Sand, gravel properties key to optimum designs

SAND CONTROL-1

M. Babs Oyeneyin
Robert Gordon University
Aberdeen

Successful gravel packed and screen well completions require a knowledge of sand as well as gravel textural properties. These completion methods keep sand and fines from entering the well bore, so that long-term production capacity of the well is ensured.

This first of a three-part series will cover key factors that influence effective sand control. The concluding parts will present guidelines for both gravel packs and screens.

Fines, more than load-bearing formation sands, pose the greater problem for the two sand exclusion techniques. Therefore, reservoir sand analysis is the main key for controlling sand.

An integrated team approach to both sand control design and implementation from well planning through drilling to final completion is the best strategy for optimizing well performance in reservoirs with sand problems.

Well completions

Completing wells in reservoirs with sanding problems is complex, and methods for controlling sand production are constantly changing. Horizontal, extended reach, and multilateral wells further complicate these completions.

Early screen designs were generally susceptible to plugging by fines and other mechanical problems. This led to a proliferation of new screen systems.

Because screens have no industry standards, the designs have generally been based on the same criteria as in gravel packs. Coupled with this, operators now have the dilemma of selecting from many screen systems. The early poor performance and lack of basic understanding of design criteria have led to a number of joint industry studies on screen performance.

Recently, the industry is witnessing more gravel packing, possibly because the industry lacks a general understanding of screen design as well as the fact that gravel packing technology has been improved.

Completion designs

For sandstone reservoirs, the completion and production process becomes complicated if the reservoir has a tendency to produce sand. If not effectively controlled, the sand can cause plugging and erosion of production lines, formation damage due to sand migration, and erosion of downhole/surface production equipment.

Completing a well in reservoirs with sanding problems is rather complex and different methods are available for controlling sand production. The completion should include appropriate control techniques in the attempt to:

• Maximize productivity
• Minimize impairment by preventing reservoir plugging
• Minimize sand migration and massive sand intrusion.

To achieve these objectives, one needs to adopt an integrated approach for designing well completions in reservoirs with sanding problems. The design should establish appropriate design criteria that includes preparing the pay zones prior to applying the sand control methods.

The mechanism for sand production is complex and is strongly influenced by the completion procedures in a given well. Irrespective of the sand production mechanism, the first step in effective sand control is to forecast the sanding potential early in the well life so that an appropriate control strategy can be specified.

The question has always been asked, "When should one initiate sand control?" Current practices tend to base the answer on a risk-weighted decision governed by the cost against the risk involved in not doing sand control.

For good results and long-term benefits, sand control is best initiated as part of the initial completion, once the sanding potential is established. The benefits of this approach include:

• Minimum permeability damage and lost production.
• Less cost because remedial measures may cost more than primary sand control and often is less predictable.
• Better success rate than most remedial measures, which have extremely low success rates and are prone to failure.

There are several methods of controlling sand production. However, for most wells especially extended reach, deviated, and horizontal wells, the most reliable are principally gravel packing and screen system control techniques.

Well-designed and implemented gravel packing is a lower-cost option and the most reliable method for sand exclusion. The method involves running a screen in the well and pumping a slurry of accurately sized pack sand around the screen. This mechanical bridging allows the production of clean reservoir fluids through the pack sand while the sand grains are filtered from the fluid.

With proper design and installation, gravel packs can yield a long-term, high-performance completion. However, many gravel packs, despite the rigorous use of the Saucier rule or other gravel sizing rules, are known to be eventually severely damaged by the invasion of fines.

Evaluation of gravel packs over the years has revealed the following:

• Evidence of fines being produced in reasonably large quantities, suggesting ineffective bridging by the gravel pack
• Evidence of high positive skin, which is a clear indication of serious formation damage mainly due to inadequate gravel sizing and consequent pack invasion.

These problems have led to new guidelines being suggested for gravel selection and for evaluating the performance of gravel packs.

Screen completions, on the other hand, involve placing a screen barefoot across the producing interval and allowing the reservoir fluid to produce through the screen slots while the formation sand is restrained by the screen.

Screen failures have been reported in many parts of the world, especially for horizontal wells. This has led to several research efforts to evaluate the problem. A joint industry group has also been formed to attempt to improve screen completion design and procedures.

