled to ensure proper crossover tool positioning and travel. The gravel packing ports are blanked-off by the production string which seals in the packer and the seal bore sub.

1.6.6.4 Fluid loss control devices

Several mechanical fluid loss control devices can be incorporated in the gravel pack liner assembly or the work string. These devices minimise fluid loss to the formation prior to and after gravel packing the well.

a. Low bottom hole pressure ball

A low bottom hole pressure ball can be added to the crossover tool and serves as a check valve to prevent the loss of fluid to the formation while the crossover tool is in the reverse circulation position. When a crossover tool is in the reverse position, the fluid pumped down the annulus can flow through the bypass ports, down the washpipe and into the formation. The ball is run in place inside the seal subs below the gravel pack port of the crossover tool.

b. Knock-out isolation valve

The knock-out isolation valve is a flapper valve (often ceramic) that is located between the gravel pack extension and the screen and allows isolation of the gravel packed interval once the job is complete. It is used for wells which experience severe fluid loss.

During gravel packing the valve is held open by the washpipe. When the washpipe is pulled the valve closes isolating the formation from wellbore fluids. This valve can subsequently be shattered by applying pressure or with the tail pipe of the production tubing.

c. Developments in surface blending and pumping equipment

The beginning of the 1990's saw many operators return to the technique of conventional water packing. This in turn has led to the development and introduction of alternative blending and pumping equipment.

Clearly a detailed description of all the available blending and pumping systems is outwith the scope of this manual. However, a brief review of commonly used equipment which is marketed by several large sand control service companies is given. For more details readers are advised to consult their local service company representatives.

Infuser sand mixing pump - During 1989 Baker Sand Control introduced their gravel pack infuser design to facilitate consistent and efficient blending of low concentration slurries, with the additional benefit of allowing the variation of pump rate and gravel concentration. The infuser consists of an auger pump having inlets for gravel and carrier fluids, and an outlet to a triplex gravel-pack pump driven by a variable speed motor. Observations to date have been positive with a number of operators reporting a number of benefits, including accurate and consistent gravel concentrations a higher rates.

CLAM system - Otis Sand Control offer a different approach for preparing conventional water slurries. The CLAM (Constant Level Additive Mixer) system which has evolved from conventional Halliburton hydraulic fracturing equipment eliminates the necessity to batch mix slurries. This system is capable of continually mixing uniform slurries (0-22 lb/gal) on the fly. Otis claim that the desired slurry can be controlled to within ±0.1 lb/gal. The basic components of the system are the CLAM blender, a high rate centrifugal pump and a high rate pressure pump.

WASP system - The WASP (Water And Sand Proportioner) is a relatively new blender specifically designed by Dowell Schlumberger to proportion and blend gravel and low viscosity carrier fluids. Controlled and continuous blending with low viscosity fluids at concentrations up to 9 lb/gal is achievable.

Other comparable systems include BJ's Service Cyclone Blender and Dowell Schlumberger's POD blender.

1.7 Fluid considerations

The main requirements for a gravel packing fluid are:

·To provide a safe overbalance against reservoir pressures.

·To minimise permeability damage by particle plugging, clay swelling and migration, or excessive fluid loss.

·Transport gravel downhole behind the screen and into the perforations in the case of an IGP.

1.7.1 Base brines

Base brines used for gravel packing operations should provide the required formation overbalance and be compatible with the reservoir rock and fluids. Unless higher brine densities are required, freshwater with 1-3% KCl or sea water are the most commonly used base brines for gravel packing applications.

The addition of potassium chloride to freshwater will inhibit interaction with water sensitive clays present in most formations. Clay de-stabilisation does not occur or can be neglected in brines with total salinities in excess of 50 kg/m3. Sea water which is compatible with most reservoirs is the obvious choice for offshore operations. 

1.7.2 Fluid compatibility

When designing gravel packs for the first time in a field, the following tests are recommended:

·Core flow tests with the same fluids that will contact the formation to determine potentially damaging effects on the reservoir rock and to design remedial treatments.

·Compatibility tests with the reservoir fluids such as emulsion and precipitation tests.

Most sandstone reservoirs contain some clay minerals. The types of clay present, the concentration and the location of the clay minerals should be known. Small formation samples (drilling cuttings, side wall samples) can be used to establish the clay mineralogy.

All these tests are fairly simple and can be done by most reputable service companies.

1.7.3 Fluid filtration

Reservoir permeability impairment due to dirty fluids can be extremely severe. It is very important to filter all completion fluids that may contact the producing formation. Additionally, all viscosified brines used for underreaming, fluid loss control or gravel packing operations should be sheared and filtered to remove unhydrated polymer lumps ("fisheyes") . WORKING CLEAN IS AN ABSOLUTE PRE-REQUISITE TO ACHIEVE LOW IMPAIRMENT.

1.7.3.1 Filtration requirements

The level of filtration depends on the formation plugging potential, practical limitations of the filtration equipment and cost considerations. It is generally accepted that there is a relationship between particle size that is damaging and pore throat size. As a rule of thumb [939]:

·Particles smaller than 1/7of the average pore throat size can freely pass through the pores and will not affect the permeability.

·Particles larger than 1/3of the average pore throat size will be filtered off on the formation face.

Ideally, the size of the particles in the completion fluid should be small enough to freely flow through the formation without plugging the pore openings. Hence the filtering criterion should be to remove all particles larger than 1/7of the average pore throat size.

The average size of the pore throat openings of a particular formation can be measured directly from core samples using a scanning electron microscope or can be estimated using various correlations. A rough estimate of the average pore throat size is given by the Kozeny equation.

The concept presented above can be used to estimate filtration requirements for each well. However this rule is not an appropriate guideline as such. Other parameters to consider are the total suspended solids, the maximum particle size at a given solids content and the treatment cost.

Recent work has shown that particles down to 1/14th of the average pore throat size can still create appreciable impairment at low fluid velocities. For such applications the filtering criterion should be aimed at removing all particles larger than 1/14th of the average pore throat size.

Many companies have found it possible to filter completion fluids to 2mm absolute without difficulty. This should be adequate for most situations.

Through cleaning of the well and surface equipment prior to gravel packing, and use of gravel with minimal fines content [989] and the application of appropriate perforation cleaning is essential to help minimise impairment. Therefore, a maximum degree of cleaning should be aimed for.

1.7.3.2 Filtration equipment

There are different types of filtration systems available for oil field use namely cartridge type, bag type or diatomeceous earth (DE) filters. High speed disk stack centrifuges can also be used to remove solids from brines.

In general the selection of a filtration process will depend on many factors such as quality of influent brine, required level of filtration, location, flow rate, volume of brine to be filtered and cost aspects.

DE filters are the most cost effective devices for filtering large brine volumes. The DE filters require a cartridge filter downstream to filter out any DE that bleeds through the filter and to "polish" the completion fluid prior to pumping it downhole. Viscosified brines should always be filtered to remove all traces of unhydrated polymers.

1.7.3.3 Quality control

Well site quality control of the fluid cleanliness can be difficult. The different methods available to monitor the solids content of a filtered brine are:

·Gravimetric analysis: a given fluid volume is passed through a fine pre-weighed filter. The filter is dried and re-weighed to measure the Total Suspended Solids (TSS) concentration.

·Particle size analysers: the most common system available is the Coulter Counter. This equipment performs well but is delicate, has to be operated by expert personnel and is a non-continuous sensor.

·On-line particle size analysers based on laser optical systems have recently been introduced and allow to monitor, in real time, the residual solids content as a function of particle size.

·Turbidity metering: turbidity can be defined as the property of a liquid that causes it to scatter or absorb light. This is usually caused by fine particles suspended in the liquid. Nephelometric Turbidity Units (NTU) are typically used to measure turbidity. Nephelometry is the technique of beaming light on a sample and measuring the amount of light scattered at a certain angle, usually 90°. The correlation between NTU readings and solids content is a function of the particle size distribution. Hence NTU readings will only give an accurate indication of solids content when the particle size distribution is known and a proper calibration has been carried out. Calibration of turbidimeters is based on Formazin standard suspensions. Formazin is a polymer of uniform particle size and shape and is a reliable turbidity indicator when prepared properly. Turbidimeters are supplied with a reference turbidity standard. Turbidimeters are perfectly adequate for field applications and it is possible to monitor solids content continuously.

These fluid quality control methods are detailed in the Workover and Completions Fluid Manual Completion&sol;work-over fluid manual. A solids content specification of less than 10 mg/l of particles larger than 2 microns is currently the best that can be technically and operationally achieved under field conditions. This is generally considered to be adequate for completion operations even under conditions of moderate losses.

When circulating a well clean, a realistic solids content target can be defined by monitoring the fluid returns cleanliness using one of the methods above. The irreducible minimum can be taken as the solids content value which can only be marginally improved over lengthy circulating times. In most cases, 20 NTU is a good yardstick.

1.7.4 Carrier fluid rheology

1.7.4.1 General requirements

Various viscous fluid systems can be used for gravel packing purposes e.g. viscosified oils or water based fluids. A brine viscosified with a soluble polymer is the most commonly used system as it offers many claimed advantages over water:

1.High gravel transport capacity.

2.Control over fluid leak-off rate through viscosity variations.

3.Control over total leak-off through gravel loading.

4.Reduces the tendency to erode equipment and tubulars.

5.Leads to less intermixing of gravel and formation sand.

6.Internal chemical breakers can be added to break the viscosity after placement.

7.Less rig time, hence less costly.

8.Better diversion over completion interval.

Soluble polymers such as HEC, XC or Shellflo-S can be used to viscosify brines for gravel packing purposes. These polymer solutions exhibit a marked non-Newtonian shear thinning character i.e. the apparent viscosity decreases with increased shear rates all other conditions remaining constant. This is a desirable property for a carrier fluid as it gives good gravel transport properties and minimises the pressure losses when the slurry is pumped downhole. The rheology of these solutions follows a power-law model above a certain, generally very low, shear rate. For a power-law fluid, the shear stress t is related to the shear rate g.

Viscosified polymer solutions are viscoelastic, that is the viscosity is dependant on the instantaneous shear rate as well as the shear rate history. Viscous polymer solutions are also thermally degradable. Each polymer type shows its own thermal stability characteristics as shown in the following sections.

The carrier fluid rheology determines three basic properties:

·The gravel transport capacity.

·For an IGP, the fluid leak-off into the formation.

·The pressure gradients within the dynamic hydraulic system.

The relative importance of each of these properties will depend on the specific well conditions. In an IGP for example, the perforation packing process is chiefly governed by the leak-off properties of the carrier fluid. Conversely fluid leak-off is not essential to obtain a good gravel pack in open-hole. Gravel packing is a complicated process which is governed by many different factors. Fluid rheology is only one of the variables that has to be properly controlled to ensure good gravel pack placement. The optimal rheology of a carrier fluid for a given set of wellbore conditions is inherently difficult to define.

a. Gravel transport capacity

The gravel settling rate is a measure of the ability of the carrier fluid to transport gravel. Correlations for particle settling velocities in non-Newtonian fluids are generally modifications of Stokes law (see Section 4). Variables that play a role are the density difference between gravel and fluid, particle size, gravel concentration and apparent viscosity which in this case is a function of shear rate, temperature and breaker activity.

Gravel settling in a viscous carrier fluid occurs at very low shear rates which can be estimated by dividing the particle settling rate by its diameter. However carrier fluid viscosities at very low shear rates are of no value in measuring particle settling rates under dynamic conditions experienced in gravel pack jobs. Because gravel packing fluids are highly shear thinning, the apparent viscosity determined at a shear rate equivalent to that being imposed by fluid motion during placement is more realistic. Viscosity determinations at low shear rates are of value in determining the gravel settling rates after pumping has stopped and the slurry is in a static state.

b. Fluid leak-off

With viscous polymer solutions, fluid leak-off is controlled by two factors:

·The resistance to flow due to invasion of the near wellbore formation by a shear thinning, viscous fluid.

·The impairment caused by a gel layer build-up on the formation face.

The main controlling variables are the fluid properties at downhole conditions, the formation permeability and the overbalance exerted on the formation.

As the fluid flows out radially from the wellbore, the velocity and hence the shear rate decreases. Given the shear-thinning properties of polymer solutions, a decrease in shear rate results in an increase in fluid viscosity. According to Darcy's law, for a given pressure differential an increase in viscosity decreases the flow rate.

One of the problems of modelling fluid leak-off during gravel pack operations is that the fluid viscosity is a function of the shear and temperature history and also of the breaker activity. When using strong breakers such as HCl with HEC solutions, the viscosity may be severely degraded before the carrier fluid even reaches the formation.

Another uncertainty is the surface build-up of gel. Laboratory tests have shown that significant pressure drops can be caused by such surface deposits. This effect can be minimised by adhering to strict fluid quality control procedures, i.e. to shear and filter viscous gels. Carrier fluid leak-off properties have rarely been reported in the literature.

c. Friction pressure losses

When pumping the fluids the friction pressure drops will be determined by the velocities and shear rates encountered in the different parts of the hydraulic system.

For gravel packing operations, the slurry viscosity is a function of the carrier fluid viscosity and the gravel concentration. 

1.7.4.2 Viscosifiers

a. HEC polymer

Hydroxy-Ethyl-Cellulose (HEC) is the most widely used polymer in the gravel packing industry and it is currently the standard against which other polymer viscosifiers are judged. It is beyond the scope of the manual to discuss the many different HEC brands that are commercially available.

The apparent viscosity of the static fluid is quite high but rapidly decreases with shear rates of only a few sec-1. At high shear rates, the data fall on straight lines and the solutions follow the "power-law" model.

The main advantages of HEC are that it is non-ionic, which implies that its solubility is not affected by salts and that it is compatible with most chemical additives, and its viscosity can easily be broken with a variety of chemical breakers. Very little residue is claimed to be created when the polymer is decomposed by breakers. Commonly used breakers for HEC solutions are acids (HCl), oxidising agents (sodium persulphate) or enzymes. Enzymes can be used in wells with bottom-hole temperatures below 50°C.

HEC solutions will also loose their viscosity under prolonged exposure to elevated temperatures, typically a few days at a temperature of 80°C. Laboratory tests have shown however that flocs consisting of degraded polymer products may be formed which result in severe formation damage.

b. XC polymer

XC polymer is a water soluble, polysaccharide gum produced by microbial fermentation. Some operators and service companies have adopted a clarified XC polymer (commercial name Xanvis) as an alternative to HEC. XC solutions are more shear thinning than HEC solutions. Depending on the relative polymer concentrations, XC solutions can have a higher apparent viscosity at low shear rates and a lower viscosity at high shear rates than HEC solutions.

A higher viscosity at low shear rates gives better gravel suspension capacities and lower viscosity at high shear rates gives lower frictional pressure drops, hence lower pumping pressures.

In principle, because of its pronounced shear thinning character, better leak-off properties can be achieved while still maintaining good gravel suspension properties. This needs careful selection of the polymer concentration as a function of the specific borehole conditions. However these requirements have not yet been clearly defined. Field experience indicates that more gravel can be placed in the perforations when using XC instead of HEC polymer.

The disadvantages of XC polymer are that:

·It is more expensive than HEC.

·It is more difficult to mix properly.

·It was found to cause more permeability damage in cores than HEC.

·Available breakers are less effective and reliable than is the case with HEC.

c. Shellflo-S

Shellflo-S is a biopolymer viscosifier produced by microbial fermentation and marketed by Shell International Chemical Company. It is commercialised as a stabilised liquid concentrate. Unlike powdered viscosifiers, Shellflo-S has a viscosity building rate relatively insensitive to brine salinity, pH and temperature. Furthermore, provided that sufficient shear mixing is applied, Shellflo-S will disperse more rapidly than powdered polymer.

Like XC polymer solutions, Shellflo-S solutions show a more pronounced shear thinning character than HEC solutions. Also, by careful selection of the polymer concentration, better leak-off properties can be achieved while still maintaining good gravel suspension properties. Additional research however is still required to define the optimal rheology for given wellbore conditions.

Shellflo-S has a unique viscosity/temperature profile. Solutions show only a small decrease in viscosity with increasing temperature until a sudden loss of viscosity occurs at a critical transition temperature as a result of a change in molecular structure. Above the transition temperature, the viscosity of Shellflo-S declines to a few cP.

By tailoring the transition temperature to specific downhole conditions, the use of external chemical breakers can be avoided. This is seen as a major advantage due to the general operational problems and unreliability of chemical breakers. Hence a Shellflo-S solution with a transition temperature somewhat less than the reservoir temperature is ideal for gravel packing operations. During gravel packing, the solution temperature is lower than the static bottom hole temperature. Eventually the solution will reach the reservoir temperature and the viscosity will drop to several cP. Shellflo-S can then be back-produced with little or no formation impairment.

Shellflo-S's viscosity is not affected by temperatures below the Tm, contrary to HEC which exhibits rapidly decreasing viscosity with increasing temperature. Hence, in contrast to HEC, Shellflo-S maintains a relatively constant viscosity as it travels down the well (assuming it experiences a uniform shear regime), as long as its temperature is below the Tm.

