Roy D. Wilcox
Baroid Drilling Fluids Inc.
Houston
Clay stability models based on surface-area, equilibrium water-content pressure relationships,' and electric double-layer' theory can successfully characterize borehole stability problems.
Because borehole problems are complex, a total systems approach to their solution is required.
Shale-swelling pressures and water requirements of drilled solids and mud additives must be considered for minimizing borehole problems with water-based mud. The application of surface area, swelling pressure, and water requirements of solids can be integrated into swelling models and mud process control approaches 3 to improve design of water-based mud in active or older shales.
Well bore stability and solids problems caused by weak and expansive shale rocks continue to be a major source of inefficiency when drilling a well with water-based drilling fluids. Improvements in polymer technology, feed systems, solids control, and basic understanding of shale behavior in various water chemistries are needed because of high-angle drilling and increasing environmental restrictions both in the Gulf of Mexico and onshore.
Drilling-fluid design for shale stability requires a total systems approach once the appropriate water-based chemistry for stabilizing the troublesome shale has been selected. Even inhibitive water-based systems can fail if improperly engineered at the well site.
Good well-site engineering requires maintaining the proper chemical concentrations and dilution rates so that the inhibitive mud can remove solids and stabilize the well bore.
Both chemistry and mud density may be required to stabilize a shale. Good chemistry does not imply that high salinity is automatically required for shale stabilization. In fact, swelling pressure theory shows that in some cases high salinity can be detrimental to shale stability.
Optimization of mud salinity, density, and filter-cake properties are all important to achieving optimum shale stability and drilling efficiency with water-based mud.
SWELLING PRESSURE MODELS
Both the electric double-layer model described by Van Olphen 2 and a clays-welling model by Low 3 can be used to calculate the swelling pressure of clays. The double-layer model is much more complex in terms of the data required and mathematical calculations.
A statistical model derived from Low's data is simpler and requires only shale-specific surface area and equilibrium water content to estimate swelling pressure. Each model has specific applicability, depending on shale type.
Recent quantification of capillary suction time (CST) data to surface area has made it more practical to test the application of the equilibrium water content and double-layer models to actual shale-swelling pressures, reduction in swelling pressure through ionic and chemical inhibition, and water requirements of solids.
Experimental data by Low on 35 montmorillonite clays with surface areas from 200 to 800 sq m/g show that the equilibrium water content under constant pressure increases with increasing surface area. Also, the pressure to drive the water out increases exponentially with decreasing water content.
At low water contents the swelling pressures, or hydration energies, are extremely high. As the water content is increased, the swelling pressure between solids decays rapidly. Thus, by meeting the water demand of a solid based on its surface area, swelling pressure between solids is reduced.
Statistical analysis of Low's data shows that the equilibrium water content is not dependent on cation exchange capacity, but more highly dependent on specific surface area. A model derived from the data presented gives the following relationship with approximately a 0.97 correlation coefficient for 35 clays at seven pressures, or 275 degrees of freedom:
[SEE FORMULA]
Fig. 1 summarizes the swelling pressure/surface area/water content relationship for solids of varying surface area. Also, a model which relates the swelling pressure to distance between particles has been derived showing:
[SEE FORMULA]
The precision of this Equation in terms of the adjusted r-squared is 0.94.
Separation distance in Angstroms is obtained from:
[SEE FORMULA]
where Vw is water volume in a unit volume, and Y., is the solid volume in a unit volume.
Separation distance is a key variable used in the electric double-layer model for calculating clay-swelling pressures. Calculated pressures using Equation 2 run somewhat less than for Equation 1.
CAPILLARY SUCTION TIME
The CST measurement has been treated quantitatively to obtain a surface area using the Kozeny equation. 4-6 Quantification of the CST value required development of a constant for the CST cell based on paper pressure drop, filtration area, and filtrate volume to calculate a specific resistance to filtration at 0.5 psi, or the paper filtration pressure.
Cake water contents were also measured simultaneously to obtain cake porosity. The equations and an example calculation for surface area of Wyoming bentonite are:
[SEE FORMULA]
Agreement between surface area obtained by CST and mineralogical analysis based on percentage composition and assigned surface areas for each mineral has been observed. However, since smectite surface areas can vary widely, it can be difficult to rely on X-ray analysis without observation of swellable layers.
