M.S. Juprasert
Chevron U.S.A. Production Co.
Bakersfield, Calif.
High acid injection rates, within the friction dominant region, proved successful in stimulating oil wells completed in a naturally fractured reservoir.
The bullhead technique was effective without any chemical or mechanical diversion.
Chevron U.S.A. Production Co.'s Santa Clara field produces through wells on Platform Grace, in 318 ft of water.
The field is on OCS Parcel 0217 Santa Barbara channel, about 12 miles offshore Ventura, Calif. Production comes from the Monterey formation, a naturally fractured reservoir containing chert, porcelanite, and dolomite.
The formation has extremely low matrix porosity and permeability. But the fracture permeability can be very high and laterally extensive and is critical to production.
After repeated unsuccessful attempts to stimulate these wells, three basic types of formation damage were discovered:
- Drilling mud invasion
- Solids precipitation
- Emulsion blocks,
Accurate identification of the damage, in 1989, led to the first successful stimulation response 9 years after the first completion.
One of the keys to successfully acidizing the Monterey is to treat the damaged zones with the stimulation fluids.
In previous treatment attempts, acidizing fluid tended to enter the undamaged zones.
FORMATION DAMAGE
Because of the poor relative performance compared to expectations based on exploratory drill stem tests (DSTs), formation damage had been suspected since 1980, at the time of the first completion of a Monterey well on Platform Grace.
From 1980 to 1985, nine stimulation attempts (Table 1) failed to increase production. These failures are partially attributed to incomplete diagnosis of the damage type.
The experience gained from these stimulations, combined with drilling records, laboratory experiments, and field data from workovers, revealed that the productivity impairment may have been caused by mud invasion, solid precipitation, and/or emulsion blocks.
MUD INVASION
The mud cake, which typically, forms on the formation face while drilling sandstone formations, can retard the movement of drilling mud into the formation. In contrast to typical sandstone pore diameters 10 100, the Monterey fracture widths of 100 500 may not allow this impermeable filter cake to be formed as easily. Therefore, overbalance can force the drilling mud deeper into the fracture tunnels until plating out by solids occurs in the fracture throats.
Normally, this plugging can be removed by reverse flow. However, the differential pressure may be insufficient to completely remove the damage, and some zones may remain partially or completely plugged.
Fluid entry surveys conducted during 1986 1990 in five wells indicated that about 40 85% of perforations did not contribute to production. The surveys also showed that plugged zones varied from well to well. The wide specific gravity range (20.9 34.4 API) of Monterey, oil produced from these commingled wells also supports the inconsistency in productive zones from well to well.
Table 2 suggests that an overbalance of 1,000 psig over reservoir pressure of 3,100 psig at 8,000 ft was experienced while drilling with 10 11 ppg mud. To analyze the effect of this overbalance, the initial production of each well was plotted (Fig. 1) against the corresponding overbalance depth.
The invasion depth was obtained from a simplified calculation of the overbalance generated bil mud weight over reservoir pressure normalized by perforation depth. While this simplification is not completely accurate, the correlation was derived from the five most productive wells located on the western side of the field.
Under the same reservoir conditions and completion design, the majority of Grace wells should perform according to this correlation. However, the performance of the remaining 18 wells was greatly influenced by different drilling and completion procedures. For example, at the same calculated overbalance depth, the wells with longer exposure time (Monterey exposed to drilling mud) showed lower initial production. Moreover, the wells in the bottom envelope may have been severely damaged by excessive drilling mud because of lost circulation, fishing, and plug back operations.
SOLIDS PRECIPITATION
The Monterey formation water at Grace contains an unusually high level of bicarbonate concentration, averaging 3,000 mg/l. In contrast, only 137 mg/l. were found in Santa Barbara channel sea water and less than 1,000 mg/l. were tested in other Monterey fields.
Based on thermodynamic equilibrium, precipitation of carbonate scales can occur by raising the pH of the formation water from its normal range of 7 8. The mud pHs were raised from 9 10 in exploratory wells to 10 12 in development wells, and this may explain why the initial production from the development wells was much lower than DSTs indicated in exploratory wells (Table 2).
To explore the alkalinity effect on Monterey formation water, composite produced water samples were adjusted to pHs of 8, 9, 10, 11, 12, and 12.5 with caustic soda, the same product used for pH control in drilling mud. Fig. 2 shows the analysis after the samples were heated to 190 F. (reservoir temperature) and the total suspended solids and the cations in the filtrate of each sample were analyzed and plotted.
An X ray diffraction analysis on the precipitates indicated that the major portion of the solids from pH 8 to 10 was calcite (CaCO3) and from pH 11 were silica gel (noncrystalline silica), and brucite (Mg(OH)2). Bailed samples of scale confirm these precipitates.