Shellflo-S shows promise for gravel packing applications due to its unique rheological properties and good temperature degradability. Field trials have confirmed this. A disadvantage of Shellflo-S is that it is more expensive than HEC, although to a large extent this is offset by its ease of handling and high filterability.

d. Permpac

Dowell-Schlumberger introduced a micellar viscosifying system for gravel packing applications. The fluid is prepared by mixing a surfactant with brine. The surfactant molecules align with water molecules to form micelles which gives viscosity to the brine. Dowell Schlumberger claim superior gravel carrying capacity, leak-off characteristics and absence of fisheyes or microgels. Permpac is also claimed to loose its viscosity when exposed to temperatures in excess of some 50°C.

Permpac has been evaluated by the production chemistry laboratory. Although the carrying capacity and filterability properties of Permpac were confirmed, both operators experienced mixing problems during the preparation of the fluid and found differing viscosity breakback characteristics.

1.7.4.3 Recommended viscosifier concentration

The carrier fluid viscosity must be high enough to transport gravel under surface and downhole conditions but, in the case of an IGP, must be low enough to allow sufficient leak-off rates to pack the perforations. Downhole carrier fluid viscosity requirements especially are not well defined. One of the problems is the difficulty to predict the downhole properties of a polymer solution as a function of breaker activity, shear rate and temperature history, although this is less of a problem with Shellflo-S.

Research is ongoing to define the optimal carrier fluid rheology for a given set of wellbore conditions.

High gel viscosities are currently specified by the industry to provide adequate gravel carrying capacity on surface. However much lower viscosities can probably be tolerated downhole or may be beneficial for the gravel placement process.

Generally polymer concentrations are specified as a function of the slurry gravel mix ratio selected for the particular job. When HEC is used as a viscosifier, the following polymer concentration is recommended:

1.7.4.4 Breakers

a. Viscosity breakback criterion

Chemical breakers need to be incorporated in brines viscosified with HEC or XC polymer solutions in order to effectively breakdown the viscosity and prevent formation impairment. Chemical breakers act by acidic hydrolysis or oxidative breakdown of the polymer chain. The proper type and concentration of breaker for a gelled brine must be determined by laboratory experiments simulating the temperature profile of the fluid during the gravel packing operation and by specifying a minimum reduction of initial fluid viscosity within a pre-determined period. The apparent viscosity breakback time is a function of the breaker type and concentration, temperature and shear rate at which it is measured. It is essential to establish breaker dosage requirements prior to each job as breaker performance is dependant on the shelf life of the products and possibly varies significantly from batch to batch.

Viscosity breakback criteria are not properly defined in the industry. Based on field experience, the recommended criterion is that the apparent viscosity should be broken to 10% of its original value one hour after the slurry is placed. Adequate conditions are thus given for the gravel pack to settle and loss of screen out can be detected immediately after the gravel pack job. Extra gravel can then be placed if necessary.

In practice, viscosified brines will begin losing viscosity immediately after the breaker is added, and continue to slowly break with time and temperature. This is a complicating factor when trying to mathematically model a gravel packing process.

With Shellflo-S the use of external chemical breakers can be avoided by tailoring the transition temperature to the specific downhole conditions. This is seen as a major advantage due to the general operational problems and unreliability of chemical breakers.

b. Breakers for HEC solutions

The viscosity of an HEC solution can be broken by acids, oxydants or enzymes. An acid breaker is preferred because control of the breakback time with acid is less sensitive to temperature and concentration than with other type of breakers. In addition, powdered agents such as persulphates and enzymes may suffer from a limited shelf life resulting in reduced breaking activity.

The following guidelines should be observed for the selection of a breaker:

·Acid breakers should be preferred to other types of breakers e.g. oxidative breakers and enzyme breakers. Acid breakers should not be used when gravel packing carbonate rich formations, when using resin coated gravel systems, and in the case of low reservoir temperatures.

·If an acid breaker is acceptable, HCl breaker of any acid strength can be used. Alternative acid breakers to consider, in order of decreasing strength are 1M formic acid + 0.1MHCl, 1M formic acid, and 1M acetic acid. These breakers cover a wide range of applications.

·Enzymes can be used below 50°C and have an optimum activity at 30°C. Further advice can be obtained from SICC London.

c. Breakers for Shellflo-S

The main advantage of Shellflo-S is that the use of external breakers can be avoided by adjusting the transition temperature somewhat below the static bottom hole temperature. This transition temperature (Tm) can be adjusted from 40 to over 100°C by changing the type and concentration of salts added to the solution. Laboratory experiments backed up by a field trial have shown that it is possible to reduce the Tm by addition of urea. Starting with a base brine of 2% KCl (Tm = 65°C) each additional 10% w/v of urea reduces the Tm about 5°C. The viscosity of the solution is also reduced almost linearly: each additional 10% w/v of urea reduces the viscosity of the solution by some 10%. At concentrations beyond 45% w/v this behaviour becomes non-linear, though still usable. Because of the reduction in viscosity on addition of urea an increased concentration of Shell Flo-S should be used to ensure the viscosity requirements are met.

1.7.4.5 Formation damage caused by viscous polymer solutions

When first introduced in the oil field, HEC gels were assumed to be totally non-damaging to most reservoirs and this may be true under laboratory conditions. However, under field conditions, great care is required when preparing HEC solutions to avoid the formation of small accumulations of partially hydrated polymer which can have a very broad size range, from microns to centimetres. These microgels or "fisheyes" are almost impossible to remove (acidise) and will result in permanent formation and pack impairment. It is therefore recommended to shear and filter all polymer viscosified brines be sheared and filtered in order to remove any particulate matter.

The utmost care needs to be exerted when preparing viscous polymer solutions. Field experience suggests that exposure of the formation to HEC or other polymer solutions can cause permeability impairment even when they are adequately mixed, sheared, filtered and include the proper breaker system. Damage caused by HEC can at best be partially removed. Residues left over from broken polymer are difficult to acidise whilst fisheyes or microgels are unlikely to be removed by acids or enzymes.

The presence of ferric ions in the wellbore may lead to cross linking of Shellflo-S, giving a sawdust like precipitate. To avoid this serious problem either sequestering agents, such as citric acid or iron reducing agents, such as erythorbic acid should be applied.

1.8 Slurry design

This section covers the design of a gravel pack slurry and the following aspects are addressed:

·Slurry composition, i.e. gravel concentration and rheology.

·Slurry volumes.

·Spacer or pad composition and volumes.

The slurry design ultimately depends on the type of gravel pack (IGP or EGP) and the specific downhole conditions.

1.8.1 Slurry types

1.8.1.1 Non-viscosified carrier fluids (conventional packing)

Over the last two decades the industry has seen a trend towards viscous slurry packing (for IGP's mainly) for a variety reasons including:

1.Quicker operations, hence less exposure to losses.

2.Better placement into perforation tunnels.

3.Better gravel transport and tighter higher quality packs.

4.Reduce fluid losses.

However recent success with gravel packing high angle intervals has led to the re-surgence of low gravel concentration, low viscosity gravel packing.

Water or non-viscous brines were the first carrier fluids used for gravel packing, hence the name conventional water packing. Until recently, the gravel concentration was limited to about 2 lb/gal maximum, but in practice often limited to about 0.5 lb/gal. This generally required only the use of a gravel pot and injection pump, where gravel is essentially added and pumped on the fly. Such a set up provides little control over gravel placement and limited the average gravel concentration. For similar reasons non-viscosified brines are not suited for transporting gravel into perforation tunnels and hence are normally, only considered for open hole applications, or packing the annulus after perforation prepacking operations.

a. Recent developments

The increasing application of water based, low viscosity systems has been assisted by:

1.The ability to filter completion brines typically using a combination of diatomaceous earth (DE) in combination with absolute cartridge filtration systems, and

2.The development of superior equipment for blending and pumping low concentration slurries.

Conventional water packing methods are in various areas, proving to be cost effective alternatives to gravel packing open (and cased) hole completion intervals:

·Gravel packing low permeability intervals,

·Highly deviated and horizontal wells,

and routinely to carry out conventional "top ups" when i.e. screen-out cannot be re-confirmed. 

1.8.1.2 Viscosified carrier fluids (slurry packing)

The concept of "slurry packing" was originally developed by Sparlin in the early 1970s to avoid the permeability reduction associated with formation sand/gravel mixing. A viscous oil-based fluid was used to transport high gravel concentrations (up to 15 lb gravel per gallon of fluid) and pump the slurry through the perforations at slow rates. The "Aquapack" system which uses a brine viscosified with HEC as a carrier fluid was designed by Shell Oil also in the early seventies.

The use of a viscous, water based carrier fluid with high gravel concentrations has become the most popular gravel packing method in the industry and the name "slurry packing" was adopted to designate this technique. The most popular fluid system is brine viscosified with a HEC polymer. Alternative viscosifier systems such as XC polymers and Shellflo-S are being evaluated and field tested.

Slurry packing was claimed to offer the following advantages over conventional packing:

·The gravel is less likely to mix with formation sand in the perforation tunnels due to the viscosity of the fluid and the lower transport velocities. As discussed previously, gravel mixing with formation sand can be a major source of impairment.

·More effective placement of gravel into the perforations.

·Reduced fluid loss to the formation and hence less risk of formation damage.

·The gravel is less susceptible to damage during pumping (crushing in pumps).

·Easier handling of the slurry on surface.

·Jobs are shorter because of the smaller fluid volumes to be pumped.

However field experience, has proven that the slurry packing technique has not exactly lived up to all its promises.

1.8.2 Guidelines for slurry design

Brine based viscosified carrier fluids have seen wide application for over 15 years. The main parameters for the slurry design are the gravel concentration and viscosifying polymer concentration.

In the past high gravel concentrations (15 lb/gal or 1800 kg/m3) together with high polymer concentrations (80 lb/Mgal or 9.6 kg/m3 HEC) was the standard slurry design used by many operators. This slurry is extremely heavy and viscous and in the case of IGPs, is likely to result in poor perforation packing because of restricted fluid leak-off and poor control of the packing operations.

The recommended standard formulation is an intermediate viscosity carrier fluid and gravel loading i.e. 65 lb/Mgal (7.8 kg/m3) HEC and 10-12 lb/gal (1200-1440 kg/m3) gravel mix ratio. Beside being less heavy and viscous, this slurry formulation presents less impairment potential due to the lower polymer concentrations.

Field experience and full scale gravel pack models have shown that this standard slurry design must be further tailored to the specific wellbore conditions. This is particularly true for "difficult" applications i.e. when any of the following conditions are met:

·Highly deviated zones (> 50-60 degrees)

·Long intervals (> 50 feet)

·Zones with a high permeability contrast

Studies have shown that satisfactory gravel transport and improved packing can be achieved in difficult wells with lower gravel concentrations and lower carrier viscosity fluids.

Gravel concentration recommendations for the different types of gravel packs are given in Section 2: Internal Gravel Packing and Section 3: External Gravel Packing.

1.8.3 Slurry properties

1.8.3.1 Slurry density

The gravel loading has a significant impact on the slurry density. This parameter governs the overbalance on the formation when pumping the slurry. High gravel concentrations may lead to a large overbalance and loss of circulation which may lead to poor pack placement. 

1.8.3.2 Slurry viscosity

It is very difficult to predict the viscosity of HEC based slurries as it is a function of fluid composition, flow characteristics, temperature, gravel concentration and breaker activity. This is a major problem when trying to model the gravel packing process.

For Shellflo-S based slurries the temperature and breaker activity do not substantially effect the viscosity until the gravel is placed and modelling is much simpler leading to better job control.

Flow loop experiments have shown that the slurry viscosity can be significantly higher than the original carrier fluid viscosity, especially for gravel concentrations higher than 10 lb/gal. Experimental correlations describing the effect of gravel concentration on slurry viscosity are discussed in the well stimulation manual.

1.8.4 Spacer or pad design

Viscous spacers or pads are generally pumped ahead of the gravel slurry to prevent roping of the slurry through the completion brine. This allows for better volumetric control of the placement operations. Another function of the pad can be to control the injectivity profile of the completion interval, but this aspect is rarely documented by operators.

Unless used to control the injectivity of the well prior to gravel packing (e.g. to obtain circulation) the pad volume should be kept to a minimum, i.e. 300 feet of work string capacity. A large viscous prepad may be required in reservoirs prone to total losses to ensure that controllable partial returns are obtained during gravel placement.

Pads should be non-damaging and the same requirements as for the carrier fluid are applicable. It is operationally convenient to keep a small batch of the carrier fluid for use as a spacer. The pad should be made more viscous than the carrier fluid by adding some additional polymer.

1.9 General gravel packing procedures

This section provides examples of recommended general procedures for gravel packing operations. Operators should tailor these programmes to their specific needs. Procedures 

1.9.1 Well preparation before perforating or underreaming

Wellbore cleaning before completion fluids can contact the formation is a critical aspect to minimise formation damage. Any source of damage that can be avoided should be avoided.

Wellbore cleaning is usually carried out by displacement of the drilling mud in the well by seawater or clean completion fluid and using a variety of pills in between. The casing should be scraped prior to perforating or underreaming

1.9.2 Assembling and running the tools

1.9.2.1 Gravel pack liner assembly

·Assembling the gravel pack tools is the contractor's responsibility but he should be given sufficient time to carry out the job properly.

·All tools should be thoroughly cleaned with e.g. a steam cleaner.

·No painted tools should be allowed to enter the wellbore. Paint flakes can very effectively plug perforations.

·Pipe dope must not be used in permanent connections, i.e. anything that stays in the hole.

1.9.2.2 Work string cleanliness

A critical requirement is to ensure that the work string is thoroughly cleaned prior to gravel packing. The best way is to clean the string joint by joint with a steam cleaner prior to running in the hole, followed by an acid wash (pickling). Pickling the work string prior to gravel packing is an efficient method for removing pipe dope, scale, rust or deposits from inside the work string. Aromatic solvents work best for removing pipe dope but create safety and environmental problems. The optimum way of pickling a work string consists in using an acid pill in combination with gravel.

Pipe dope is a perennial problem in gravel pack operations; it is an efficient impairment agent and is very difficult to remove by remedial methods. Limited application of pipe dope is a difficult operation to control, even when the rig crew has been provided with only toothbrushes! The requirement for pipe dope can be eliminated by employing a dedicated work string with special metallic thread coating to prevent galling. Additional rig time is however required to pick-up a dedicated work string. If pipe dope cannot be avoided, then automated pipe dope applicators are strongly recommended in order to minimise the total quantity used.

2 Internal gravel packing

2.1 Perforating considerations

Field experience and laboratory studies indicate that perforations are a primary source of well performance problems, especially in gravel packed completions. Under certain conditions perforation tunnels can cause a very high pressure drop for produced or injected fluids. In comparison, the pressure drop through the screen and the gravel sheath in the screen/casing annulus is relatively insignificant.

2.1.1 Pressure drop in a gravel packed perforation tunnel

A gravel packed perforation tunnel can be modelled as a linear flow cell filled with porous material. The exact geometry of the flow cell is difficult to ascertain as the flow pattern in the perforations and the nearby formation is a combination of linear and radial flow. The length of the linear flow cell is a matter of judgement but the minimum length should be the sum of the casing wall and cement sheath thickness and a term related to formation damage 

·Turbulent flow effects can cause a significant pressure drop in addition to the laminar (Darcy) pressure drop component.

·Formation sand can effectively plug a perforation tunnel.

·Perforation diameter and flow rate are the main controlling variables for the pressure drop.

·Gravel size plays a less significant role due to the high permeability contrast with formation sand.

Although the calculations above consider turbulent flow, they are still conservative because they do not consider multiphase flow, wettablity and relative permeability effects.

Mixing of gravel and formation sand leads to reduced permeabilities and impaired gravel packs.

2.1.2 Perforation geometry

The four basic perforation design variables have the following general order of importance for cased hole gravel packs:

·Perforation diameter (provided a threshold penetration is exceeded).

·Effective shot density.

·Penetration depth beyond a damage defined threshold level.

·Gun phasing.

The benefits that can be obtained from a reduction of the flow velocity in a perforation by using large diameter perforations and high shot densities. Modern perforating guns enable perforations to be made with an entry hole diameter up to 1.25 inch and a density up to 12 shots per foot. Some guns allow for even higher shot densities. However the perforation diameter can only be maximised at the expense of penetration depth as there is a limit to the energy available in a shaped charge. The other constraint is of course the cost of the perforating operation.

2.1.2.1 Penetration depth

Perforation penetration beyond the damaged zone i.e. the zone invaded and damaged by the drilling or completion fluid, is of particular importance for productivity. As a rule, deep penetrating charges should be selected whenever deep formation damage is suspected (indicated by mud losses when drilling through the reservoir) or when the borehole has been severely washed out. In all other cases, the perforation diameter should be maximised at the expense of total penetration.

2.1.2.2 Perforation diameter

Whenever possible the perforation diameter should be maximised to reduce the pressure drop in the gravel packed perforation tunnels when the well is on production. Entry hole sizes up to 1.25 inch can be obtained with modern perforating guns. When the specific production conditions are known, a pressure drop analysis can assist in the selection of the optimal entry hole size.

2.1.2.3 Effective shot density

In theory, a non-gravel packed completion with four perforations per foot should give the same productivity as an open-hole completion. In a cased-hole gravel packed completion however, only a fraction of the perforations may be open to flow with the rest being impaired, plugged with formation sand or a mix of gravel and sand. To overcome these effects, the highest possible shot density should be considered. Field experience should establish whether the maximum perforation density is justified in view of the additional costs involved. In any case, intervals to be gravel packed should be perforated with a minimum of 6-9 shots per foot.