An empirical correlation model between CST value corrected for solids content and surface area with a 0.89 correlation for 90 samples is in Reference 5. Currently, both the empirical model and analysis of cake porosity along with CST values are being used for surface-area estimation.
FLUID INTERACTION
The CST measurement was introduced as a method for monitoring drilling-fluid interaction with shale cuttings. Fig. 2 shows basic interaction curves of Wyoming bentonite with both monovalent and divalent cations used to inhibit active shales in drilling.
The reduction in CST values with increasing ionic concentration is a result of a decrease in effective surface area of the solids through flocculation. Maintaining excess concentration prevents going back up the curve and increasing surface area.
The exponential decay in CST value with increasing salinity in Fig. 2 is also comparable to the exponential decay of surface potential with ion concentration and distance in the electric double layer model:
[SEE FORMULA]
where d is distance, and K is the characteristic thickness of the double layer with:
[SEE FORMULA]
The total electric potential midway between the plates can be represented by the schematic in Fig. 3.
The total swelling pressure between clay plates is dependent on the electric potential u between plates, the surface potential Uo, and the separation distance. The true potential u cannot be determined experimentally, but must be calculated in accord with double-layer theory. 7
The ability of cations to reduce the surface potential depends on their charge density and ability to diffuse into the pores. Flocculation rate studies which assess the stability ratio are based on diffusion rate of ions into the pores.8
In highly compacted shale formations, with close interparticle separation, the ability of ions to diffuse becomes limited. Some CST data have shown that increasing ion concentration crowds the surfaces, resulting in increased dispersion. These detrimental effects have also been confirmed in the borehole.9
The addition of mud additives may increase or decrease the filtrate ionic activity. Generally, flocculants decrease the ionic activity and dispersants increase activity. However, since all materials can behave as flocculants or dispersants depending on their concentration to solids, actual filtrate must be tested to assess its stabilizing properties.
MATERIAL BALANCE
The key to maintaining a good water-based drilling fluid for stabilizing shale is to maintain the proper liquid-to-solid balance in the filter cake. Depending on shale formation's osmotic pressure or sand permeability, approaches for controlling water loss to the formation will vary.
In permeable sands, studies 10 have shown that the proper size bridging material is the key to reducing wall-cake buildup.
In active shale formations, the proper control of water loss is actually achieved by maintaining the proper liquid-to-solid balance in the cake, or porosity. 11 Active shale formations which produce compressible particles are subject to flocculation and spontaneous dewatering above a critical solids concentration with only slight increases in pressure. 12
Accounting for water requirements of both formation solids and drilling-mud solids and additives is essential for improving design and optimizing dilution rates of a water-based mud. The basic concept of design for bound water can be derived from porosity, pressure, surface-area relationships.
While much work remains to be done on the subject of solids-water demand, basic equations can be used to estimate material requirements based on porosity considerations. The basic equation for porosity, which decreases with
[SEE FORMULA]
As porosity decreases with increasing pressure, so does the water content. If porosity and volume of dry solids are known, then volume of wet solids, Vws, can be calculated from:
[SEE FORMULA]
As an example, if porosity of cuttings is 0.75, and assuming complete dispersion of solids, then Vws for 1 ft of 12.25-in. hole or 0.145 bbl of solids is:
[SEE FORMULA]
The volume of dry solids plus bound water becomes 0.584 bbl. This new volume of solids plus bound water becomes the volume of solids to design the mud around.
Thus, Vws becomes Vs'. Then the total volume of water mud to build around Vs' to keep a constant mud liquid-to-solid ratio becomes Vwm = Vs'/(1-o) where Vs' = Volume of dry solids plus bound water in Equation 9.
If mud cake porosity is 0.75, then the volume of mud to build around 0.584 bbl of solids plus bound water to keep a constant liquid-to-solid ratio in the cake is:
Volume water mud+ solids+ bound water = 0.584/ (1 -0.75) = 2.33 bbl of total volume or 2.33-0.584 = 1.752 bbl of mud + 0.43 bbl of bound water+0.145 bbl of solids (10)
By introducing a solids control efficiency of 60%, the amount of mud required to build around 1 ft of 12.25-in. hole is:
[SEE FORMULA]
Equation 11 assumes gauge hole.
Factors such as formation surface area and solids control efficiency can seriously alter the calculated dilution rate.