In addition to the damage caused by drilling mud invasion, both exploratory and development wells in the Santa Clara field were also damaged by carbonate scales.
Furthermore, the development wells were severely damaged by brucite and silica scale. It should be noted that the elevated concentration of calcium in drilling mud (mostly from make up water) and CaCl2 brine used in completion fluids of the early development wells would have aggravated the calcium carbonate scaling tendency (Table 2).
EMULSION BLOCKS
Emulsions can occur between two immiscible liquids and can be stabilized by fine particles or asphaltenes collected at the interface.12 Lab analysis of Grace's produced crude indicated asphaltene as high as 10% and paraffin of 3%. Oil and water emulsions in the formation near the well bore can drastically reduce well productivity.
Extraneous water from drilling mud and killing fluid entering the formation could have become emulsified with the crude. The majority of Grace wells produce emulsified crude in varying degrees (water in oil emulsion). Most of these emulsions are believed to be caused by agitating and mixing the produced fluids with lift gas.
Experience has shown that Grace wells suffered a 50 100 bo/d production loss per well after completion of workover operations. These production losses could have been caused by emulsion blocks. The blockage was also evidenced by the emulsion recovered from Well A 11 perforating guns. An additional 320 ft were perforated with strip guns in April 1990, with negligible production gain.
This emulsion consisted of 63% oil, 35% water, and 1-2% solids. An X ray diffraction analysis indicated that the major constituents of the solids are calcite, barite (BaSO4), and halite (NaCl).
A very stable thixotropic emulsion similar to the one seen in Well A 11 was created by mixing potassium lignite drilling mud with wellhead crude samples. The plastic viscosity measurement of this emulsion indicated a seven fold increase from pure mud. A beaker of this emulsion could be held upside down without pouring out the emulsion. The same tests were conducted for other Monterey producing fields, their emulsions are movable, and only a 2 3 fold increase in plastic viscosity was observed.
Most of the Grace wells were drilled with inhibited, dispersed mud systems, while the other fields used low solids, nondispersed systems.
TREATMENT DESIGN
The stimulation for Grace wells is designed to remove or bypass the damage caused by drilling mud, scales, and emulsion, so that natural formation permeability can be restored. The stimulation process consists of treating the well with surfactant, allowing it to soak overnight, then stimulating with acid. Well unloading time (1 2 days) is considered as an acid soaking, period. All treatments are bullheaded in single stages without chemical or mechanical diverters.
Once damage types are identified, proper selection of treating chemical is the key to successful stimulation. Although chemicals identical to those used in past stimulation work are being used, the additive arrangement, volume, and placement are different.
SURFACTANTS
In practice, a surfactant should be employed in all acidizing jobs to prevent damage caused by water and emulsion blocking, oil wetting, and clay dispersion. Typically, less than 1% by volume is mixed in the acid.
However, removing damage requires many times the volume of surfactant needed to prevent damage. It is recommended that a 1% solution of surfactant in a carrier fluid be injected for each 100 gal/ft of interval treated.1 The current treatment is designed for 20 30 gal/ft of 3% surfactant mixed in xylene.
Xylene is used as the carrier fluid to ensure surfactant miscibility. Because xylene is noncorrosive, it will not adversely affect downhole equipment during an overnight soaking period. Xylene also has the capability of dissolving both asphaltene and paraffin deposits that might be shielding scales and fines from acid contact.
ACID
Mud acid treatment 3 normally consists of sequentially injecting three fluids as follows:
- Preflush
- Hydrochloric acid/hydrofluoric acid mixture (12% HCI and 3% HF)
- Postflush.
The preflush and postflush are usually HCI acid, ranging in concentration from 5 to 15%. All acids contain additives (e.g., surfactant, corrosion inhibitor, desludging agent) as required.
The primary purpose of using HF acid is to dissolve formation clays, sand, and drilling mud. The HCI acid preflush reacts with calcareous material in the formation and also displaces connate water ahead of HCI/HF acid mixture to prevent the formation of sodium and potassium fluosilicates, and calcium fluoride, insoluble precipitates capable of plugging the formation. The postflush is required to isolate the reacted HF from brine and push secondary damage, if any, away from the well bore.
A very large increase in well productivity was observed after mud acid treatment in other Monterey fields, although in the past this type of treatment was employed exclusively in sandstone formations.