2.1.2.4 Gun phasing

For normal applications (deviation < 60 degrees), the perforation pattern should cover the complete circumference of the borehole. This minimises the geometrical skin component and ensures homogeneous placement of the pack as dehydration of the slurry occurs in all directions around the casing/screen annulus.

In highly deviated and especially in horizontal wells, it is debatable whether perforating should be limited to a fraction of the lower hemicycle. The main argument in favour of this practice is the difficulty to ensure complete packing of the annulus and hence to achieve effective sand control. Low side perforation minimises the risk of gun sticking due to sand ingress in the wellbore when shooting but reduces the chance of tightly packing the high side of the annulus.

2.1.3 Perforating method

The major concern when perforating prior to gravel packing is to effectively clean the perforation tunnels i.e. to evacuate all perforation debris and to remove the "crushed" zone. Gravel packing will trap any debris and damaged reservoir rock which is left behind after perforation cleaning. This results in permanent impairment as it is very difficult to remove this material.

Perforating guns can either be tubing conveyed or wireline conveyed. The superiority of one or the other method has not yet been demonstrated in terms of well productivity, especially in a gravel packing context. Hence the choice of the perforating method should be based on operational and cost considerations in addition to technical considerations.

2.1.3.1 Tubing Conveyed Perforating (TCP)

Tubing conveyed perforating enables long intervals to be shot with underbalance in one run and to backsurge and backflow immediately after perforating in one run, thus eliminating the need for subsequent perforation washing. TCP also enables the inlet flow area to be maximised by using high shot densities (up to 12 spf) together with large entry hole sizes that minimise pressure drops across gravel filled perforations. TCP also enables the well to be back flowed and killed through the same work string.

The cleaning efficiency of underbalance perforating depends on the surge profile which governs the rate and volume of fluid flow through the perforations. The surge profile is controlled by several factors: reservoir parameters, wellbore geometry, underbalance pressure differential and underbalancing method.

It is recommended that underbalance is created by using air, nitrogen or natural gas instead of a liquid e.g. diesel. A compressible wellbore system results in a much longer flow period exposed to the underbalance pressure before equalisation and increases the effectiveness of perforation surging.

2.1.3.2 Wireline perforating

Wireline perforating can be performed with the wellbore fluid over or under balancing the formation pressure. When perforating with overbalance, there is no mechanism for removing debris generated during the perforating process. In wells which do not require sand exclusion, much of this material is removed during the early production life of the well. Prior to gravel packing, perforations need to be cleaned by washing or backsurging.

a. Perforation washing

The objective of perforation washing is to break down and remove the formation between perforations. A cavity can be created behind the casing which can then be packed with highly permeable gravel. However invasion of fines generated when washing and fluid losses increase the risk of permeability impairment. Perforation washing tools rely on opposed swab cups to isolate sets of perforations and establish a circulation path behind the casing.

The efficiency of the washing operation can be measured by the perforation wash factor which is defined as the volume of sand removed per unit length of perforated interval. Wash factors measured in the field can vary considerably, from 0.02 up to 1 to 4 ft3/ft (or 0.002 to 0.4 m3/m).

The main controlling variables for perforation washing are:

·Pump rates: The amount of formation sand removed by perforation washing is largely dictated by the wash rates that can be achieved. Rates of up to 8-10 bbl/min may be required for washing to be effective. The flow rate required to breakdown the near wellbore formation and to create a cavity is a function of the specific wellbore conditions and has to be established locally. If this flowrate cannot be achieved then acidisation should be considered.

·Fluids: Brines are more effective in washing perforations than viscosified fluids. Fluids should be clean and non-damaging to the formation.

·Washing pressure: It is important to know the pressure drop through the tool in order to calculate the pressure to which the formation is exposed.

b. Perforation backsurging

Backsurging is a perforation cleaning method favoured by various operators especially in the Gulf of Mexico. The method uses a sudden pressure underbalance to move fluids into the wellbore and flush debris from perforation cavities and tunnels. The potential for plugging is theoretically less than with washing.

The "backsurge" tool consists of a packer and a backsurge chamber isolated by two valves. The tool is operated by rapidly opening the bottom valve to expose the perforated interval to near atmospheric pressure. Penberthy has shown in large scale model testing that perforation surging is capable of removing about the same quantity of formation sand per foot as perforation washing. However perforation surging may not open all the perforations and the amount of formation material removed per foot is erratic.

The main controlling variables for perforation surging are:

·Chamber volume: The efficiency of the backsurging operation is mainly a function of the chamber volume used. Typical Operator practices can vary from 1 to 10 ft3 per foot of perforations (0.1 to 1 m3 per meter).

·Pressure differential: Backsurging utilises atmospheric pressure and high pressure differentials can result. Excessive pressure differentials can collapse perforation tunnels.

2.1.3.3 Recommended perforating method

Underbalance perforating with a tubing conveyed gun is generally the preferred method of preparing a cased hole for gravel packing as it is often the quickest method (depending on the length of the interval) and supposedly gives the best results. Research and field studies indicate that underbalanced perforating is more effective in cleaning perforations than backsurging or washing, but this may be a function of the type of reservoir rock. Underbalance perforating is also possible with wireline guns.

Perforating overbalance has to be followed by perforation washing or surging. Perforation washing can potentially remove more formation damage from the wellbore vicinity in an unconsolidated sand than underbalanced perforating or surging. However washing causes contamination of the completion fluid and may cause impairment with clays and fines.

Underbalancing at the instant of perforation appears to be the critical difference between underbalanced perforating and perforation surging. The true merits of each method are still a matter of debate and local experience is essential for predicting the effects, if any, on productivity. With lack of proven technical superiority of a particular method, the economics of perforating are an important consideration when selecting a perforation method. Cost comparisons between different perforating methods should take into account the rig time spent for washing or surging operations in addition to the time spent for perforating. Tubing conveyed perforating tends to be more expensive than wireline perforating and washing, unless the intervals happen to be relatively long and deep.

2.2 Gravel placement considerations

2.2.1 The ideal IGP

The requirements for a successful internal gravel pack are:

·To tightly pack the perforations with gravel. This is the most important requirement as it prevents formation sand movement and creates a highly permeable flow path for the produced fluids. A perforation tunnel that is not packed may collapse when the well is on production. This perforation will then be impaired due to the high pressure drop caused by the presence of formation sand.

·To maintain a clean gravel/sand interface to avoid permeability impairment associated with mixing of gravel and sand.

·To place a dense homogeneous gravel sheath around the screen in order to prevent backflow of the gravel placed in the perforations and control formation sand opposite poorly packed perforations.

·To tightly pack the perforations with gravel. This is the most important requirement as it prevents formation sand movement and creates a highly permeable flow path for the produced fluids. A perforation tunnel that is not packed may collapse when the well is on production. This perforation will then be impaired due to the high pressure drop caused by the presence of formation sand.

·To maintain a clean gravel/sand interface to avoid permeability impairment associated with mixing of gravel and sand.

·To place a dense homogeneous gravel sheath around the screen in order to prevent backflow of the gravel placed in the perforations and control formation sand opposite poorly packed perforations.

·To ensure that the near wellbore formation remains unimpaired by drilling and gravel packing operations.

Good gravel placement can be ensured by proper control of all of the following variables:

·Mechanical variables i.e. equipment selection: screen size, washpipe dimensions.

·Hydraulic variables i.e. the slurry design: rheology, gravel concentration, viscosity breakback.

·Operational variables i.e. circulating or squeezing, leak-off rate, pump rates and pressures.

All of these variables have to be given the same level of attention!

2.2.2 Fluid flowpaths

The slurry can flow through the annulus and into the perforations. The carrier fluid can leak-off into the formation or through the screen and flow the washpipe/screen annulus. The flow distribution is governed by the frictional pressure losses in the dynamic hydraulic system.

2.2.3 Factors affecting gravel placement

2.2.3.1 Factors affecting gravel placement in perforations

Gravel is deposited as the slurry flows into the perforations and dehydrates into the formation. The amount of gravel placed in the perforations is proportional to the volume of slurry pumped into the perforations. Fluid leak-off is controlled by the following factors:

·The overbalance on the reservoir pressure which is a function of the slurry hydrostatic head and operational variables such as the pump rate and the control or choking of fluid returns.

·The formation and perforation zone permeability which can vary significantly over the perforated interval.

·The geometry of the perforation tunnels which governs velocities and frictional pressure drops.

·The carrier fluid rheology. Low viscosity promotes fluid leak-off but a minimum viscosity is required to avoid gravel drop-out over the perforation tunnel length.

The perforation packing efficiency can be defined as the amount of gravel placed in the perforation per unit volume of slurry pumped in the perforation and is a function of the following variables.

·The fluid carrying capacity which determines how efficiently gravel is transported "round the bend" into the perforation tunnel. This is mainly a function of viscosity, the difference between gravel and fluid density and the gravel grain size.

·The gravel mix ratio.

·The inclination of the perforation tunnel. In highly deviated wells, gravitational forces will impede the placement of gravel in perforations situated on the high side of the wellbore.

A high perforation packing efficiency is desirable in order to reduce the fluid volume lost in the formation and thus minimise formation damage. The importance of minimising fluid loss is a function of the impairment potential of the carrier fluid.

Fluid rheology plays a dual conflicting role (leak-off and carrying capacity) in the perforation packing process. An optimal rheology must exist for a given set of wellbore conditions to pack the perforations while minimising impairment. This optimum has not been defined yet and further research is required.

Modelling the gravel placement process is a complex problem because many factors are difficult to predict. The slurry rheology of HEC and XC polymer solutions under downhole conditions is a function of temperature and shear rate history and the breaker activity. These problems are much reduced with Shell Flo-S. Similarly, the effective formation permeability is difficult to determine as it is a function of the perforating process and the perforation cleaning method used. The flow of a viscous fluid into the formation also influences the leak-off process.

Tightly packed perforation tunnels are crucial for the productivity of the completion. The main obstacle to adequate perforation packing may however be the interaction with the annular packing process.

The volume of gravel placed behind casing can thus be considered as a gravel pack quality indicator although in general the average cavity will not be well characterised. In practice, this volume can only be indirectly determined. Gravel pack logging tools are inadequate for this purpose and the volumetric balance methods are generally inaccurate.

2.2.3.2 Factors affecting gravel placement in the annulus

Gravel transport in the annulus is mainly a function of carrier fluid rheology, flow velocities and wellbore inclination.

Deviation has a major impact on the gravel placement process in the annulus. In vertical and moderately deviated wells (< 60 degrees), gravity helps in packing the annulus from bottom-up. Full annular gravel placement is relatively easily achieved. If bridges are formed in the annulus, these are likely to collapse as pumping stops and the differential pressure is removed. Gravel settling will then prevent formation of voids if adequate gravel reserve is available. In these wells, fluid rheology and gravel concentration are not critical parameters. It is generally accepted that high gravel concentrations and high viscosity fluids can substantially increase the risk of bridging especially in a small annulus.

In highly deviated wells (higher than about 60 degrees) the gravel transport mode changes significantly. Gravel will fall out of the slurry as it exits the crossover port and will form a bed. The equilibrium bed height is a function of fluid properties, gravel concentration and fluid velocity. This bed will progress along the wellbore if adequate circulation rates and flow paths are maintained along the annulus and bridging does not occur. When the end of the screen is reached, a dune propagating in the opposite direction will fill the annular void on the high side of the hole. Slurry design and flow parameters are critical factors in these wells and there is a high risk of not achieving a complete annular pack. Voids formed in the screen/casing annulus below a bridge will subsist after pumping stops because friction between the gravel grains prevents downwards settling of gravel.

The gravel pack in the annulus can be qualitatively evaluated with the help of logging tools which can detect top gravel and voids, depending on their size and shape.

2.2.3.3 Interaction between annular and perforation packing

Since we are generally successful in packing the annulus, attention must focus on properly packing the perforation tunnels. The problem is that the annular packing process interferes with the placement of gravel in the perforations. In fact it is probably very difficult, if not impossible, to achieve complete packing of all perforations and the annulus simultaneously using current gravel packing procedures.

If the annulus is allowed to pack too rapidly, then the gravel will pack-off the perforation entrance before the tunnel is fully packed.

All of these variables have to be given the same level of attention!

2.2.4 Fluid flowpaths

Different fluid flowpaths are available when gravel packing. The slurry can flow through the annulus and into the perforations. The carrier fluid can leak-off into the formation or through the screen and flow the washpipe/screen annulus. The flow distribution is governed by the frictional pressure losses in the dynamic hydraulic system.

Lower tell-tales (LTT) are commonly used to force slurry circulation to the bottom of the annulus. Flow must enter through the LTT screen as the lower end of the washpipe is sealed into a sealbore. There is a more subtle fluid flowpath that may defeat the effectiveness of a LTT. The carrier fluid can bypass into the washpipe/screen annulus, re-enter at the bottom of the casing/screen annulus to flow through the LTT. The most practical way to ensure that slurry flow is forced to the bottom of the annulus is to minimise the screen/washpipe clearance.

2.2.5 Factors affecting gravel placement

2.2.5.1 Factors affecting gravel placement in perforations

Gravel is deposited as the slurry flows into the perforations and dehydrates into the formation. The amount of gravel placed in the perforations is proportional to the volume of slurry pumped into the perforations. Fluid leak-off is controlled by the following factors:

·The overbalance on the reservoir pressure which is a function of the slurry hydrostatic head and operational variables such as the pump rate and the control or choking of fluid returns.

·The formation and perforation zone permeability which can vary significantly over the perforated interval.

·The geometry of the perforation tunnels which governs velocities and frictional pressure drops.

·The carrier fluid rheology. Low viscosity promotes fluid leak-off but a minimum viscosity is required to avoid gravel drop-out over the perforation tunnel length.

The perforation packing efficiency can be defined as the amount of gravel placed in the perforation per unit volume of slurry pumped in the perforation and is a function of the following variables:

·The fluid carrying capacity which determines how efficiently gravel is transported "round the bend" into the perforation tunnel. This is mainly a function of viscosity, the difference between gravel and fluid density and the gravel grain size.

·The gravel mix ratio.

·The inclination of the perforation tunnel. In highly deviated wells, gravitational forces will impede the placement of gravel in perforations situated on the high side of the wellbore.

A high perforation packing efficiency is desirable in order to reduce the fluid volume lost in the formation and thus minimise formation damage. The importance of minimising fluid loss is a function of the impairment potential of the carrier fluid.

Fluid rheology plays a dual conflicting role (leak-off and carrying capacity) in the perforation packing process. An optimal rheology must exist for a given set of wellbore conditions to pack the perforations while minimising impairment. This optimum has not been defined yet and further research is required.

Modelling the gravel placement process is a complex problem because many factors are difficult to predict. The slurry rheology of HEC and XC polymer solutions under downhole conditions is a function of temperature and shear rate history and the breaker activity. These problems are much reduced with Shell Flo-S. Similarly, the effective formation permeability is difficult to determine as it is a function of the perforating process and the perforation cleaning method used. The flow of a viscous fluid into the formation also influences the leak-off process.

Tightly packed perforation tunnels are crucial for the productivity of the completion. The main obstacle to adequate perforation packing may however be the interaction with the annular packing process.

The volume of gravel placed behind casing can thus be considered as a gravel pack quality indicator although in general the average cavity will not be well characterised. In practice, this volume can only be indirectly determined. Gravel pack logging tools are inadequate for this purpose and the volumetric balance methods are generally inaccurate.

2.2.5.2 Factors affecting gravel placement in the annulus

Gravel transport in the annulus is mainly a function of carrier fluid rheology, flow velocities and wellbore inclination.

Deviation has a major impact on the gravel placement process in the annulus. In vertical and moderately deviated wells (< 60 degrees), gravity helps in packing the annulus from bottom-up. Full annular gravel placement is relatively easily achieved. If bridges are formed in the annulus, these are likely to collapse as pumping stops and the differential pressure is removed. Gravel settling will then prevent formation of voids if adequate gravel reserve is available. In these wells, fluid rheology and gravel concentration are not critical parameters. It is generally accepted that high gravel concentrations and high viscosity fluids can substantially increase the risk of bridging especially in a small annulus.

In highly deviated wells (higher than about 60 degrees) the gravel transport mode changes significantly. Gravel will fall out of the slurry as it exits the crossover port and will form a bed. The equilibrium bed height is a function of fluid properties, gravel concentration and fluid velocity. This bed will progress along the wellbore if adequate circulation rates and flow paths are maintained along the annulus and bridging does not occur. When the end of the screen is reached, a dune propagating in the opposite direction will fill the annular void on the high side of the hole. Slurry design and flow parameters are critical factors in these wells and there is a high risk of not achieving a complete annular pack. Voids formed in the screen/casing annulus below a bridge will subsist after pumping stops because friction between the gravel grains prevents downwards settling of gravel.

The gravel pack in the annulus can be qualitatively evaluated with the help of logging tools which can detect top gravel and voids, depending on their size and shape.

2.2.5.3 Interaction between annular and perforation packing

Since we are generally successful in packing the annulus, attention must focus on properly packing the perforation tunnels. The problem is that the annular packing process interferes with the placement of gravel in the perforations. In fact it is probably very difficult, if not impossible, to achieve complete packing of all perforations and the annulus simultaneously using current gravel packing procedures.