These dilution requirements also must be met on a per-minute basis to maintain a constant system. Thus at 100 ft/hr, or 1.66 fpm, the following is obtained:
[SEE FORMULA]
Approximately 75% is whole mud and 25% is dilution water.
The calculations are designed to show the importance of water to cuttings for drilling-fluid design purposes and operations.
Development of specific requirements by drilling area and mud type is required to minimize solids-related borehole problems, because swelling pressure's are greatest at the lowest cake water content.
STABILIZING ACTIVE SHALES
Various drilling-fluid systems are currently being used for stabilizing active shales. The primary materials are PHPA (partially hydrolyzed polyacrylamides), starch, grafted polymer calcium lignosulfonate lime, CMC (carboxymethyl cellulose), and ferrochrome lignosulfonate. While this generally represents the order of shale-stabilizing properties, all systems can be properly engineered to produce good shale stability in a wide variety of shales. System economics based on dilution factors separate mud performance.
PHPA
PHPA systems have been used to control shales for many years.13 PHPA is a high molecular weight polymer and varies in anionic charge. Low-charge anionic PHPA's are more suitable for pit-solids flocculation, while the 30% anionic PHPA's are used for control of shale stability. PHPA's build viscosity in fresh water, but not in salt water. Salt water PHPA's with sodium chloride concentrations of 10-20% have become popular in the Gulf of Mexico for controlling gumbo shale and preventing gas hydrates. CMC's are used to control fluid loss or cake porosity. Also, calcium carbonate has been found to be an effective bridging agent in porous sands.
These systems are very solids sensitive, and mud cakes tend to dewater under low pressure. Fluid loss and solids content in both active shales and sands are important properties required to control shale stability with this system. Optimum dilution rates and solids-control equipment are mandatory. Also, mud-density requirements must be met to control hydrating shale.
KCI-STARCH-CMC
KCI-starch-CMC has been used for many years to control swelling shales.
Optimum KCI content has been found to vary from 3 to 6% or higher (e.g, 8-12% in the North Sea) depending on shale activity.
Fluid-loss control with starch is very important for reducing shale swelling. Upper limits on balancing shale osmotic pressure with KCI can be reached in compacted shales. Solids control is extremely important for this system.
LIME-KOH-PGCL
The lime-KOH-polymer grafted calcium lignosulfonate (PGCL) system, which has been introduced recently and used extensively over the last several years, 14 15 16 takes advantage of the inhibitive properties of a lime-based mud. CST data in Fig. 1 show that lime and KOH act as powerful clay inhibitors.
When applied in active plastic shales, both gauge hole and drilling-fluid rheology control can be maintained at higher solids levels than either of the nondispersed polymer systems. Excess lime of greater than 3 lb/bbl must be maintained along with adequate concentration of the PGCL, which acts as a coating deflocculant."
While this system tends to run very thin, solids content above 6-7% by volume must be avoided.
As with many drilling-fluid systems, Theological properties can be deceiving with regard to downhole performance.
POTASSIUM LIGNITE
The potassium lignite (K-Lig) system was developed in the early 1970s. 17 A recent publication 18 by T. Hemphill documents its use in the Gulf of Mexico in the 1980s showing the stabilization generated by 2,000 ppm potassium-treated lignite. Apparently, these systems exhibited tremendous solids tolerance and still produced gauged well bores, while avoiding many gumbo-related solids problems.
SHALE BEHAVIOR
Shales are formed by the compaction and dewatering of colloidal clay material from river deposits. The compaction process of the varying surface-area clay minerals results in formations of varying bulk densities and shale particle aggregates of variable 1/2 distance between plates. Shale compaction or separation distance over a wide range of bulk densities and surface areas is illustrated in Fig. 4.
Once a 1/2 spacing or surface area has been estimated using Equation 3, then a swelling pressure can be calculated if the water content or bulk density is known (Tables 1 and 2).
Fig. 5 shows how osmotic pressure of shales varies with separation distance, using both electric double layer and equilibrium water content.
CASE HISTORIES
Three cases illustrate how CST values were used to determine mud additives needed in wells with tight spots, hole enlargement, and low mud weight.