Application of mud acid in one of Grace's wells in 1990 was an economic success (Table 1). The precipitation of HF acid reaction products can potentially reduce productivity as secondary damage. Therefore, it was decided not to use HF acid in treating Well A 8 and subsequent jobs. Uncertainties regarding HF acid application at Grace include:
- Abundant HCI-soluble material A good rule of thumb is if the formation is more than 20% soluble in HCI acid, it should be treated with HCI only.1 About 10 15 % of dolostone (CaMg(CO3)2) and marl (CaCO.3 and clay) are present in Grace Monterey. Unfortunately, it is very difficult to quantify the amount of carbonate, silica, and brucite scales resulting from alkaline drilling mud. (These scales dissolve in HCI acid.)
- Lack of immediate clean up To minimize the possibility of secondary damage to the formation, the spent HF acid should be produced back out of the formation within a few hours of treatment completion.3 However, Grace well are equipped with gas lift systems which require 1 2 days to unload the casing.
ACID VOLUMES
In 1989 after the stimulation of Wells A 4 and A 30, it became apparent that inadequate acid volume was one cause for failures in early stimulation attempts. Most acid jobs are designed by rules of thumb, with the quantity of acid per foot of perforation determined from experience.
The acid volumes used in stimulating Monterey wells offshore California ranged widely, up to as much as 330 gal/ft. A theoretical penetration depth of 20 ft from the well bore can be achieved in 1 ft of 1 % porosity (typical Monterey fracture porosity) reservoir with 100 gal of acid. To treat suspected plugging beyond the immediate well bore, the innovative approach was to increase the acid volumes (from 70 to 130 gal/ft) and select the least damaged well, A 18. The optimum acid volumes for treating the more damaged wells (Fig. 1) have yet to be determined.
RETARDING, SEQUESTERING
Because of the abundance of calcareous materials in the Monterey at Grace, HCI acid tends to rapidly react with carbonates and spends near the well bore. As a result, the optimized acid program included acetic acid to retard the HCI acid and achieve deeper penetration.
Acetic acid reacts with calcareous materials to form calcium acetate, which acts as a buffer to HCI. The CO2 released by the reaction of HCI and acetic acid on carbonates, retards the rate of reaction of both the HCI and acetic acid.1 This allows HCI to remove flow restrictions deeper in the formation and to react more uniformly around the well bore.
Fig. 3 shows the comparison of the main geochemical constituents in the spent acid returns compared to those in the Monterey formation water. A two fold increase in total dissolved solids was primarily caused by the chloride, bicarbonate, and acetate. Citric acid was added as a sequestering agent to the HCI acid to chelate the dissolved calcium, magnesium, and iron ions and prevent redamaging the formation by reprecipitation.
Of particular concern is the redeposition of insoluble iron compound ferric hydroxide near the well bore. These deposits can cause permanent plugging.
Together with the acetic acid added for retardation, citric acid can act synergistically to keep metal ions in solution for a long period of time.1 This application is very crucial for Grace wells because the spent acid water from each treatment requires an average of 30 days for total recovery.
WETTABILITY RESTORATION
As a practical matter, it is desirable to leave the fracture surfaces water wet because making them oil wet would adversely affect relative permeability to oil. A mutual solvent in acid treatments can often improve both the frequency of success and well productivity.4 The mutual solvent ethylene glycol monobutyl ether, Egmbe, has appreciable solubility in both oil and water. The optimized treatment design uses Egmbe in the xylene preflush and the HCI acid stage to solubilize surfactant, thereby increasing surfactant effectiveness.
In the NH4Cl afterflush stage, Egmbe acts as a detergent capable of removing oil wetting materials from fracture surfaces and leaves the surfaces water wet.1
ACID PLACEMENT
To place the acid in the damaged zones, the majority of historical acidizing treatments for offshore California wells involved pumping acids down the production string in stages separated by chemical diverting agents. Some more recent jobs were foamed or nitrified.
Although the most effective placement control requires wish tools and straddle packers for mechanical diversion, chemical agents attempt proper acid distribution with the cost savings of bullheaded acid placement. These chemical diverters, unfortunately, have had little or no effect in altering the production profiles. This indicates that the acid was either ineffective in treating the damaged zones or not properly placed.
As shown by production logs from other Monterey fields, the significant production increases after acid treatment came from the zones that were already contributing.
The optimized stimulation method employed at Grace bullheads treating fluids in a single stage with no chemical diversion or nitrification. This method should be effective in distributing the acid as long as the pumping rates remain within the friction dominant region (Fig. 4.
For typical Grace wells, a minimum pump rate of 7 bbl/min is required, with the surface pumping pressure varying from well to well depending on depletion. For example, Wells A 8 and A18 required only 3,000 3,500 psi at 8.5 bbl/min, whereas Well A 17 required 3,500-4,300 psi at 8 bbl/min.