If the annulus is allowed to pack too rapidly, then the gravel will pack-off the perforation entrance before the tunnel is fully packed

The leak-off rate must be high enough to ensure that the perforations are packed before the screen/casing annulus is filled. Conversely, if excessive leak-off occurs, the gravel may prematurely bridge in the annulus especially across high permeability streaks. Premature bridging is not always obvious in vertical wells because the bridge may collapse as soon as pumping ceases. Perforations below the bridge will however remain unpacked.

The correct fluid leak-off to fluid returns ratio is a matter of debate. Surprisingly this critical topic has never been properly researched and no specific guidelines can be given. Ideally, all perforations should be fully packed before gravel in the annulus screens out the perforation entrance. In a vertical well, the annulus normally packs from bottom-up but the perforation packing order will be a function of the effective leak-off in each perforation and this is a self diverting process. Progress has been hampered by the lack of reliable logging methods to measure cavity geometry and gravel volume behind the casing, and the lack of realistic gravel packing numerical and physical simulators.

An optimal pumping schedule would aim to preferentially fill the perforations in the early stages of the job (while still maintaining slurry flow to the bottom of the annulus) and complete the annular pack at the end of the job. In other words, one would aim for a high leak-off rate at the beginning of the job while gradually increasing fluid returns as the job progresses. In practice it is difficult to effectively control operational variables. When high gravel/gel mix ratios are used (> 12 lb/gal) the pack may be complete in a few minutes time, leaving little opportunity to adjust operational variables. Furthermore, significant time lags are inherent to the hydraulic system. Hence optimal gravel pack placement is only seen possible through automation of the gravel placement process. This requires a step change in oil-field pumping equipment.

Alternatively, the perforation packing process can be separated from the annular packing process i.e. prepacking. This can be achieved by squeezing gravel into the perforations prior to running the screen assembly. Prepacking techniques are considered to offer the best scope for improving perforation packing efficiency.

2.2.5.4 Effect of interval length and deviation

The maximum length of interval that can be gravel packed will vary with the wellbore conditions. In general, as the deviation angle increases it becomes more difficult to pack a long interval. Reservoir heterogeneity increases the difficulty of gravel packing long intervals. Moreover perforation cleaning of a long interval by under balanced perforating may be ineffective.

Although the amount of gravel placed behind perforations cannot be measured accurately, field experience suggests that perforation packing effectiveness declines rapidly with interval length. Some authors go so far as to suggest that perforation packing is very poor for intervals longer than 10 feet, even in vertical wells as shown.

The design of a gravel pack for long intervals, especially when highly deviated, is typically chosen to promote annular fill at the expense of perforation fill. The most efficient way to obtain good annular packing is to size the washpipe correctly. The ratio of washpipe OD to screen ID should be at least 0.8 in long and/or highly deviated wells. However, as the screen/casing annulus packs from bottom-up the carrier fluid is forced to travel down the washpipe/screen annulus. The pressure required to flow down this annulus increases as the job progresses and may become excessive in long intervals. Hence a long zone may be required to be broken down in shorter intervals. There are however no specific guidelines as to what is the maximum allowable length.

As the deviation increases, gravity will prevent filling of voids left in the annulus after the job and it becomes essential to avoid bridging. Hence annular packing should be promoted at the expense of perforation packing. If necessary, remedial action should be taken following a gravel pack log.

2.2.6 Gravel placement methods

2.2.6.1 Squeeze pack

In a squeeze pack, the slurry is forced to dehydrate in the perforation tunnels thereby ensuring gravel placement in cavities behind the casing. This can be achieved either by placing the gravel pack tool in the squeeze position or by closing-off the returns.

A squeeze pack promotes a radial build-up of the gravel pack inside the screen/casing annulus as gravel nodes build-up on the packed perforation entrances, with a consequent high risk of bridging. Non uniform injectivity profiles increase the risk of bridging as the slurry preferentially dehydrates opposite high permeability zones. Bridges may not be so problematic in vertical wells as they may collapse as soon as pumping stops. However the perforations below the bridge run a high risk of being poorly packed. In deviated wells and/or long intervals however, neither the perforations nor the annulus below the bridge will be adequately packed.

Also in a squeeze pack, it is not possible to dehydrate slurry below the lowest perforation. This will create a void which may be subsequently filled when the pack settles with the risk of leaving voids elsewhere in the pack. Squeeze packs are generally not recommended and should in any case be restricted to gravel packing very short intervals i.e. less than 5 meters

2.2.6.2 Circulation-squeeze pack

Lower tell-tales are frequently used to force a self-induced squeeze pack. The crossover tool is left in the lower circulating position for the entire gravel placement operation.

The pressure drop for returns to flow through the telltale increases dramatically as the annulus packs-off and the circulating pack effectively becomes a squeeze pack, which is undesirable in most cases. This method, which does not require any operator intervention during the job, has however been adopted by many operators because of its operational convenience .

2.2.6.3 Circulation pack

The objective of a circulation pack is to fill the annulus and the perforations from the bottom up. The wash pipe must be spaced out so that return circulation occurs as close as possible to the bottom of the screen (Fig. 723). As the gravel pack begins, slurry fills the annular space outside the screen and the slurry starts to dehydrate at the bottom of the screen opposite the end of the washpipe. As packed gravel accumulates around the bottom of the screen, a higher pressure drop is required for the fluid to enter the screen and to flow in the screen/washpipe annulus to the end of the washpipe, thereby inducing fluid leak-off through the perforations. It is important to correctly size the washpipe OD to enhance this effect.

Fluid leak-off can also be controlled by action on the fluid returns. However, when high gravel concentrations are used, the pack can be completed in a few minutes and little time is available to effectively control the operations. More elaborated instrumentation than that currently available on the wellsite today is required to allow effective control of gravel placement i.e. real time measurement of operational variables like fluid returns. Ideally the gravel placement process should be completely automated.

2.2.6.4 Perforation prepacking

Conceptually, the ideal cased hole gravel packing method is to separate the perforation and annular packing process, i.e. to prepack the perforations. Prepacking methods are claimed by other operators to be very effective at packing perforations, controlling fluid loss and to result in high well productivity. Gravel can be squeezed into the perforations prior to running the gravel pack liner assembly. This method is to some extent a revival of the old "washdown" gravel packing technique. If it is necessary to control fluid losses after squeezing gravel behind the casing the perforations can be sealed with fluid loss control pills.

As previously discussed severe formation damage can occur during the installation of an IGP. This is thought to be due to combination of losses during completion operations and inadequate tunnel fill. It is well known that standard gravel packing operations result in a conflicting process of tunnel versus annular fill, and with such a scenario, preferential leak off can occur and in many instances this will result in the formation of bridges over certain perforations.

a. Prepacking

Prepacking - the concept of splitting the process of perforation packing from gravel packing the annulus is not new, and has been carried out routinely by a number of operators.

Flling the perforation tunnels immediately after perforating is thought to reduce the likelihood of tunnel damage by:

·Minimising the possibility of tunnel/cavity collapse.

·Invasion of incompatible (possibly unclean) fluids.

Furthermore the chance of filling the perforations is increased as the annular packing stage often results in additional packing of perforations.

Recent work has been in support of an extensive field test programme. Although reports to date indicate prepacking to be an operational success, no clear gains in productivity have been demonstrated.

Nevertheless prepacking followed by circulation packing is the currently recommended technique for difficult wells: long highly deviated zones, reservoirs which exhibit high permeability contrasts, and zones where excessive losses are anticipated.

b. Types of prepacking

The following is a summary of the basic prepacking techniques:

Prepacking without acidising - commonly referred to as the BP technique. This involves perforating the zone underbalance, pulling guns above the completion interval and packing the perforations with low concentration slurry. Perforation tunnels are then seal using a LCM system. Thereafter the annulus is conventionally packed. If required an additional LCM pill is spotted inside the wire wrapped screen prior to pulling the work string.

This techniques has been used extensively by BP in the North Sea where high permeability reservoirs dictate the use of particulate LCM.

Prepacking while acidising - as above this operation can be carried out before or after running the gravel pack liner. However in order to minimise losses and the risk of tunnel collapse it is preferred to pack perforations prior to pulling the perforating guns. Again LCM can be spotted where required.

Uniform treatment is difficult as acid will preferentially flow into the first set of open perforations or into high permeability zones. Possible diversion techniques include gelled or foamed acids. A relatively new but promising method is to use viscous pills loaded with gravel to divert acid stages and simultaneously prepack perforations

A typical gravel diverted acidisation procedure is as follows:

·The treatment should be carried out in stages with gelled prepack diverter slurry placed between stages. The diverter pill containing the gravel required to pack the perforations opened by the preceding stage. Each acid stage should be over flushed far enough into the formation to reduce the potential for damage due acidisation side effects.

·The final acid stage should be followed by the gravel pack slurry required to complete the pack.

A typical design practice is to target between 10-20ft of perforations, with gravel concentrations in the range of 1-2 lb/gal.

Chevron have reported a significant increase in productivity in both oil and gas wells completed with this method compared to conventional gravel packs.

Auger system - this system involves spotting gravel across the completion interval, and "screwing" a specially designed screen into the gravel bed. It is claimed that this technique promotes tighter more effective packing of perforations.

c. Prepacking fluids

Gravel laden water, HEC and Shellflo-S slurries where tested under deviated well conditions at various pump rates and sand concentrations.

·The use of high viscosity slurries (100 cP) at a shear rate of 100 sec-1) leads to filling of perforations from the top to the bottom of the interval, node formation and risk of premature bridging, leading to incomplete filling of perforation at the bottom of the interval. High viscosity slurries are not recommended.

·The use of medium viscosity slurries (40 cP at a shear rate of 100 sec-1) leads to simultaneous filling of perforations, along the entire interval, node formation, tightly packed perforations, but with a substantial risk of premature bridging due to a weak diverting effect. With this slurry a pumping rate of 2 bpm and gravel loading of 3 ppg seems ideal.

·The use of low viscosity slurries (20cP of a shear rate of 100 sec-1) leads to filling of perforation from bottom to top of the interval. Gravel is swept into the perforations from the top of the gravel bed that develops in the casing. Provided leak off rate is sufficient, tightly packed perforations can be achieved, also for deviated wells. Because the perforations are treated layer by layer, there is not much risk of premature bridging in the casing. This option is very attractive and should be tried out in the field.

Based on the above studies the use of water as a prepack carrier fluid is not recommended, mainly because very high leak off rates are required to fill perforations with gravel. However the study did confirm that water is ideal for gravel packing the annulus.

2.3 Slurry design for IGPs

2.3.1 Gravel loading and polymer concentration

A gravel concentration of 15 pounds per gallon of fluid is commonly used by many operators as a standard slurry recipe for IGPs. This probably represents the maximum gravel concentration that can be pumped under field conditions. In many cases however, the use of lower gravel concentrations together with lower polymer concentrations may offer clear advantages. For an IGP, the pro's and con's of low, medium or high gravel concentrations are as follows:

Low gravel and zero polymer concentration (0.2-2 lb/gal gravel)

·Dune formation in highly deviated and horizontal wells leading to dense annular pack.

·Large fluid volume required to pack perforations.

·More intermixing compared to viscosified slurries.

·Minimises bridging in the case of restricted screen-casing clearance.

·High risk of screen plugging in the case of circulation packing.

·Ineffective placement of gravel in upward facing perforations in high deviated wells.

Low gravel and low polymer concentration (1-2 lb/gal gravel)

(< 58 lb/1000 gal HEC, <0.2 w/w % active Shell Flo-S)

·Is essentially an enhanced water slurry to promote better transport of gravel into perforations.

Medium gravel and polymer concentrations (3-6 lb/gal gravel mix ratio)

(58 - 62 lb/1000 gal HEC, 0.2 - 0.4 w/w % active Shellflo-S)

·Generally helps to prevent bridging, especially when the casing/screen annulus clearance is small.

·Compared to high gravel and polymer concentration increases the volume of slurry to be dehydrated in the perforations. The importance of this aspect is a function of the impairment potential of the fluids used.

·Compared to high gravel and polymer concentration placement times are longer. Although rig time increases, more time is available to control the gravel placement operations, i.e. to ensure that the adequate balance between losses and returns is achieved during placement.

·Is beneficial for gravel annular placement in highly deviated wells.

·Should result in better placement over entire interval in the case of prepacking.

High gravel and polymer concentrations (> 7 lb/gal gravel mix ratio):

(> 62 lb/1000 gal HEC, > 0.4 w/w % active Shellflo-S)

·Causes significant increases in slurry viscosity and density. The resulting increase in hydrostatic head and friction pressure losses may prevent circulation of fluids to surface.

·Increases the risk of bridging especially in long intervals and when a highly heterogeneous formation is being packed. High fluid viscosity and high polymer concentrations promote layering on the screen surface and the build-up of nodes at perforation entrance. This may cause bridging especially across high permeability intervals.

·Minimises the fluid volumes to be squeezed and the time required for placement.

·May cause "U-tubing" of the wellbore fluids when pumping the slurry, with consequent loss of volumetric control.

·Less risk of plugging the screen in circulation pack.

·Diverts fluid over interval for medium permeability contrasts.

Slurries with high gravel and polymer concentrations have generally resulted in gravel packs that effectively control sand. However when applied indiscriminately, this recipe may contribute to poor gravel placement in the perforation tunnels and hence restricted well productivity. The inability to circulate or the poor operational control of gravel placement (it may take a few minutes only) is most likely the cause of poor gravel placement in many cases. Impairment may also be caused by the high polymer concentrations required to suspend the gravel.

High gravel and polymer concentrations

(i.e. > 12 lb/gal gravel mix ratio) should be avoided in the following cases:

·Non-uniform formation injectivity profile, especially if high permeabilities are encountered at the top of the interval.

·With low screen/casing clearance (one inch or less).

·Long intervals (> 50 feet) and/or highly deviated (> 60 degrees) wells.

·If no circulation can be achieved or poor control of placement operations is expected.

2.3.2 Recommended placement procedures

Prepacking methods, where the process of perforation tunnel fill is separated from the annulus packing operation offers the best scope for improving pack factors and hence increased productivity.

Circulation squeeze packs are commonly carried out using a lower tell-tale and leaving the crossover tool in the lower circulating position. The pack essentially becomes a squeeze pack after the lower tell-tale is covered with gravel. This method is generally not recommended for intervals longer than 15 feet.

Field experience and experimental investigations allow placement recommendations for specific conditions, for example:

1.Low angle wells (< 30°) and/or short intervals (< 15 feet)

The circulation pack method is recommended.

2.Intermediate angle wells (30°-60°) and/or long intervals and/or intervals with high permeability contrasts.

Prepacking followed by circulation packing is recommended in this case. Low or medium gravel and polymer concentrations are recommended for the prepacking phase, low or medium gravel/polymer concentrations are recommended for the annular pack.

The use of low density gravel is expected to further increase pack factors, but this has still to be tested within the Group.

3.High angle and horizontal wells (> 60°)

The circulation pack method is the preferred method in this case, where low gravel concentration with water or brine is applied. Flow experiments in Shell Oil have shown that a minimum fluid velocity in the screen-casing annulus must be sustained to achieve successful annular packs in such cases.

2.3.3 Gravel volume

The total gravel volume required is the sum of:

·The annular volume from sump packer to top screen,

·The gravel volume expected to be placed behind casing,

·The reserve gravel volume

Because of the uncertainty over the expected gravel volume behind the casing, a safety margin is recommended, e.g. 10 % of the total volume calculated. Using an excess should not itself cause problems since circulating out excess slurry can be usually be easily accomplished.

The amount of gravel that can be placed behind casing is a function of the perforation design and the perforation cleaning method used as well as the formation type. Larger cavities may exist behind the casing when perforations are washed instead of backsurged. In practice it is difficult to estimate the void volume behind casing. This can be done by measuring the volume of formation material recovered on surface, if any, during washing or backflowing operations and/or by measuring the level of fill within the casing before and after these operations.

The gravel volume effectively placed also depends on the perforation packing efficiency of the gravel placement procedures used. Hence the gravel volume to be placed behind casing should be derived from local field experience. When no field experience is available, use a minimum of 0.3 ft3 per foot perforated interval in the case of a new well.

In the case of wells that have produced significant sand quantities prior to gravel packing, (cumulative sand production data is rarely reliable) a gravel volume of 1 ft3 per foot perforated interval can be taken as a guideline. If the slurry volume proves too small, additional batches of slurry can be placed.

2.3.4 Example of gravel volume calculation

 

2.4 Gravel pack liner assembly

2.4.1 Screen considerations for IGPs

2.4.1.1 Screen size

The following considerations apply for the selection of the screen size:

·A minimum radial clearance of one inch (between the largest screen/pipe OD and the casing ID) is required to allow for washing over the screen.

·A larger annulus clearance reduces the chance of bridging especially in the case of difficult packing conditions i.e.: long intervals (> 50 feet) and/or non homogeneous injectivity profile and/or highly deviated zones (> 60 degrees). However too large clearances could affect the placement of gravel in the annulus if flow velocities become too low, especially in highly deviated wells. In practice a clearance of 1-2 inch is recommended.

·A screen ID larger than the tubing ID offers little advantage.

·For multiple completions, the screen internal diameter is governed by the size of the production tubing required to reach a lower zone.