TIGHT SPOTS AND SLOUGHING
The first full application of CST to a well in Northwest Colorado was based on laboratory studies and previous field experience for specific shale intervals that showed that maintaining low CST values of mud filtrate against shale would produce better borehole stability.
While basic theory was understood, quantification of CST into surface area and for pressure calculations was not possible at the time. However, when CST values were very low and shale problems persisted, the problem was not from the mud. In this case, either mud weight was needed to be increased or fluid loss lowered and filter cake quality improved.
Fig. 6 shows the progress of the well from 9,000 to 14,000 ft as judged from CST data in distilled water, mud filtrate, and caliper logs. Also, comments which indicated trouble spots are shown.
At 10,360-10,404 ft CST values in filtrate against shale increased from 35 to 80 sec, with increase in sloughing and tight spots at the same time. Pilot tests showed CST values could be reduced by a combination of KCI, lime, and fluid-loss control polymer.
Treatment was implemented, the CST numbers were reduced from 80 to 36 sec, and sloughing stopped.
Tight spots due to highly compacted shale were encountered from 10,600 to 11,000 ft. The hole remained stable while logging, and casing was run at 11,040. Out of casing, active shale was encountered, KCI was increased, and no hole problems were observed. The increased washout right out of casing was attributed to hydraulics. Mud weight was increased to hold back swelling shale at 12,190 and 12,370 ft.
Swelling-pressure calculations in Table 1 show that the double-layer model gave reasonable estimates of the mud-density requirements to hold back the shales based on surface area in distilled water and 5,800 ppm NaCL shale salinity. Estimates by the Low model were high for these shales at very close separation distances.
Evidence indicates that the Low model is more applicable at somewhat greater separation distances.
The KCI polymer system was fully required to stabilize this active mountain shale. If a sodium chloride polymer system or a lime-based system with deflocculant to control rheology had been used, serious problems would have developed.
This statement is supported by expected average CST values of these systems against shale samples. For instance, while CST values of a KCI system generally ran about 40 sec against shale, NaCL polymer values of 60-70 sec and lime deflocculant of 100 plus sec would be expected.
Hole enlargement would have been expected to be significantly greater with higher values. Also, possible hole failure due to generation of larger swelling pressures would be greater with more time in the hole.
Determination of KCI concentration to control the active shale was continually needed to maintain stability. Excess KCI concentrations were not used throughout the well because of severe corrosion problems due to acid gas.
Optimized KCI concentrations were maintained to control the shale and minimize corrosion.
HOLE ENLARGEMENT
Two wells were drilled with a fresh water (2,000 ppm chloride) deflocculated-lime system. CST data were used to track and improve inhibition. Data were collected in the laboratory and used for subsequent recommendations.
Because of fast drilling rates in offshore operations, it is difficult to run CST tests fast enough with manpower limitations at the well site. As mentioned previously, young plastic shales can tolerate more dispersion.
Well 1 had good solids control equipment while Well 2 did not. Excess lime was increased in the second well to improve inhibition. The average CST values in distilled water and mud filtrate along with caliper logs are shown in Fig. 7.
The first well with high CST values of filtrate against shale of 300 sec had good solids control equipment and low-gravity solids were maintained under 60 lb/bbl. Bit balling and plugged flow line were not observed on this well. However, the hole was severely washed out.
In Well 2, CST values of filtrate against shale were reduced to 140 sec, and the hole was near gauge. Solids-control equipment was very poor, and low-gravity solids built up in excess of 100 lb/bbl.
Plugged flow lines occurred from 8,000 to 10,000 ft as a result of excess low-gravity solids.
MUD WEIGHT
A well drilled in the Gulf of Mexico with an NaCl PHPA polymer system experienced stuck pipe around 8,000 ft. CST values of close to 900 sec in distilled water or 185 sq m/g were obtained. Prior to getting stuck at 8,000 ft, mud weight was increased to 11.8 and then 12.1 ppg.
The hole was sidetracked and weight was raised to 12.5 ppg at 7,000 ft and 13.5 ppg at 8,000 ft without any hole problems.
The results in Table 2 show maximum shale pressure by Equation 3, based on surface area, shale density, and water content listed. A mud weight of approximately 4.73 ppg above hydrostatic of 8.54 ppg at 7,500-8,000 ft based on shale-swelling pressure of 1,844 psi for a total weight of 13.3 ppg was required.