The maximum surface treating pressure traditionally is predicted, ignoring friction in tubing, from the frac gradient minus the hydrostatic gradient. According to conventional practices for avoiding fracturing the formation, the surface pumping pressure should be limited to pressure below this maximum treating pressure. The treating pressures employed at Grace wells, however, were much higher than the average maximum treating pressure, calculated at 2,000 2,500 psi.
This raises the question of whether Grace wells were stimulated by acid fracturing or by acidizing the existing fractures or both. Pump pressure did not indicate breakdown during acid treatment of these wells; on the contrary, the pump pressure kept increasing throughout the jobs.
FLUID DYNAMICS
Fluid injection during the course of well stimulation treatment is influenced to a large extent by the resistance to flow in the production tubing. The differential pressure curve (Fig. 4) is the result of using the Darcy-Weisbach formula to calculate pressure gradient inside Well A 18's production string (from wellhead to mid perforation) at different acid injection rates.
The pump pressure curve is obtained by subtracting the differential pressure from the static bottom hole pressure. The top curve depicts the pumping pressure for a well stimulated after initial completion at the original reservoir pressure. The bottom curve reflects pumping pressure at 1990 reservoir conditions. Note that the actual field data (stimulated May 20, 1990) are in good agreement with the prediction.
Fluid density, and friction to flow are the two major components that affect the pressure gradient of the fluid flow in vertical pipe. For injection, the hydrostatic head (density) is acting in the same direction as the flow, and the friction is in the opposite direction. As shown in Fig. 4, at zero pump rate the differential pressure is equal to the hydrostatic. Friction resistance develops as the pumping starts. Friction development is directly proportional to the pump rate.
In this density dominant region, the weight of the fluid is partially reduced by friction. For Well A 18, the friction becomes equivalent to the hydrostatic at 7.2 bbl/min pump rate, i.e., the pump pressure is the same as the bottom hole pressure. Beyond this pump rate, the friction resistance overwhelms the hydrostaticfriction dominant region.
Less pressure may be exerted on the formation face when treating the well in the friction dominant region because of the lack of potential energy from the hydrostatic head that was overcome by friction. This potential energy in the density dominant region is responsible for moving fluid down the tubing. The pressure at the pump, therefore, is equivalent to the pressure at the formation face.
Injection profile at the perforations is controlled by formation permeability and the differential pressure at the formation face. The treating fluid will move preferentially through zones of higher permeability and less damage. The quantity of fluid moving into these particular zones is controlled by the differential pressure (Poiseuille's Law).
Consequently, stimulating wells in the friction dominant region may result in a better injection coverage-less fluid is being taken by higher permeability zones because of lower differential pressure, at formation face, and more fluid is available for lower permeability zones, higher pump rate.
RESULTS
Since 1990, five Grace wells have been successfully stimulated using the acid treatment design discussed. The production response has been phenomenal, with an 85% increase in the first well (A 18) and a 100% increase in the last four wells (A 8, A 17, A 14, and A 28).
The first well, A 18, was treated at 8.5 bbl/min, about 70 times its production capacity, resulting in a 130 bo/d oil gain and 3 year job life. Because of mechanical problems, an attempt to evaluate acid effectiveness with production logs in Well A 18 was terminated. However, different gravity oils are being produced after the acid job and this may indicate contribution of new zones. Production logs were not run in the remaining four wells.
It should be noted that, of all the five wells stimulated, the best response came from the most damaged wells (Fig. 1). Wells A 17 and A 28 responded with 180 bo/d and 160 bo/d, respectively. The effect of proper acid treatment design is apparent in the successful stimulation of Well A 17. The previous treatment in June 1983 used almost identical chemicals but different acid coverage and pumping rates. That treatment was a technical and economic failure (Table 1).
Because of the economic and technical success of the redesigned acid stimulation, the treatment now has been performed in other offshore Monterey fields and in onshore fields in the San Joaquin Valley, Calif.
REFERENCES
- Allen, T.O., and Roberts, A.P., Production Operations Well Completions, Workover, and Stimulation, Oil & Gas Consultants International Inc., Vol. 2.
- Menon, V.B., and Wasan, D.T., "Characterization of Oil Water Interfaces Containing Finely Divided Solids with Applications to the Coalescence of Water in Oil Emulsion, A Review," Colloids Surfaces, Vol. 29, 1988.
- Williams, B.B., Gidlex., J.L., and Schechter, R.S., Acidizing Fundamentals, Henry L. Doherty Monograph Series, SPE, Vol. 6, 1979.
- Gidley, J.L., "Stimulation of Sand stone Formations with the Acid-Mutual Solvent Method," JPT, May, 1971.