2.4.1.2 Screen overlap

A minimum screen overlap over both top and bottom perforations of 5 feet should be used. A larger overlap provides some flexibility in the screen design as the length can be rounded of to a number of whole joints. The overlap also dictates a minimum gravel height in the annulus as screen out occurs only when the entire screen is covered. This may be important to ensure that the entire perforated interval is covered, especially in highly deviated wells.

Screen overlap is not a critical design consideration and a length of 10 to 20 feet above top perforations is often used.

2.4.2 Lower tell-tale (LTT)

A lower tell-tale is supposed to give a positive indication of when the slurry reaches the bottom of the annulus and to enhance pack quality (fewer voids. It is the subject of continuing industry debate as to its best use and effectiveness. In general however the use of LTTs cannot be recommended for the following reasons:

·The use of LTTs makes the screen assembly and gravel pack tool (slightly) more complex and expensive.

·Discussions with vendors have shown that the use of LTTs gives little proven benefits. There is ample speculation within the service industry that an LTT improves pack quality but the reason for enhancement is not entirely clear. In low deviation wells, an LTT is of little use as gravel settling promotes complete annular packing. In highly deviated or long intervals, gravel pack models have shown that the most efficient method of maintaining flow in the screen/casing annulus and to ensure good packing is to minimise the washpipe/screen annulus and to set the washpipe end at the bottom of the screen.

·The LTT is supposed to provide a displacement bench-mark for the gravel packing procedure. However in many cases there is hardly any surface pressure indication of when the slurry reaches the LTT. If U-tubing has occured when pumping the slurry, a sudden surface pressure increment may only occur when the displacement fluid catches up with the slurry, long after the slurry has screened out on the LTT.

·The gravel pack tool should be moved to the upper circulating position after screen out on the LTT to avoid packing under squeezing conditions. However in many cases this is largely ineffective as it is difficult to detect the right moment to do so.

·Operators often leave the gravel pack tool in the lower circulating position during the whole packing operation. Once compacted gravel covers the tell-tale the resistance to flow becomes higher and forces fluid leak-off into the formation. Gravel packs executed in this manner effectively become squeeze packs, which is undesirable in most cases.

In summary, lower tell-tales should be generally be avoided as they are costly, increase the mechanical complexity of the liner assembly and are of little proven help in ensuring good placement. The best method to ensure good gravel placement is to correctly size the washpipe and closely monitor the progress of the packing operations.

2.5 Procedures

2.5.1 Killing the well

The manner in which the well is killed after underbalanced perforating and backflowing will affect both the placement of gravel and the ultimate well productivity. The same level of precautions as taken during gravel packing operations should be applied when killing the well. There are two methods for killing the well after backflow operations: bullheading or reverse circulating.

Bullheading is unacceptable as formation and gun debris will be pumped back into the perforations, defeating the whole object of backsurging. This method is however operationally more convenient as hydrocarbons are kept from reaching surface.

Specialised downhole accessories (e.g. Halliburton OMNI valve) allow isolation of the formation downhole and enable the well to be reverse circulated with zero fluid loss to the formation. This avoids squeezing formation or gun debris back into the perforations. These downhole tools are typically used for Drill Stem Testing (DST) and more information on these tools can be found in the DST manual, EP 89-0490.

Reverse circulating through a DST valve above the TCP packer will remove hydrocarbons in the work string above this level. A significant volume of hydrocarbons may however remain trapped below the packer and will be released when unseating the packer. Bullheading all hydrocarbons below the DST valve should be considered as it allows:

·Positive prevention of brine contamination and associated problems (well control).

·Fluid loss agents can be placed immediately thus preventing any losses.

·The amount of debris pumped back into the perforations should be limited if the well is closed in for some time after backflowing to allow debris to settle.

2.5.2 Fluid loss control after perforating

Fluid loss control may become necessary after perforating operations as it is not always possible to avoid significant losses while maintaining the minimum operationally required overbalance (usually 100 - 200 psi). This is a function of formation properties, fluid overbalance at the formation face and perforation efficiency.

The maximum acceptable fluid loss rate above which fluid loss control systems are necessary must be established. If the completion brine is non damaging to the formation the main concern from a well productivity point of view is that fluid loss control systems are difficult to remove efficiently even by acidisation due to diversion problems. After an acid treatment, many perforations can effectively remain plugged, resulting in poor gravel placement and poor well productivity. Hence, if fluids are non-damaging (i.e. filtration is up to scratch), the limit is the maximum fluid loss rate operationally acceptable and is mainly a well safety issue. With proper planning and under certain conditions, some operators may allow up to 50 bbl per hour losses to the formation.

Conventional fluid loss control systems can be grouped into two categories: soluble particle systems and viscous fluid systems. A relatively new method of fluid loss control consists of prepacking the perforation tunnels with gravel. Field experience should assist in selecting the system which allows to control fluid loss rates and result in minimum residual impairment.

2.5.2.1 Soluble particle systems

Particulate fluid Loss Control Material (LCM) include acid, oil or water soluble material of selected grain size (for details refer to the Workover and Completion Fluids Manual. These systems can be very effective in controlling fluid loss by the build-up of a thin, impermeable layer of particulate matter on the formation face. Subsequent removal of this layer requires placement of a solubilising fluid i.e. under saturated brine for soluble salts or hydrochloric acid for calcium carbonate material. Removal efficiency is a function of solubility and contact time but fluid diversion is the main problem. Incomplete removal of LCM will prevent proper placement of gravel in perforation tunnels. Removing LCM after gravel packing may result in voids which may jeopardise the success of the gravel pack.

The use of calcium carbonate bridging agent is not recommended in the context of gravel packing as it requires the use of acid washes for removal. Formation collapse has been observed as a result of exposure to acid just prior to gravel packing. The use of oil soluble resins is not recommended either as they are difficult to dissolve and require hydrocarbon flow from rather than into the perforations.

A technique for controlling fluid loss is prepacking perforations with gravel after perforating. Typically the TCP string is used to squeeze gravel into the perforations after the backsurge operations. The packed perforations can further be sealed with a graded salt system prior to round tripping. The graded salt is removed by flushing under saturated brine prior to packing the annulus. This method is considered to offer much scope for obtaining unimpaired, gravel packed perforations. If perforations are not fully packed, graded salt will enter the perforation tunnels and will be extremely difficult to remove, which will lead to productivity loss. If there is any doubt that the perforation tunnels are poorly prepacked i.e. low pack factors, it is considered prudent to avoid the use of graded salt and opt for the use of a viscous fluid system. If the bottom hole temperature is appropriate Shell Flo-S is recommended.

2.5.2.2 Viscous fluid systems

Fluid loss can be controlled using high viscosity fluids, generally prepared with HEC or XC polymers. Rather than forming an impermeable filter cake on the formation face, the viscous fluid penetrates the formation. As the fluid flows out radially from the wellbore, the velocity and consequently the shear rate decreases. Given the shear-thinning properties of polymer solutions, a decrease in shear rate results in an increase in fluid viscosity. According to Darcy's law, for a given pressure differential an increase in viscosity decreases the rate. This effect can be calculated using a power law fluid model. Effective fluid loss control will depend on the volume of the pill. Typically, fluid loss is not eliminated but can be reduced to an acceptable rate.

Typical polymer concentrations for fluid loss control pills are the range of 100 to 120 lb. HEC per 1000 gal. of fluid. A classic pitfall in using viscous pills is improper gel preparation due to a shortage of time and lack of planning. High viscosity pills must be sheared and filtered in order to remove any particulate matter.

2.5.3 Prepack acidisation

2.5.3.1 When to acidise

Adequate leak-off is essential to obtain good gravel placement in the perforation tunnels. In case fluid loss control material has been used to kill the well, it is mandatory to remove it from the perforations, with a water, solvent or acid treatment. Acidisation must be considered if any potential for formation damage exists prior to gravel packing. Acidisation may follow an injection test performed prior to gravel packing to ensure that the perforated interval has enough injectivity.

There is no fixed guideline for what is an acceptable injectivity index. Operators should determine which criteria best suits their specific conditions. The following should be considered:

·Most gravel packs are placed at a pump rate of 2 to 4 bpm. As a rough rule of thumb, 50 to 70 % of the slurry needs to be diverted into the perforations during placement giving the injectivity range required.

·The bottom hole pressure when injecting or packing must be kept below the formation fracturing pressure.

The injectivity test is normally carried out with the crossover tool in the squeeze position, after setting the packer and checking the crossover tool positions. One or two rates are enough to establish the injectivity index. The work string must be cleaned prior to injecting any fluids into the formation.

2.5.3.2 Design guidelines

The design of a prepack acid job is critical as spent fluids cannot be back produced and a poorly designed job can do more harm than good. The following guidelines should be followed:

·If not already done, the work string must be pickled prior to the treatment to remove pipe dope, rust, scale.

·The acid types, concentrations and additives employed should be based on the formation mineralogy, type of damage expected, etc.

·Low strength acid should be preferred to reduce the possibility of precipitating acid by-products.

·Acid volumes should be kept to a minimum, 10-20 gal/ft is enough to target the perforation area and higher volumes e.g. 50-100 gal/ft perforation should be used if deeper damage is suspected.

·The treatment should be over flushed to minimise the potential for formation impairment due to precipitation of by-products. Typical practice: 150-200 gal/ft perforations.

·Pumping should be continuous throughout the acidisation and gravel packing operation to avoid the collapse of perforation tunnels weakened by acidisation.

2.5.4 Gravel packing

2.5.4.1 Pumping the slurry

As the progress of a gravel pack job is monitored by the volumes of fluids pumped, it is essential that the slurry arrives at the crossover tool in one homogeneous slug. Several authors recommend a velocity of 500 ft/min to prevent premature sand outs or roping of the slurry during the pumping operations [935], but the evidence supporting this criterion is weak. According to this criterion the minimum pump rate required in a 3 1/2inch work string is some 4.5 bbl/min. Further experimental and theoretical work is being progressed.

In many cases the slurry will free-fall as soon as it enters the work string. This is due to the difference in density between the slurry and the completion brine which is generally high especially with high gravel loadings. As the fluids in the wellbore are unbalanced the system "U-tubes" i.e. the fluid return rate is higher than the pumping rate; in other words voids are generated in the work string. If "U-tubing" occurs then the slurry may reach the cross-over port long before expected and volumetric control of the placement operations is lost. This will prevent proper gravel placement in the case of an IGP. One practical method to ensure that U-tubing does not occur is to keep the pump rate high enough to maintain positive pressure at the wellhead (the pump pressure may only indicate the friction pressure loss in the lines from the pump to the wellhead).

An alternative is to spot the slurry with the crossover tool in the reverse circulating position. If a low bottom hole pressure ball is used in the crossover tool then the formation is completely isolated from circulating pressures while spotting the slurry. U-tubing of the wellbore fluids can be prevented by choking the fluid returns. This method has one drawback as great care has to be taken to shift the crossover tool in the circulating position before slurry reaches the crossover port (use a healthy safety factor e.g. 5 barrels). Failure to do so may result in a stuck gravel pack string because of gravel in the annulus above the packer.

When the pad is within 5 barrels of the GP port, the pump rate can be slowed and the crossover tool lowered into the circulating position. The pump operator should react quickly to any unexpected sharp pressure increase. This may be caused by bridging of the gravel in the casing/screen annulus. The blank liner can easily be collapsed under excessive pressures and an expensive fishing job will result.

2.5.4.2 Placing the gravel (circulation pack)

The pump rate affects the annular and perforation gravel packing efficiency. In highly deviated wells, circulation rates should be kept high (2-3 bbl/min) to prevent premature bridging. In less deviated wells, circulating rates of 1/2to 2 bbl/min are adequate and allow more time to for the perforations to fill.

Circulation packs are in a sense a gradual squeeze. As the gravel pack begins, slurry fills the annular space outside the screen. As gravel accumulates at the bottom of the screen opposite the end of the washpipe, the pressure drop required for the fluid to enter the washpipe increases. Hence leak-off increases while fewer perforations remain unpacked.

Returns must be monitored during placement of the pack to ensure that adequate leak-off occurs. Measuring fluid lost after the job only serves the purpose of filling a number in a job report form. As gravel placement may be over very quickly, a real time measurement of the fluid returns (positive displacement flowmeter) is required to be able to adjust the balance between fluid returns and leak-off. There is no fixed guideline for the ratio of fluid leak-off to fluid returns. An acceptable range is 50 to 70 %.

2.5.4.3 Screen out

When screen-out occurs, pump pressures will increase sharply. The screen out pressure is supported by the gravel fill in the casing/screen annulus. This pressure is transmitted only a short distance within the pack and the formation is in principle not exposed to screen-out pressures. Hence formation breakdown pressure considerations should not be used as the criterion for screen out pressures.

Screen out pressures should not exceed 1000 to 1500 psi over the initial circulating pressure for the following reasons:

·It is unlikely that more gravel can be placed as the pack cannot be compacted in this fashion.

·Excessive pressures can damage downhole equipment. Incremental pressures must stay well within the blank liner collapse pressure.

Once screen-out occurs, pumping is stopped and the bleed-off rate should be observed. A slow bleed-off is a sign of a good pack as it indicates that carrier fluid leak-off is impeded by the presence of gravel in the annulus and the perforation tunnels.

A screen-out should be re-confirmed immediately, e.g. a few minutes after the initial screen-out. The reason being that if a premature bridge has formed, it often collapses as soon as pumping stops and packing can then be resumed without unnecessary time loss.

2.5.4.4 Reversing excess slurry

If the screen-out is confirmed, the annulus should be pressured-up to 500-1000 psi and the crossover tool picked-up to the reverse circulating position. Pressurising the annulus avoids having slurry entering the annulus as this may result in a stuck crossover tool. Connections should be spaced out to avoid stripping a connection through the Hydrill. To avoid bridging in the tubing do not wait too long before reversing the excess slurry. Circulate 1 1/2to 2 tubing volumes at high rates. It is best to have a check valve (low bottom hole pressure tool) in the crossover tool to isolate the pack from circulation pressures.

In order to estimate the volume of gravel placed outside the casing the volume of gravel returns should be measured. It is however difficult to measure accurately return volumes. The following procedure, developed by Baker, is recommended.

Returns are diverted to the blenders when gravel is spotted. With the blenders thoroughly mixing the return slurry, several samples of the fluid are taken. Using a centrifuge or a graduated cylinder, an average volume percentage of gravel is determined. The blenders can be shut down momentarily to determine the total volume of return slurry. Multiplying the total volume by the volume fraction of gravel reversed out gives a reasonably accurate value of the total volume of gravel reversed out of the well. The amount of gravel placed in the perforations can then be estimated.

2.5.5 Repacking

After the excess slurry has been circulated out, the pack should be left to settle. If a viscosified carrier fluid was used, then the pack should be left undisturbed for the time corresponding to the design viscosity break back time, usually an hour.

Screen-out should then be re-confirmed by an injectivity test. The "confirmed screen-out" criterion is then based on the circulation rate and/or pump rate achieved during the injectivity test. There are no firm guidelines to give here. Each operator should rely on field experience to establish threshold values.

If an additional batch of slurry is required, then a batch of no more than half of the original slurry volume should be mixed. Consideration should be given to reduce gravel concentration and viscosity to ease secondary placement operations. Slurry for re-packing can best be placed with the gravel pack tool in the reverse position.

2.5.6 Reporting

Concise and accurate job reporting is essential to allow for subsequent analysis and review of gravel packing operations. Suitable job report forms are provided in Section 22.

The Perforation Pack Factor (PPF) or amount of gravel placed behind casing is a qualitative indicator of the success of an IGP. This can be estimated if the volume of gravel reversed out of the well is known.

It should be noted that it is notoriously difficult to accurately measure the volume of the gravel returns, nevertheless it is recommended that as accurate measurements as possible are made to enable the quality of the technique and operation to be evaluated.

2.5.7 Gravel Pack Simulators

The lack of firm guidelines for gravel packing is partially due to the fact that large scale physical gravel pack simulators do not model the gravel pack process realistically (linear leak-off, no diversion effects, low perforation cavity volume). Furthermore, data gathered in the field are inaccurate and incomplete, for example the void space behind the casing prior to gravel packing is unknown, the uniformity of perforation filling along the interval is unknown, leading to evaluation difficulties.

An improvement in the understanding of the gravel pack process is expected to yield better job design and control. This can be realised by numerical simulation where process sensitivities can be tested and validated experimentally if necessary.

3 External gravel packing

Two types of External Gravel Packs (EGP) can be distinguished, the Open Hole Gravel Pack (OHGP) and Milled Casing Underreamed Gravel Pack (MCUGP). For both types, the completion interval is underreamed to enlarge the wellbore diameter and remove the zone of damaged permeability due to drilling operations. Drilling and casing milling operations are beyond the scope of this manual, readers are advised to consult local service companies for more information.

Most of the concepts applied in cased hole gravel packs are also valid for open hole gravel packs. Hence only procedures and equipment requirements that differ will be addressed in this section.

3.1 Underreaming

3.1.1 Effect of underreaming on well productivity

EGPs should theoretically provide a higher productivity than open hole completions or cased hole gravel packs because the restrictive casing perforations are eliminated and the underreamed borehole is filled with highly permeable gravel which improves radial flow into the well.