Based on an industry survey,¹ the possible causes of screen failures include:

• Screen plugging caused by high-pressure drop across screens, hot spots of localized production, and fines and shaly sand
• Incorrect procedures, materials or equipment selection including installation problems for screens, corrosion due to acid, improper cleanup, ineffective oil-base mud removal, ineffective sand control, inappropriate screen selection, and screen erosion especially in gas wells
• Poor understanding of the reservoir, especially in reservoir fracturing, grain-size textural properties such as size distribution, sorting, and sanding up due to water production.

Some fast-track suggestions on improvements have included designing and using the so-called super-clean mud, planning and testing the entire cleanup schedule for all muds, and testing the screen and gravel sizes prior to use.

This fast-track approach will no doubt lead to increased cost.

The first step in designing mechanical control methods is accurate textural sand properties. Sand analysis has been identified as a major problem affecting gravel packing/screen completion techniques.

Sand analysis

Successful gravel packing and screen systems requires a knowledge of sand as well as gravel textural properties. These textural properties include:

• Overall sand/gravel size distribution
• Sand/gravel sorting
• Sand/gravel shape.

Current field practices involve simply knowing the average percentile sizes, especially the 50 percentile size for gravel-pack design and the 10 percentile size for screen slot gauge.

Gravel selection

A number of equations exist for selecting gravel but the most popular is the Saucier rule that is based on the 50 percentile size of the formation sand.

Most current applications use the modified versions of the Saucier rule to select gravel size. This is also applicable to proppant size selection of prepacked screen systems and fracpacks.

The Saucier rule is defined as d50(gravel) = 5 to 6 3 d50 (sand).

Conventional gravels are typically silica sands but a number of highly symmetric, well-rounded synthetic gravels are available. Synthetic gravels are relatively more permeable and have smaller pores that provide better bridging effectiveness against sand migration.

Field experiences show that the Saucier rule may be too general and limited in that it does not account for the overall sand texture, well bore configuration, and operating conditions.

As shown in Fig. 1 [47,879 bytes], two different sands can have the same mean size, but different size distribution, sorting, possibly shape, etc. Therefore gravel pack or screen response to these sands would be different.

Likewise, gravel texture dictates its eventual pore size distribution and permeability. These are key parameters for ensuring effective bridging and maintenance for optimum producing capacity.

Formation sand

Experience shows that the load-bearing formation sands may not be a main problem with respect to the impairment of many gravel packs and screen systems. Rather, it is the presence of reservoir fines as well as solids in completion fluids that may be responsible for the high skin observed.

Thus, knowledge of formation sand textural properties is essential. A first step is to properly sample the formation over the pay interval. The following six-step procedure is recommended:

1. Obtain representative samples from all producing intervals for every well completed. These can be obtained with comprehensive mud logging of cutting samples through optimum hole cleaning strategy for a well, especially for horizontal wells, and complementary whole core sampling.
2. Do a full sieve analysis after disaggregating samples washed clean of reservoir fluids. In some cases, dry sieve analysis may yield poor results, especially if very fine particulates are present in large numbers. The electromagnetic forces between grains tend to promote aggregation and thus false evaluation of the percentage of fines present in the formation sand. The dry test sieving should be complemented with the wet sieving technique and/or use of a more-sophisticated particle analyzer.² Sieve analysis should be done for each interval sample. For heterogeneous pay sands, mixing the individual samples to form a combined average sand size distribution is recommended. Attempts to use the coarse sand as the representative sample for determining gravel size may lead to fines invasion and plugging while the use of the finer grains may lead to the selection of a smaller gravel which may result in total sand exclusion but poor flow capacity.
3. Plot both the cumulative sieve analysis curve and the weight percent profile for each interval sample as well as the average size distribution.
4. Compute the 5, 10, 50, 90, and 95 percentile sand sizes to quantify size distribution from coarse through the mean to the fines present in the sand.
5. Compute the sorting with either Equations 1 or 2 (see equation box [28,128 bytes]). In these equations d is the sieve size in millimeters at the 84 percentile on the cumulative sieve analysis curve, and f84 is the phi-unit at the 84 percentile on the size distribution curve. Based on this correlation, the sorting can be categorized as follows:
   • Very well sorted-S < 0.35
   • well sorted-0.35 < s < 0.5
   • moderately well sorted-0.5 < s < 0.71
   • moderately sorted-0.71 < s < 1
   • poorly sorted-1 < s < 2
   • very poorly sorted-s 2.
   In the modified Begg's equation the sorting classification is:
   • Well sorted-S < 1
   • poorly sorted-s 1.
   Sands can also be classified according to their degree of uniformity defined as C = D40/D90, where C is the uniformity coefficient and D40 and D90 are the sieve sizes at the 40 and 90 percentile. For C < 3, sand is uniform but for c > 5 sand is nonuniform. Some confuse the uniformity coefficient with sorting. It is possible for a nonuniform sand to be well sorted, but every poorly sorted sand would be nonuniform. While uniformity covers a smaller band of size distribution, sorting determines the overall size distribution and accounts for the finer grains.
6. Carry out the appropriate petrographic analysis. The mud logger or production geologist should use the most up-to-date technique for shape analysis. A Quantimet microscope is recommended. The analysis provides quantitative values of size distribution and especially shape in terms of sphericity.