Weight was raised to 12.5 ppg by 7,000 ft and 13.5 ppg by 8,000 ft. The results in Table 2 indicate that low mud weight caused stuck pipe on the original hole and that the increased mud weight to hold back the shale on the sidetrack is in line with calculated shale pressures based on surface area and bulk density data.
SOLVING THE PROBLEM
Shale stability can be monitored and improved by controlling mud chemistry, density, and material balance.
The primary difficulty in designing for shale problems is the lack of available data which can be tied into a process control approach. Understanding of the shale problem from a process control viewpoint and quick useful measurements must be integrated into a complete systems approach.
Factors such as available rig equipment, solids control, mud activity, and dilution rates must be considered based on engineering calculations which account for water required by solids.
ACKNOWLEDGMENTS
The author would like to thank Baroid Drilling Fluids for permission to publish this work.
Also, the author would like to acknowledge the following Baroid staff members for their efforts: K. Barr, D. Bilka, F. Cornay, M. Fillingame, D. Foremen, G. Guth, W. Guillot, D. Jamison, A. Liao, L. Morales, S. Shaffer, D. Siems, B. Vaughn, B. Wethington, G. Keely, and B. Schiff.
REFERENCES
- Low, P. F., "The Swelling of Clay, 11 Montmorillonites," Soil Science, Vol. 44, No. 4, 1980 pp. 667-675.
- Van Olphen, H., An Introduction To Clay Colloid Chemistry, John Wiley & Sons, 2nd Ed., 1977.
- Beasley, R. D., and Dear, S.F., "A Process Engineering Approach to Drilling Fluids Management," SPE paper presented in San Antonio, Oct. 8-11, 1989.
- Wilcox, R.D., and Jarrett, M.A., "Polymer Deflocculants: Chemistry and Application," SPE paper presented in Dallas, Feb. 28-Mar. 2, 1988.
- Wilcox, R.D., "Application of Filtration and Colloid Stability Principals To Solving Shale Stability Problems," presented at American Filtration Society Regional Meeting on Liquid Solid Separation in the Oil and Gas Industries, Oct. 29, 1989, Houston.
- Liao, W.A., and Siems, D.R., "Adsorption Characteristics of PHPA on Formation Solids," to be presented at SPE/IADC Drilling Conference, Feb. 27 - Mar. 2, 1990, Houston.
- Madsen, F.T., and Vonmoos-Max Muller, "Swelling Pressure Calculated From Mineralogical Properties of a Jurassic Opalinium Shale, Switzerland," Clays and Clay Minerals, Vol. 33, No. 5, 1985, pp. 501-509.
- Kruyt K.R. Ed. Colloid Science, Elsevier Publishing, Amsterdam, Vol. 1, 1952, pp. 278-300.
- Wilcox, R. Baroid Internal Studies.
- Fisk, J.Y., and Jamison, D.E., "Physical Properties of Drilling Fluids at High Temperatures and Pressures," SPE Drilling Engineering, December 1989, pp. 341-346.
- Siems, D.R., Baroid Internal Studies.
- Tiller, F.M., "The Role of Porosity in Filtration," Chemical Engineering Progress Vol. 49, No. 9, September 1952.
- Clark, R.K., et al., "Polyacrylamide/Potassium-Chloride Mud for Drilling Water-Sensitive Shales," JPT, June 1976, pp. 710-727.
- Neal, J.A., et al., "Lignosulfonate copolymer stabilizes lime muds," OGJ, Mar. 25, 1985.
- Gilmore, W.H., and Sanclemente, SPE/IADC 18640, "A Case History Review of the Effects of Drilling Fluids on Shale Stability in the Central Gulf of Mexico," presented in New Orleans, Feb. 28-Mar. 3, 1989.
- Elsen, J.M., Mixon, A.M., Broussard, M.D., and LaHue, D.R., SPE 19533, "Application of a Lime-Based Drilling Fluid in a High-Temperature, High-Pressure Environment," presented in San Antonio, Oct. 8-11, 1989.
- Mondshine, T.C., "Tests show potassium-mud versatility," OGJ, April 22, 1974.
- Hemphill, T., "Using Potassium to Stabilize Gulf of Mexico Shales," World Oil, November, 1989, pp. 81-90.
Copyright 1990 Oil & Gas Journal. All Rights Reserved.