In theory, an 8 inch hole underreamed to 16 inch and then gravel packed should have 15 % higher productivity than an open hole completion. In practice, this effect is rarely seen because of the various impairing mechanisms that may be caused by the underreaming and gravel packing operations. For example shale streaks in the reservoir may release clay cuttings which can severely impair the formation.Nevertheless EGPs generally show a better productivity than comparable IGPs. This is due to the avoidance of restrictive perforation tunnels, making an EGP more forgiving with respect to impairment mechanisms.

3.1.2 Underreaming operations

Underreamers come in many different types and sizes. Expandable underreamers have arms that expand when hydraulic pressure is applied and retract in the tool body when the pumps are shut-down. A variety of cutters can be attached to the arms i.e. roller cones, PDC cutters or hardened cutter blades. Blade type underreamers are only suitable for soft unconsolidated formations. The maximum diameter to which a hole can be underreamed is a function of the tool used and the reservoir rock strength. Conventional tools allow to underream up to twice the tool diameter in medium hard to soft formations.

3.1.3 Underreaming fluids

Great care must be taken to prevent formation impairment when underreaming as any damage will subsequently be locked behind the gravel pack. Impairment can be caused by several mechanisms: by build-up of an impermeable filter cake with formation debris, by clay swelling due to fluid incompatibility or by reservoir invasion of particulate matter. Formation damage can be minimised by:

·Thoroughly cleaning the wellbore prior to underreaming.

·Minimising the overbalance on the reservoir.

·Using filtered, compatible fluids.

·Using efficient solids removal equipment.

·Using a properly sized acid or water degradable particulate system to control losses if required.

Viscosified clear brines are commonly used as underreaming fluids to minimise impairment. Viscosity is required to provide solids transport capacity and will help to control fluid loss. HEC or HEC/XC polymers are generally used as viscosifying agents. Shellflo-S should be seriously considered in view of its better carrying capacity and degradability.

Fluid cleanliness is extremely important for underreaming operations. Viscous underreaming fluids should be prepared with the same precautions as for a gravel pack carrier fluid i.e. polymer solutions should be sheared, filtered and mixed with a viscosity breaker, if required, prior to pumping downhole. When volumes are excessive, viscous fluids are often sheared only. An efficient solids removal equipment is mandatory. If price and brine availability allows, part of the circulating system should be frequently changed out in order to reduce the suspended solids content.

Losses are inevitable when underreaming with clear brines. In very permeable or depleted reservoirs, a degradable particulate system will be required to maintain hole stability and minimise formation damage. Bridging solids shape and size distribution should be such that a thin cake forms quickly [938]. Calcium carbonate or salt particle systems can be used. Subsequent removal of the bridging cake is however difficult and often incomplete. The use of graded salt particles is a recent development. The advantage is that the bridging cake can be dissolved by circulating under-saturated brine. Graded salt systems have recently been used to successfully drill horizontal wells in poorly consolidated sandstone reservoir.

3.2 Gravel placement considerations

3.2.1 Factors affecting gravel placement

EGP completions only require that the annular space between the screen and the underreamed hole be packed with gravel. Hence fluid leak-off into the formation is not essential to place the pack but contributes to compacting the pack by slurry dehydration.

Basically, the factors affecting gravel placement in the casing/screen annulus for an IGP also apply here. The main difference is that the annular clearance is much larger in the case of an EGP.

Especially with thin fluids, there is an inherent risk of mixing gravel with sand as the slurry is being circulated downhole. Hence circulation rates are generally limited to about 4 barrels per minute in normal applications, where fluid velocity in the annulus is not a critical factor for gravel placement. In highly deviated intervals fluid velocity in the annulus should be designed with a view to optimise gravel transport in the annulus.

As with an IGP, gravel placement is increasingly difficult with increasing deviation, interval length and fluid leak-off variations along the interval.

3.2.2 Gravel placement methods

The one trip crossover circulation method is normally used to place an EGP. The cheaper option is to use the "over the top" system together with low viscosity fluids (conventional packing).

3.3 Slurry design for EGP's

3.3.1 Gravel loading and polymer concentration

Slurry design is generally less critical for an EGP than for an IGP because there are no perforations to pack.

For normal applications (deviation < 60 degrees), both low viscosity and high viscosity carrier fluids have been used with equal success. The fluid carrying capacity is not a critical factor as gravity helps to pack the annulus and packing occurs from the bottom of the well upwards. In the context of open hole gravel packing, merits of slurry and conventional packing are as follows:

Conventional packing (non-viscosified brines, 0.2 to 2 lb/gal gravel concentration):

·Cheaper with respect to fluid costs, may however require long pumping times and hence generally more costly overall.

·More likely to de-stabilise the formation sand and cause impairment by mixing of gravel and sand.

·Significant fluid losses may occur.

·Prolonged exposure of the formation to completion fluid.

Slurry packing (up to 80 lb/Mgal HEC, 15 lb/gal gravel concentration):

·Short pumping times.

·Less likely to cause gravel/sand mixing because of the lower pump rates, higher viscosity and reduced exposure times.

·More expensive than conventional packing

3.3.2 Gravel volume

The gravel volume should be calculated using a similar approach as for cased hole gravel packing. A calliper log should be considered prior to running the screen to insure the hole is open and to provide an indication of the quantity of gravel to be placed. 20 to 25 % excess should be added to the calculated amount of gravel. Field experience may indicate that a different value is more appropriate.

3.4 Gravel pack liner assembly

3.4.1 Screen dimensions

It is recommended that the same screen sizes be used in open hole as for cased hole completions.

3.4.2 Tell-tale

As for an IGP, the use of a Lower Tell Tale (LTT) when slurry packing is not recommended. The benefits of a LTT are even more doubtful here as the screen/openhole annulus is many times larger than the screen/washpipe annulus. However the washpipe should be properly sized.

Upper Tell-Tales (UTT) should be used only when conventional packing (packing with non viscosified brines) is carried out. 

3.4.3 Blank liner

Part of the gravel reserve should be in the open hole portion of the completion. More reserve gravel will be available because of the relatively larger annular volume in the screen/open hole annulus. Hence the blank pipe should extend 10 feet below the casing shoe. Another reason is that a void can easily develop just below the casing shoe if the screen extends into the casing and if there is a small screen/casing clearance. A bridge may form as the gravel in the open hole settles and a void is created below the casing shoe.

3.5 Procedures

3.5.1 Acidising

Prepack acidisation must be carried out if acid degradable loss control material has been used when underreaming.

3.5.2 Gravel placement

4 Gravel packing in special applications

4.1 Completing horizontal wells through unconsolidated sands

Most  common applications:

1.To promote cone suppression.

2.To laterally connect multiple reservoirs and/or high permeability features.

3.To increase productivity through increased exposure to the formation.

The vast majority of conventional vertical/deviated wells employ gravel packs when sand control is necessary. However, in horizontal wells the technical difficulties and additional cost make this traditional option difficult to justify in the majority of cases. Clearly, completion costs become disproportionate with long lateral drain holes, particularly when simple completions are not feasible. 

4.1.1 Wellbore stability

There is a common mis-conception that openhole horizontal wells are inherently more stable than their vertical counterparts. From a rock mechanical point of view this is not the case for a normally tectonically stressed area, as the horizontal wellbore will be perpendicular to the principal (vertical) stresses. In fact by extending the above further, it can be clearly shown that from a theoretical point of view a cased horizontal hole, perforated in the vertical plane only is the most stable horizontal completion configuration. However with such a configuration massive failure will occur more suddenly as post failure stabilisation is not possible.

This can be partially explained by considering the following:

·The drawdown imposed on the formation is comparatively less due to the increased inflow area. This is thought to greatly offset the effect of hole angle and to a lesser extent orientation.

·Fluid velocity, hence drag forces are much lower in horizontal wells. This together with reduced drawdown pressure may in some cases help substantiate less stringent sand control requirements in horizontal wells.

·Changes in wellbore fluid composition frequently lead to increased sand production. As mentioned above, horizontal wells may help defer water (gas) breakthrough and hence indirectly promote sand exclusion.

·In cased hole completions, although the hole is horizontal perforations might be shot in the vertical plane.

·In uncased horizontal holes the formation may collapse on the liner or screen, creating a new stable situation without excessive damage and with little sand production.

Regardless of the above, most of the horizontal wells drilled to date have been justified on the basis of optimising production, reservoir sweep, etc. With time and experience, sand exclusion requirements may prove to be less stringent - offering potential to reduce completion costs compared to vertical gravel packs. Clearly horizontal wells offer a way of offsetting problems associated with conventional wells and traditional sand control practices.

In summary, open hole horizontal wells through normally stressed reservoirs are more likely to be unstable compared to vertical wells drilled through similar formations. Nevertheless, Group experience to date indicates that neither wellbore stability nor sand control are insurmountable problems when planning horizontal wells through poorly consolidated reservoirs. However, long term production history from existing horizontal developments may prove otherwise.

4.1.2 Completion types

Unconsolidated formations are generally completed with a variety of screen designs: including dual screen prepacks, perforated outside prepacks, slotted liners and pre-drilled liners or appropriate combinations. Although attempts have been made to gravel pack highly deviated wells, to date no attempt has been made within the Group to gravel pack a truly (90°) horizontal well.

Within the industry the most common completion type for horizontal wells through unconsolidated sands is the installation of a liner or screen. This section describes the pro's and con's of the most widely used completion techniques.

4.1.2.1 Liners and Screens

In cases where large grain, well sorted reservoirs are being developed, screens used without gravel packing have proven effective in controlling sand production without a significant drop in short to medium term production. Furthermore, in many low rate oil well applications, slotted liners or conventional wire wrapped screens in most cases turn out to be the most "cost effective" options. For higher rate oil (and gas) wells, or in areas where the cost of remedial action is prohibitive (eg. subsea wells) prepacked screens are normally recommended to provide added insurance against screen erosion. In short or medium radius wells where aggressive build rates are required, and/or in applications through poorer quality, finer sands the use of pre-packed screens should be considered with some caution due to concerns about mechanical integrity, quality control, and plugging potential.

a. Pre-drilled liners

Completions using pre-drilled uncemented liners offer nothing more than a future logging conduit. They prevent a unconsolidated formation completely filling the wellbore. Pre-drilled liners although commonly used by numerous Group companies are not viewed as sand control.

b. Slotted liners

Presently the least expensive form of sand control. Slots are cut parallel to the longitudinal axis and uniformly distributed around the circumference of the liner whilst retaining the mechanical integrity of the liner.

The disadvantages commonly quoted are:

1.Smaller effective slot inlet area.

2.Slots tend to plug easily with formation fines, corrosion products, and precipitates during installation and during production.

3.In most applications the required slot sizes, below say 300mm, are difficult to achieve using conventional machine tools, hence exclusion of formation sand cannot be guaranteed.

4.Slots may erode through fines production.

Slot plugging is not, apparently, adversely affecting short to medium term production performance of such horizontal well completions. However, in view of the long completion lengths it is possible that plugging may not be noticeable, or indeed, stable bridges may be formed across the slots. Nevertheless, to reduce plugging potential the inflow area of the liner should ideally be maximised without reducing the mechanical integrity of the liner. This however has a direct impact on machining costs, twice as many slots cost nearly twice as much. Industry guidelines recommend that a minimum of 2% inflow area should be provided.

c. Wire-wrapped screens (WWS)

WWS have around 20 times the inflow area compared to an equivalent slotted liner. Additionally, manufacturing methods are available which are capable of mass producing WWS with 50mm slots. The main drawback with WWS for such applications is that they are more costly, which in view of long completion intervals results in significant up front expenditure. As with slotted liners the use of WWS only should be limited to low erosion risk applications, which is frequently the case with horizontal wells.

d. Pre-packed screens (PPS)

Success stories on the use of PPS alone as a primary form of sand control in horizontal wells have been reported. Judging the performance of such sand control measures is always difficult, there is no doubt however that PPS have been successfully used in unconsolidated reservoirs containing clean, well sorted, large grain sands and gravels.

As with other forms of screen, the application of PPS in poorer quality reservoirs must however be viewed with caution as formation slump, or natural packing will lead to an area of low permeability around the screen. This reduction in permeability will be a function of grain size, sorting and clay content. As with any completion design which uses screens alone, this effect can be reduced by minimising the clearance between the screen and wellbore.

By design PPS are less likely to fail because of erosion. However, plugging of the gravel sheath during installation due to dirty completion fluids or by formation fines during production is a major problem.

Another major design consideration is the high risk of cracking the brittle resin consolidation when running screens through tight dog-legs, or while flexing during handling and make up. If the resin is cracked while installing the screens (or before) this can lead to an erosive type failure at a later date, with potentially serious ramifications, especially in higher rate gas wells.

Design limitations for various screen types marketed by the major service companies are readily available. Several of the major screen companies have subjected standard (both WWS and PPS and slotted pipe) screen designs to tests that involve pulling screens through casing strings shaped to represent a medium - short radius horizontal wells. Readers are advised to consult local screen vendors for more information.

The additional expense (3x WWS), availability and quality control problems are the other main disadvantages of pre-packed screens.

e. Gravel packing

As previously mentioned most horizontal wells within the industry where sand production problems were anticipated have been completed with liners or screens in open hole. Although the problem of formation damage was generally thought to be less important in horizontal wells, work conducted has concluded that formation damage is just as important in horizontal wells. A number of operators and authors advocate gravel packing prepacked screens to minimise plugging potential, and therefore maintain economic production rates through extended well life. Although this philosophy may lead to the technically optimal solution, it may, in many cases be less attractive due to the incremental cost of gravel packing.

Over the last decade great research effort has focussed on defining key parameters involved in the process of gravel packing highly deviated or horizontal wells. Numerous organisations and researchers have constructed full size physical models to observe the process.

In near vertical wells, gravity assists in overcoming placement failures in gravel packed wells due to settling. In highly deviated wells ( > 60° deviation) the opposite is true, gravity tends to exacerbate placement difficulties due to duning. In the above models using brine researchers have observed two waves of gravel flow: firstly the alpha wave - which deposits gravel on the lower side of the hole, and secondly the beta wave which propagates back towards the ports filling the gap between the alpha and top of the top side of the hole. Although a detailed description of the process is outside the scope of this review, the technology related to gravel packing high angle wells can essentially be directly applied to horizontal wells.

If there is a clear need to gravel pack a horizontal well and the project can support the incremental cost then a number of key design parameters have to be considered:

1.Sand transport efficiency - the ability of the carrier fluid to suspend gravel:

-Carrier fluid rheology.

-Gravel loading etc.

2.Pump rate:

-Fluid flowrate (from 2-4 bbl/min).

3.Tool and screen configuration:

-Washpipe OD screen ID ratio > 0.8.

-Centralization and clearance (0.75-1.0").

-Elimination of blank sections.

-Continuous washpipe to bottom.

4.2 Water injection wells

The requirement for sand control in a water injection well is debatable and no unique technique can be recommended. Sand control may be required in view of the following:

·Injection water tends to slowly reduce the the formation strength by dissolution of the cementing material and this is aggravated if the well is acidised.

·Backflow may occur when injection stops and crossflow between layers that have been differentially pressurised may occur. Significant quantities of sand may then accumulate in the wellbore resulting in impairment.

·Backflow flow may be required to clean out impairing material (gaslift often being provided for this purpose.

Laboratory tests have shown that significant amounts of silica can be dissolved from gravel pack and formation sand when large quantities of water or steam are injected into a sandstone reservoir. The rate of dissolution is dependent on the flow rate of water through the gravel pack and the formation, the pH of the injected fluid and the temperature

Consolidated gravel packs are used where conventional gravel packs tend to give only short lived protection against sand inflow because of dissolution of the gravel and formation sand by the injection fluid and fluidisation of the pack when backflushing the wells. Consolidated gravel is gravel pack sand which is coated with a resin consolidation system. The more widely used system is gravel pre-coated with a partially cured thermo-setting resin.

5 Post gravel packing and remedial operations

5.1 Bean-up procedures

The objective of a bean-up policy is to is to reach well potential within the shortest period of time without jeopardising ultimate well integrity and productivity.

During sand control workshops, bean-up procedures were found to vary greatly and were not always clearly justified. Some Opcos have however established that rapid bean-up times are detrimental to the long term performance of their wells.

5.2 Gravel pack logging

5.2.1 Objectives

The objective of gravel pack logging is to inspect the quality of the pack, either immediately after it has been set, or after some production has taken place. Gravel pack logging is not required on a routine basis after gravel packing. Recent experience is that most gravel packs are successful from a sand exclusion point of view. Gravel pack logging may be needed in the following circumstances:

·When premature screen-out is observed or when no screen out is possible.

·When difficult or non-routine gravel packs are carried out, e.g. long (> 50 ft) or highly deviated (> 60 degrees) hole sections.

·To detect pack slumping with time, e.g. slumping caused by gravel dissolution in steam injection wells.

If an incomplete pack is indicated, secondary packing operations can be employed.The logging technique should thus be capable of determining the top of the gravel pack and the presence of voids in the pack. By comparing logs taken at different moments in time, pack slumping with time and the development of cavities in or beyond the pack can be monitored.