Gravel analysis

API specifications exist for testing many commercial gravels especially in terms of strength, acid solubility, percentage of fines, etc. However, these tests do not generally include assessment of the size distribution and impact on gravel porosity, pore size distribution, and permeability.

Currently, the industry has no established standards for testing gravel pack permeability.

In general, commercial gravel permeability, porosity and pore size distribution, and key properties for effective bridging and production capacity are a function of not only the average size but also of grain size distribution, shape, and turbidity. In most cases, quality control addresses only the grain size in terms of the percent of oversize and undersize particles present.

Usually no one attempts to analyze the grain size distribution in terms of spread and skewness.

Spread refers to the distribution of the particle sizes around the mean, and skewness refers to where the mean falls.

Variation in the weight percent of the particle sizes can lead to different pore size distribution, porosity, and permeability values. Reports of variations between 30 and 100% on permeability have been reported.³

Laboratory analyses of a variety of measurements show variations in the permeabilities and porosity of commercial gravels of the same size.³ For example, low fine 80/120 U.S. mesh gravel had 19.01-24.25 Darcy permeability. Corresponding porosities were 33-35%. The 20/40-mesh gravel had 183-342 Darcy permeability with a corresponding porosity of 30-36%.

The 12/20 mesh commercial gravel had 585-1,267 Darcy permeability with a corresponding porosity of 30-37.8%.⁴

Synthetic gravels with similar sizes as conventional gravels have also been found to exhibit much higher permeabilities, smaller pore size distribution, and porosity.^{3 4} Adjusting the grain size distribution with respect to the individual weight percents can yield different pore size distribution as well as permeability. These are two key properties required for effective sand control and maintenance of optimum production capacity.⁵

It is therefore crucial that the quality control procedure should be extended to analyzing for size distribution, especially when sourcing gravels from different suppliers.

Gravel packs vs. screens

Theoretically, with good planning and adoption of optimum design criteria, any of the two mechanical options for sand control can be used. Nevertheless, there are limitations on the application of the different systems. A summary guideline on the applications is presented in . Table 1 [119,053 bytes]

References

1. Perdue, J.M., "Completion Experts Study Gulf of Mexico Horizontal Screen Failures," Petroleum Engineer International, June 1996, pp. 31-33.
2. Hunt, T., "Monitoring particles in liquids, Filtration and Separation, March 1995.
3. Cocales, B., "Optimizing Materials for better gravel packs," World Oil, December 1992, pp. 73-77.
4. Oyeneyin, M.B., Peden, J.M., Hosseini, A., and Ren, G., "Factors to consider in the effective management and control of fines migration in high permeability sands," SPE Paper No. 30112, 1995.
5. Ren, G., Oyeneyin, M.B., Peden, J.M., Bigno, Y., and Hosseini, A., "3-D Modelling of gravel pack structure and its application to optimum gravel size selection," SPE Paper No. 24999, 1992.

The Author