Owing to the presence of metal screens and casing, only nuclear techniques are suitable to fulfil all these requirements. Two different types of logging devices are suitable, the gamma-gamma (photon) and dual spaced neutron (compensated neutron) tools. Both provide qualitative indications of changes in the completion interval i.e. the sum of effects caused by wellbore fluid, screen or liner assembly, gravel, cement and formation lithology to some extent. Because fewer variables affect its performance and interpretation, the photon tool is preferred for gravel pack evaluation although the FDC/CNL combination has in many cases also resulted in satisfactory logs.

5.2.2 Gamma-gamma tools

5.2.2.1 Principle

Like the usual density tools, gamma-gamma tools emit gamma rays into the formation around the tool and detect the intensity of backscattered radiation. The detector count rate decreases with increasing average bulk density (more absorption) in the volume investigated. This volume is roughly given by the source-to-detector spacing and extends radially up to a few inches in the formation beyond the gravel pack. The tool response is thus a function of the amount of gravel, but also of the amount of steel, fluid, cement and formation density. The relative importance of these components, however, depends on many parameters, but in particular on the source-to-detector spacing. This feature is explicitly used in dual detector tools.

5.2.2.2 Available tools

a. Dual detector tools

The Gravel Pack Porosity Tool GPPT (Schlumberger) and the Dual Photon log (Western Atlas) are specially designed production logging tools with a diameter of 4.3 cm (111/16"). Both tools have two detectors, set at different spacings from the source. Since the depth of investigation increases with increasing spacing, the two detectors have a different sensitivity to the presence of gravel in the annular space. This configuration allows a better qualitative assessment of the gravel sheath.

b. Single detector tools

There is a variety of tools with a single detector, which are either modifications of tools made for a different purpose, such as the Nuclear Fluid Density Meter tool NFD (Schlumberger), or dedicated such as the Photon log (Atlas Wireline). Because these tools give less information, they are more difficult to interpret, unless a baselog is available as a Reference.

5.2.2.3 Interpretation

The standard interpretation technique for dual detector tools is to plot the count rates of the near and far detectors on linear scales. The scales are chosen such that the logs overlay in good quality gravel (usually at the bottom of the pack). At places where the pack is poor, the count rate at both detectors will increase, but at different rates, and a separation is then seen.

The log readings at full fill and no fill may be used as calibration points, provided casing and cement thickness are the same. Log readings in between these values may be linearly interpolated in terms of percentage of fill. The resulting quantitative evaluation is of little practical value due to the broad range of variables affecting the measurement. The minimum detectable change under favourable conditions is some 10%, which corresponds to one to two litres (0.04 - 0.08 cf) of gravel in a 7" casing, 4" screen IGP.

Single detector logs can readily be interpreted when used in time-lapse mode, e.g. before and after a re-pack operation. The interpretation of a stand-alone run is much helped by the availability of a schematic drawing of the hardware (because density tools are very sensitive to hardware effects) and the open hole density log to recognise features that are not related to gravel pack quality.

5.2.3 Compensated neutron tools

5.2.3.1 Principle

Neutron tools emit high energy neutrons and detect those that have slowed down to thermal energy. Since the slowing down process is dominated by hydrogen, the detector output is a measure of the amount of hydrogen in the vicinity of the tool. In the application for gravel pack logging, the count rate increases with decreasing average void ("porosity") in the volume investigated, provided that the hole is liquid filled. The investigated volume is roughly given by the source-to-detector spacing and may extend up to several inches into the formation beyond the gravel pack. In gas-filled holes, the hydrogen content is too low to obtain an interpretable log.

5.2.3.2 Available tools

All current (open hole) neutron tools may be used. They have a dual detector system and are omni-directional. Tools are available in different diameters, down to about 7 cm. (2 3/4inch) (e.g. CNT, Schlumberger and NEUT, Western Atlas).

5.2.3.3 Interpretation

The standard interpretation technique is to plot the count rates of the near and far detector on linear scales. The scales are chosen such that the curves overlay in places with a good gravel pack. In intervals with a poor pack the count rates will decrease, but, because of the different depths of investigation, by a different factor and a separation will be seen. Variations in the formation properties are often reflected on the count rates, in particular when the gravel is homogeneously packed. The open hole neutron log (count rates) may help to explain such variations.

5.2.4 Other nuclear techniques

Pulsed neutron capture (PNC) logs, such as the TDT or PDK, may be used, but there is little experience of using them within the group. These tools emit neutrons, but detect gamma radiation. Consequently, the near and far count rates react to the presence of gravel in a similar way as neutron logs, but some effect of the hardware may be seen.

Recently, a method has been proposed measuring silicon activation with pulsed neutron logging. This method is valuable when very heavy completion fluids are used such that the density contrast between completion fluid and gravel is too low to be detectable with gamma-gamma tools. It would appear, however, that in such a case PNC or neutron logs would also be successful.

5.2.6 Recommendations

1.Dual detector gamma-gamma tools should be run in preference to neutron tools and may be interpretated semi-quantitatively. Single detector tools may be run when the completion scheme is simple, or in time lapse mode (e.g. before and after a repack job).

2.Near and far count rates should be overlayed over intervals with a good gravel pack.

3.Poor gravel pack will then show as a separation of the curves while both count rates change; a decrease in the case of neutron logs and an increase in the case of gamma-gamma logs.

4.A CCL should be run with the gravel pack log for proper depth match.

5.Changes in hardware (in particular collars) are clearly visible on the gamma-gamma logs.

6.Neutron logs may reflect changes in formation properties; this can be recognised from correlation with open hole neutron logs.

7.Neutron logs are not interpretable in gas-filled wells.

8.Completion scheme should be drawn on the plot.

9.Open hole logs (far detector count rate) should also be displayed.

5.2.7 Developments in gravel pack logging

Recent gravel pack development work has focused on optimising the process of gravel placement into perforation tunnels. This work has led to the introduction of alternative carrier fluids and better defined operational procedures, with a number of ongoing field trials. Evaluation of these trials and the effectiveness of routine gravel packs requires a quantitative assessment of perforation tunnel fill. With conventional logging equipment this is not possible.

Halliburton Logging Services (HLS) have however recently introduced two new tools:

1.TracerScan and

2.RotaScan, which have been developed to assist in assessing gravel placement.

These tools detect radioactive (gamma ray emitting) materials placed in the wellbore and near wellbore region. By coating gravel uniformly with a radioactive label these tools can be applied for evaluation of gravel inside and outside casing. Tagging with different isotopes facilitates the selective detection of carrier fluid, prepacked gravel and gravel placed in the screen-casing annulus. In collaboration with HLS, KSEPL at the beginning of 1992 initiated a study to calibrate tool response of both tools with screens placed around the tool, and for various (gravel filled) perforation cavity sizes and geometry. The remainder of this sections discusses the capabilities and limitations of these tools.

TracerScan is a spectroscopic gamma ray tool that can be used for the detection of radioactively labelled gravel placed in the annulus and/or perforation tunnels. The tool has a vertical resolution of 30 cm and a depth of investigation of approximately 20 cm. The tool facilitates 360° detection. The measured signal gives information on both the depth and density of placed gravel.

The TracerScan tool can measure:

1.Perforation tunnel fill along the perforated interval. However, it must be noted that only a qualitative assessment can be made.

2.Filling of the annulus. Voids in the annular pack, as well as top and bottom of the annular fill can be detected.

RotaScan is a total energy, directional gamma ray tool. The main component is a sodium iodide scintillation detector within a rotating tungsten shield containing a slotted apperture. The tool detects a circumferential section of 40°. A three axis accelerometer is used to determine the inclination and azimuthal orientation of the tool. Vertical resolution is 30cm and depth of investigation is approximately 50cm behind casing.

The tool shows potential for evaluating gravel placement in highly deviated wells. For instance placement in high and low side perforation tunnels may be compared.

5.2.7.1 Operational considerations

·When using the above mentioned tools it is recommended to tag the pre-pack gravel and annular pack with different isotopes.

·Half-life of the tracer should be long enough to ensure that, if the operation is delayed, no extra quantities of tagged material are required. Tracers having half-lives of 90 days are often used. Tagged gravel is normally prepared and transported to location by a specialised company.

5.3 Production logging in gravel packs

The presence of a gravel pack will necessarily affect the resolution of production logging tools. For this reason they are likely to have a limited usefulness in short intervals, except as a flow/no flow indicator.

For intervals longer than 10 ft or in reservoirs which are clearly stratified (e.g. large shale breaks) it is quite possible to obtain information on production contribution from different intervals. This is of interest for reservoir management purposes or as feedback to aid in the design of future gravel pack procedures. It may also provide information which is required for calculating gravel pack skins (length of zone contributing to production).

5.4 Gravel pack stimulation

This section addresses the design of stimulation treatments for impaired gravel packed wells. A diagnostic of well impairment should first be established and the stimulation treatment must be tailored to the type of damage identified. It is important to realise that some types of damage are difficult or even impossible to remove and impairment prevention is always more effective than damage removal. Some of the concepts presented here need further research and development to arrive at optimal damage removal.

5.4.1 Impairment diagnostic

The first step is to carry out a production test and BHP survey to measure the well impairment. This data is required to establish the severity of the problem and, by comparison with post-treatment data, to quantify the benefits of the stimulation treatment.

There are many potential sources of impairment in a gravel packed completion e.g. screen plugging, gravel contamination with formation sand or debris, plugged perforations, etc. Skin data, which is generally a more reliable indicator of well impairment than PI data, does not allow however to properly identify the impairment mechanisms apart from geometry and turbulence effects. It is difficult to discriminate between true formation damage and gravel pack impairment as there are simply too many variables in the problem.

The following approaches may help in identifying the source of impairment:

·A systematic well history (starting at the drilling phase!) and gravel pack job review should be established. Production engineers may often have only a portion of the well history available and may not be aware of important events like severe losses when drilling the reservoir. Hence some detective work is usually required.

·The computer program STIMSEL under the ICEPE portfolio is designed to help in pinpointing the extent and nature of impairment. The use of this tool requires caution however as it uses theoretical correlations to describe the productivity of gravel packed completions. This may result in overestimating the true formation damage as the effects of e.g. gravel/sand mixing or perforation plugging are included in the skin due to damage.

·When a large database has to be analysed, the use of statistical methods such as discriminant analysis should be considered.

·Production logging can also help in the diagnosis of well impairment by establishing the production profile in a well. KSEPL is currently developing software to aid in the interpretation of such profile with respect to impairment.

5.4.2 Design of stimulation fluid

The type of acids, the concentrations and additives employed should be based on the mineralogy of the formation, the type of damage presumed and specific wellbore conditions (e.g. BHT).

As the exact source of damage is rarely known, both laboratory and pilot tests should be carried out to identify the optimal stimulation treatment. Similarly, stimulation job reviews are an essential part of the optimisation process.

Ideally one should aim at developing stimulation treatments that can simultaneously remove different types of damage. Combinations of different treatments can be made e.g. two different damage sources could be targeted by pumping larger stimulation fluid volumes or by applying a longer soaking time. However tackling more than one problem at a time may cause undesirable side effects and may require a complicated treatment which becomes difficult to handle.

5.4.2.1 Screen cleaning

At the installation phase, screens may become plugged with mud, paint from downhole tools, pipe dope, and any material collected from the casing wall. When the well is producing, screens may become plugged with sand, wax or scale deposits.

Unless the plugging material is of organic nature (e.g. wax) the obvious choice is HCl acid or mud acid. Mud acid is normally preceded by an HCl preflush, as a general measure to minimise secondary precipitates.

If organic deposits are suspected, solvents such as alcohols blended with surfactants, mutual solvents or plain hydrocarbons like kerosene or xylene can be applied. Before using such chemicals, laboratory tests to select the the optimum solvent are essential.

The preferred placement method is to spot the acid with coiled tubing at the bottom of the well and subsequently pull the coiled tubing while pumping. This will ensure coverage of the entire screen if the fluids remain in the wellbore.

Mechanical methods such as jetting or washing are not recommended as they can potentially fluidise the pack or inject foreign matter into the gravel pack and cause further impairment.

5.4.2.2 Gravel cleaning

Gravel contamination with foreign material leads to impairment of the pack. The contaminating material can have many different origins:

·Pipe dope, paint debris,etc...

·Formation debris e.g. shale, feldspars...

·Precipitates due to incompatibility between formation and completion fluids.

·Mud remnants.

·Polymer residues.

Acid solubility is an important gravel quality specification item. The lower the quartz content, the higher the acid solubility as impurities (e.g. clay and feldspar) are more soluble in HCL and HF acid than quartz.

Incompatible fluids can cause impairment by precipitate deposition at the formation interface. The effect will be intensified if the fluids contain solids or if the viscosifier leaves a residue behind e.g. when poorly mixed, unsheared and unfiltered, partially broken HEC was used. If the impairment is due to precipitates then 15% HCl is the suggested treatment.

If however the problem is related to HEC residue, lower acid concentrations can be used. Laboratory and field work suggested that HEC breakdown by strong acids might create more insoluble residue, specifically at higher temperatures. However, when properly mixed, breakdown of HEC with acid presents no additional impairment problem.

Alternatively, treatments with enzymes (if the reservoir temperature is 65°C or lower) or hypochlorite may be applied. Both types of treatment have shown positive effects.

The volume of the treatment should be restricted to the volume of gravel to be treated, e.g. 1.1 times the volume of the pack. A diverting agent should be used to promote full coverage of the pack.

5.4.2.3 Near wellbore formation

The near wellbore formation may have been impaired during drilling, cementing, perforating, underreaming or gravel packing operations. As such gravel packed wells do not differ from perforated wells except for the access to the formation face. It is even more difficult to direct the stimulation fluids to the impaired region in gravel packed wells than it is in conventionally completed wells.

5.4.3 Diversion

It is difficult to achieve effective diversion in gravel packed wells, because there are many routes for the acid to by-pass the targeted zone.

Mechanical techniques such as selective placement tools, or inflatable packers can be used to isolate a particular zone for injection or stimulation of fluids. Ball sealers are generally not suitable for gravel packed wells due to the configuration of the screen.

Graded solid particles (e.g. benzoic acid or oil soluble resins) can be used to create a thin, low permeability cake to plug off high injectivity zones.

Examples of diverters are:

·MMOWG (Halliburton), a solution of ammonium benzoate from which finely divided benzoic acid will precipitate when mixed with acid.

·J363 (DS), a sodium benzoate that with proper surfactants will rapidly recrystallise to fine grained benzoic acid.

·DIV III (Nowsco) similar to Halliburton's MMOWG.

·Oil soluble resins.

In practice it may take months to fully dissolve the diverters in crude oil. Attempts to speed up the dissolution by applying solvents, such as methanol (benzoic acid) are usually not successful or may even have a negative effect, e.g. a diesel overflush to dissolve oil soluble resin will lead to the formation of sticky, impermeable residues. A solvent preflush preceding the stimulation treatment may be considered in such cases.

Diversion is also possible with viscous pills e.g. polymer solutions injected ahead of the stimulation acids. The diversion mechanism relies on the shear thinning properties of polymer solutions. The effectiveness of this method will be a function of the volume of polymer placed. One disadvantage is that viscous pills, may themselves cause impairment.

A typical foamed acid stimulation pumping schedule may read as follows:

1.Prepare the well and well site for the treatment by rigging up the coiled tubing unit and pumping equipment. Pressure test the surface lines the blow out preventer, and tubulars to the required pressure.

2.Run the coiled tubing in the hole and clean with the hole with foam. Spot 0.12 m3/m (10 gal/ft) of foamed 10% w/w HCl, while pulling the coiled tubing up at a velocity of 1 m/min. Maintain pumprate at 159 l/min (1 bbl/min).

3.Run the coiled tubing back to original depth. Spot 0.25 m3/m (20 gal/ft) of foamed mud acid. Pull the coiled tubing at 0.5 m/min while pumping at 159 l/min (1 bbl/min).

4.Inject 0.12 m3/m (10 gal/ft)of unfoamed 3% NH4Cl brine, while lowering the coiled tubing at 1 m/min and pumping at 159 l/min (1 bbl/min). An additional similar volume of NH3Cl brine should be injected after running in the coiled tubing to total depth.

5.Displace the tubing/coiled tubing annulus and while pulling the coiled tubing, the tubing with filtered water, suitable brine, oil, etc. Keep the well full with displacement fluid to avoid swabbing acid foam into the well bore while pulling.

Shell Oil also reported good results with viscosifying and/or foaming separate diverter stages in between acid stages, i.e. using a combination of foam and polymer (HEC). The exact mechanism of such diverted treatments is not fully understood.

5.4.4 Additives

Additives are required to combat the side effects inherent to the use of acid, i.e.:

·Corrosion of completion items.

·Precipitation of secondary reaction products, e.g. ferric hydroxide.

·Formation desintegration and fines release.

·Formation of emulsions.

·Clean-up problems.

This topic is discussed in the Production handbook and the Well Stimulation Manual [970]. As a general rule, the recipe should include a corrosion inhibitor and a sequestering agent to prevent iron precipitation.

5.4.4.1 Corrosion inhibition

In the context of gravel packing, special attention should be given to corrosion aspects. Screens are often manufactured from high alloy steels and the effect of acid on these alloys can be devastating, especially in view of the close tolerances imposed on screen slot width. Current corrosion inhibitors are chiefly designed to protect carbon steels, although most contractors claim that they can effectively protect high alloy steels with higher inhibitor concentrations. A number of service companies were requested to recommend corrosion inhibitors for these high alloy steels.

5.4.4.2 Sequestering

In case normal HCl is used, it is recommended to add a minimum of 10 kg/m3 citric acid (sequestering agent) to avoid precipitation of ferric hydroxide. If mud acid is used, the sequestering agent can usually be left out. More specific recommendations can be obtained by using SEQUES, a program to calculate sequestering agent requirements which is available under ICEPE.

Other additives should only be used after tests have demonstrated the need for them. For example a silt suspending agent should only be used if tests show that a significant amount of fines will be generated. Likewise a demulsifier should not be applied if the crude oil/spent acid does not show an emulsification tendency.

5.5 Remedial sand control

Remedial sand control in the context of this section is defined as a treatment designed to reduce sand production to tolerable levels in existing wells with or without sand exclusion previously installed.

As discussed in the introduction to this manual an increasing number of operators and Opcos are now willing to manage the risk of sand failure. The requirement for reliable, cost effective remedial forms of sand control is becoming increasingly important.

5.5.1 Industry developments

Well completion and maintenance costs can be significantly reduced if specified pumping and or coiled tubing units are deployed for workover operations. In many areas, rigless workovers and completions are becoming more popular with numerous operators (including several Opcos) reporting higher success ratios with through tubing remedial operations. The major sand control service companies are now actively marketing a suite of remedial sand control services.

Remedial sand exclusion includes many different techniques from beaning down production to reduce drawdown to sand bailing. There are several basic through tubing methods methods which can be considered for remedial operations. Method selection is generally driven by the mechanical configuration of the well (casing size, tubing size, completion accessories, length of completion interval, formation type etc).

5.5.2 Through tubing sand control

Historically through tubing sand control was considered for marginal wells were a major rig re-entry could not be justified. However, field experience clearly indicates that the chance of successful remedial sand control is substantially reduced if significant quantities of sand have been produced. This is generally attributed to the presence of cavities (which leads to poor placement) and additional problem of excessive losses exacerbated by reservoir depletion.

Due to new procedures advanced in line with recent developments in coiled tubing technology, through tubing sand control is becoming an increasingly used technique.

5.5.2.1 Mechanical techniques

The deployment of through tubing screens (WWS, PPS) inside tubing, casing or inside failed gravel pack liners, conveyed by wireline or coiled tubing conveyed systems has been widely applied by a number operators.

With such techniques screens alone are used or gravel can be placed either before or after running the screens.

With the those techniques general gravel pack design considerations normally apply. The following being particularly important to note:

·Screen overlap of 3m above and below the completion intervals.

·Screen OD will be restricted by smallest completion accessory.

·The wellbore must be clear opposite the completion zone.

·Clean completion practices is, as normal, a job prerequisite.

·Prepacking existing and new perforations is generally recommended.

Although such methods are potentially cost effective solutions, they should be viewed with some caution, as "cluttering" the wellbore with addition hardware may well present a host of additional problems. Mechanical through tubing sand control solutions are therefore generally considered as a last resort. Experience within the Group with this type of workover is limited.

5.5.2.2 Chemical systems

The most attractive advantage of chemical consolidated completions is that they lend themselves to overall completion flexibility - and eliminate the need for major rig re-entries.

Chemical through tubing sand control techniques employ chemical resins or resin coated gravel (or a combination) injected into the formation to provide in-situ grain to grain bonding.

a. Chemical consolidation

A follow up on chemical consolidation techniques has led to the successful application of several new chemical consolidation systems suitable for both initial and remedial through tubing sand control. Chemical consolidation is routinely carried out in SPDC (and to a lesser extent in Shell Gabon) where it has proven to be an efficient, cost effective alternative to gravel packing.

Design considerations - The success of such treatments is primarily dependent on the quality of the formation and effective coverage over the completion zone. As discussed in Section 13 the operational limitations of chemical consolidation are presently : BHST temperature (35-110°C) range, reservoir quality (< 20% clay) and interval length (typically 3m)

One of the main design considerations with chemical through tubing sand control is the volume and geometry of possible voids behind casing created by excessive sand production. At present these cannot be quantified, nor can produced volumes be sensibly estimated. This often leads to difficulties with optimal placement design. This design difficulty is "overcome" in several areas by assuming that general dilation of wellbore material occurs. For example under certain conditions of rock strength and in-situ stress levels the produced sand volumes may largely be compensated by a general loosening - redistribution of the rock matrix. Nevertheless through tubing chemical consolidation is often designed on the basis of pumping excess volumes to allow for inaccuracies in pore volume or assumed cavity size.

Overflush systems are generally prefered for such applications due to the possible presence of cavities behind the casing, maintaining completion equipment resin free, operational simplicity and shorter curing times. If a phase separation system is used, the resin could precipitate out in the cavity, falling to the low side, plugging the formation and reducing productivity.

Placement methods - Chemical consolidation can directly be injected through the production tubing, a dedicated work string, coiled tubing or snubbing unit. However as previously discussed in Section 13 the risk of pumping overflush systems directly through production tubing, and its effect on the future performance of subsurface equipment (SSD's, TRSSSV's) should be properly addressed.

The following factors should be carefully considered when evaluating chemical consolidation as a remedial tool.

1.Remedial chemical consolidation should be applied through a dedicated work string: coiled tubing or snubbing unit. Where possible the job design should avoid killing the well. In rare cases where several intervals are required to be treated sequentially the selective placement tool may be considered

2.Volume to be pumped is a function of the length of interval to be treated and pump rate restrictions. Note that excessive pressure drop may limit coiled tubing applications to 1 bbl/min (or less).

3.In the event of operational problems contingency must be incorporated to still allow displacement of the work string.

4.Depth control and wellbore deviation.

Utilising coiled tubing may present some limitations with respect to depth control when employing treating packers. For such applications above 60° deviation "stiff" work strings (eg snubbing strings) may be more appropriate.

5.Repeated consolidation attempts are generally not recommended as an excessive drop in permeability will occur.

6.Unless an existing zone is expected to suffer high skin values, re-perforation of the completion interval should not be considered.

7.Once a massive sand failure has occured, wells repaired with chemical through tubing sand control are statistically likely to fail on average after 3-6 years. Additionally, the time to second failure is generally much less than the time to first failure.

8.Treatment of an interval with chemical through tubing sand control before massive failure occurs generally leads to longer sand "free" production.

9.For longer treatment intervals it is possible to carry out chemical consolidation by modifying the chemical systems to facilitate self diverting.

b. Resin coated gravel

This technique evolved due to the lack of success of early chemical consolidation systems. A typical operation involves cleaning out existing perforations (and the wellbore) and squeezing sufficient quantities of resin coated (applied on the fly or during the manufacturing phase) gravel through a dedicated work string into perforation tunnels. Essentially this type of operation is a one step pre-pack and consolidation.

Most of the commercially available systems are run in water based carrier fluids. The slurry is typically batch mixed and squeezed through perforations until screen out occurs. After curing, surplus gravel is drilled and cleaned out prior to putting the well back on production. For a successful treatment sand must be placed behind every perforation, which is an inherently more difficult operation than for example uniformly placing epoxy resins.

5.5.3 Planning remedial sand control

Despite the fact that the most effective sand control techniques are those implemented early in the life of a well before sand production occurs, there remains great economic incentive to delay the installation of downhole sand exclusion (Section 3). However with this development strategy (and subsequent operations philosophy) in mind, remedial sand control must be "designed" into initial completions to facilitate well management and maintenance.

In the past, through conventional type completions, the success of remedial treatments was dependant on the performance of high expansion ratio packers. To date results with such service tools are poor. In this respect experience has shown that conventional completions using a host of completion accessories restrict access and future wellbore operations, and ultimately increase operating costs. However, in combination with full bore completions (i.e. monobore) such tools present opportunities for more reliable, less costly through tubing workovers.

In summary, if sand exclusion is likely to be a requirement during a wells life, the required functionality should be specified during the design stage, and not as an after thought.

5.5.4 Remedial operations: liner vibration and rotation

There are a number of remedial operations which may assist in removing pack bridges and voids after gravel pack installation. This section discusses some systems commonly available.

The civil engineering industry has been using vibration tools to remove voids in setting cement (concrete) for many years. Since the late 1970's similar tools have been developed for the oil and gas industry and have in some cases shown to be practical in promoting a tight gravel pack, with claims that vibration reduces the degree of fluid loss required to fill perforation tunnels and increases perforation pack factors. This section reviews two commercial systems available from Solum and Dresser Atlas, the former being a work string conveyed systems the latter wireline conveyed.

5.5.4.1 Solum system

Gravel packing with liner vibration has been adopted as suitable means of improving gravel placement in highly deviated wells. Rotating the work string throughout the gravel packing operation in high angle wells prevents gravel from building up on the lower side of the hole reducing the risk of early bridging. A number of workers have investigated the effects of vibration on the degree of compaction and rate of fluid loss to the formation during the gravel placement process.

The Solum Vibra-Pak is a rotary compactor liner vibration system which claims to attain maximum gravel compaction while gravel packing. The tool is designed to utilise rotary power to transmit vibration inside the gravel pack liner. A tailpipe rotation of 60rpm results in 3 Hz exitation, which is claimed to be effective 10m either side of the tool.

Vibration maximises compaction by eliminating bridging and promoting hexagonal packing as opposed to the less dense cubic pack. The tool has also been used to excite casing during critical casing cementations. To date the tool has been run in over 300 wells (mainly in the US).

However, two spring centralisers placed either side of the tool where severely damaged with the top centraliser losing 1 blade, and the bottom one losing three. On the basis of this (limited) yard test the following conclusions where made:

·The tool should be field tested during actual gravel packing operations to fully evaluate its performance.

·Centralisers should not be run with the vibrator.

5.5.4.2 Dresser Atlas

A combined wireline system utilising a nonfocussed gamma -gamma density device coupled with a mechanical (eccentric-cam) vibrator is available from Dresser Atlas for the identification and correction of gravel pack deficiencies.

When pack voids are indicated on a first pass log, an eccentric weight located in a vibrator housing is rotated at high frequencies (typically 3000 rpm) which transmits mechanical energy through the well bore fluids to the "packed" annulus. The housing protects high speed collision of the weight and liner. Vibration induced turbulence within the pack combined with gravitational forces disperse sand bridges and increase pack. It should be noted that several passes are normally required.

6 Other mechanical sand control methods

6.1 Non-gravel packed screen

6.1.1 Scope for application

Non-gravel packed screens (SL, WWS or PPS) are probably the oldest form of sand control and success stories have been claimed by operators in a variety of instances. Some operators report the successful use of non-gravel packed screens in shallow oil and water wells where the formation sand is typically non-uniform, medium to coarse grained. 

·Formation sand will eventually move into the wellbore causing a productivity decline due to the reduction in effective borehole diameter. This may be further aggravated by mixing of different sand sizes, clays, shale debris,etc.

·Slots are susceptible to plugging during installation of the screen and during production.

·Because a bridging criterion is generally used, the slots can erode before stable sand control is achieved.

It has been generally accepted that non-gravel packed screens eventually will cause significant impairment in a cased hole completion because of high pressure drop in sand filled perforation tunnels. In open hole completions, this effect is of lesser magnitude due to the absence of perforation tunnels. However some impairment should result from the collapse of the formation sand around the screen. Another problem is that the retrieval of a failed screen will be more problematic than in cased hole.

Field experience has generally shown that non-gravel packed screens have limited application. Nevertheless, this method should certainly be considered in the following cases:

·When more efficient sand control methods are not technically or economically feasible.

·For production testing of exploration/appraisal wells.

·Horizontal wells

6.1.2 Screen dimensions

The OD of the screen should be as large as possible to minimise formation sand movement. A minimum casing radial clearance should be maintained to allow for wash-over operations if required.

6.1.3 Slot sizing

Recent Group experience with horizontal completions has encouraged the application of slotted liners as low cost completion technique. It is therefore anticipated that an increase in the use of slotted pipe will be observed, especially in marginal developments 

6.1.3.1 Slot sizing criteria

For gravel pack applications an absolute stoppage criterion has to be used to retain high permeability gravel in place.

For non gravel pack applications applying the same criterion has not been previously possible since machining slots smaller than 0.012" proved impractical for conventional machine tools, due to high tool breakage costs. Hence wire wrapped screens have been used almost exclusively.

To overcome this problem Coberly investigated the possibility of formulating a bridging criteria. This work carried out in the late 1930's concluded that stable sand bridges would form over slot widths twice as large as the 10 percentile formation sand grain diameter. Field experience however, proved that too much sand may be produced prior to the formation of bridges. Hence the generally (industry) accepted criterion is to size slots equal to, or smaller than the 10 percentile sand grain size.

With solid state laser technology high quality slots (0.006) are achievable in all commonly used materials for oil tubulars, including stainless steels.

6.1.3.2 Required inflow area

For all applications the ideal inflow area should be maximised. In other words the largest number of uniformly distributed slots should be aimed for, without affecting the mechanical integrity of the liner, bearing in mind the mechanical loading during installation (dog leg severity etc) and operation (packer setting forces etc). However, there is a clear economic incentive to minimise the number of slots to be cut, as costs are proportional to the number of slots.

6.1.4 Prepacked screens (PPS)

The concept of using PPS alone for sand exclusion is not new. Successful reports of this technology were first documented by the water production industry during the 1930's. The main application was then viewed to be in shallow, relatively coarse, clean reservoirs where formation slump, it was claimed resulted in a "reasonably" uniform pack.

More recently, offshore (and onshore) oil and gas developments have required highly deviated and horizontal wells which have posed major gravel placement problems. In search of insurance against the increased possibility of pack voids and ultimate gravel pack failure, many operators led the resurgence of PPS during the early eighties. Essentially the PPS was used to replace the WWS in such applications, installed to act as the last line of defence against sand production and possibly the loss of the well or costly remedial action.

There are a number of PPS designs marketed by major screen manufactures and service companies. Generally they are constructed either as a screen on which gravel is bonded or as concentric screens packed with gravel. 

6.1.4.1 Industry experience with PPS

A number of authors have observed numerous problem areas associated with the use of PPS. Initially such screens suffered from damage (eg cracking) to the gravel coating while running into the well. To some extent however this can be alleviated , if catered for in the well design (minimising doglegs etc) and proper screen protection through adequate centralisation.

A potentially more serious problem is that PPS are very susceptible to plugging (fines,scale, pipe dope etc) while running in hole and/or during circulation and during production due to fines mobilisation. Hence, it is recommended to run PPS open ended to reduce the risk of screen impairment. Proponents of PPS claim good completion practices (eg. clean wellbores, filtered completion fluids etc) help overcome this problem. Nevertheless, industry studies have shown that very small volumes of fines in the completion fluid result in major plugging problems.  

General observations

Success stories on the use of PPS alone as a primary form of sand control have been reported. However it should be re-iterated that judging the performance of such sand control measures is always difficult, as in such cases the need for PPS as the main form of sand control is debatable. Nevertheless, there is no doubt that PPS have been successfully used in unconsolidated reservoirs containing clean, well sorted, large grain sands and gravels. Their application in poorer quality reservoirs must however be viewed with caution as formation slump, or natural packing will lead to an area of low permeability around the screen. This reduction in permeability will of course be a function of grain size, sorting and clay content. It should be noted that this effect can be reduced by minimising the clearance between the screen and wellbore. In general PPS application should be viewed in conjunction with gravel packing where the risk of failure due to an incomplete pack, and the cost of an unplanned workover is relatively high.

6.2 External casing packers

A novel method, claimed to prevent sand production, was published in the middle seventies, and is based on the use of inflatable External Casing Packers, (ECP's), positioned across reservoir intervals. The concept behind this idea is that inflation of an ECP with cement, up to a pressure that approaches the fracture breakdown pressure, would impose high radial stresses on the borehole wall, and improve sand stability. The cement is allowed to harden, and the interval is then perforated through the and packer to establish production.

The use of ECP's is well known as an efficient method of zonal isolation. The packer will only improve perforation stability if, under given field conditions, it causes a reduction in the value of the perforation hoop stress. This would lead to a lower differential stress at the wall of a perforation, and therefore improve perforation stability. The packer's influence will depend on the wellbore inclination, the initial stresses and the packer inflating pressure.

In conclusion:

·Inflatable external casing packers will have an initially favourable effect on perforation stability, except for vertical or near vertical wells in extensional tectonic/ tectonically relaxed environments. In these cases the packer has an unfavourable effect.

·For (rare) areas where it can be assumed that all three principal effective stresses are equal, the packer has a favourable effect on perforation stability.

·For horizontal holes, the packer will have a favourable effect on perforation stability.

·It is possible that the positive effects of the packer pressure may diminish with time, owing to formation creep.

·The favourable effects only delay the onset of sand production. After sand production is initiated, the packer will have no further influence on preventing continuous sand influx, or massive sand failure.

·The packer cannot be relied upon as an effective long term method for preventing sand influx.

Some of the above conclusions were subsequently confirmed in a high risk (but potentially highly rewarding) field test with a 40 feet ECP in the Seria field, Brunei. The ECP was installed in a reservoir with a known history of sand production (nearby wells were all completed with gravel packs).

Following perforation (under drawdown) through the cemented ECP, the well produced initially sand free at a rate double that of a gravel packed well on the same reservoir. After 3 months of production, water cut started, and this was followed immediately by massive sand failure. The well was subsequently worked over and a gravel pack installed.