Focused Fluid Flow Along Faults in the Monterey Formation, Coastal

Focused ¯uid ¯ow along faults in the Monterey Formation, coastal California

Peter Eichhubl* James R. Boles Department of Geological Sciences, University of California, Santa Barbara, California 93106, USA

ABSTRACT ding. The inferred ¯ow geometry illustrates namics and associated transfer of mass and the combined effect of fault permeability heat, it is of interest to quantify the extent to Fluid ¯ow in fractured siliceous mud- structure, permeability anisotropy of the which conductive faults focus basinal ¯uid stone of the Miocene Monterey Formation surrounding formation, and hydraulic head ¯ow. In a sedimentary sequence undergoing of California is inferred to be highly fo- distribution in controlling basinal ¯uid ¯ow dewatering and ¯uid expulsion during pro- cused toward map-scale faults that locally in faulted sequences. grade burial and diagenesis, the extent of ¯ow contain extensive amounts of carbonate and focusing by the presence of faults may be minor silica cement. The distance of cross- Keywords: faults, ¯uids, strontium, iso- quanti®ed by the ratio of fault-perpendicular stratigraphic ¯ow, as inferred based on the topes, hydrocarbon, Monterey Formation. ¯ow distance over the distance of fault-par- strontium isotopic composition of carbon- allel ¯ow in the source region. In strati®ed ate fault cement, is close to the thickness of INTRODUCTION sequences and for faults inclined with respect the Monterey Formation of 700 m in one of to the strati®cation, it may be more practical two study locations, Jalama Beach, and less Faults may affect basinal ¯uid ¯ow as bar- to consider the ratio between the distance of than the formation thickness at another lo- riers or as preferred conduits, depending on formation-parallel ¯uid ¯ow ⌬L over the dis- cation, Arroyo Burro Beach. Fluid is thus the difference in permeability between fault tance of cross-stratigraphic ¯ow ⌬H within derived from within the Monterey Forma- zone and the surrounding formation. Faults the source unit, tion rather than from underlying older that are preferred conduits due to increased units. Based on mass-balance estimates of permeability parallel to the fault are likely to ⌬L F ϭ (1) the ¯uid volume required for fault cemen- control the localized precipitation of pore ce- ⌬H tation at Jalama Beach, the minimum dis- ment and ore minerals and the migration and tance of formation-parallel ¯ow into the accumulation of hydrocarbons. Faults condu- (Fig. 1). Fault systems with a large focusing fault zone is 4 km and possibly Ͼ12 km. cive for ¯uid ¯ow can be subdivided into three ratio F will be more effective in channeling The inferred distance of ¯ow parallel to the hydraulic regimes (Fig. 1): (1) a source region ¯uid expulsion from the source rock to a po- formation into this fault thus exceeds the where pore ¯uid is drawn into the fault zone tential sink within the sequence or to the distance of cross-formational upward ¯ow from the surrounding formation of higher hy- Earth's surface. However, in petroleum sys- along the fault by at least a factor of six. draulic head; (2) a sink region of lower hy- tems faults with large F will also be more ef- The mass-balance estimate requires that draulic head where pore ¯uid is released back fective for leaking hydrocarbons out of ¯uid ¯ow along the fault is channeled into into the formation or onto the Earth's surface; breached reservoirs. a pipe-shaped conduit rather than distrib- and (3) an intermediate neutral conduit of no For any fault of given length, the ¯uid fo- uted along fault strike. Fluid ¯ow from the signi®cant ¯uid exchange with the surround- cusing factor F is a function of the perme- surrounding formation into fault pipes is ing formation that may separate the source ability anisotropy of the ¯uid source rock and inferred to follow a radial rather than uni- from the sink region. The effectiveness of ¯u- of the distance to adjacent faults (Fig. 1). or bilateral ¯ow symmetry, using bedding- id transport along a fault can be expressed as Large ratios of F would be expected in strat- con®ned sets of extension fractures and the total ¯uid volume that has migrated along i®ed source rocks with the highest permeabil- stratabound breccia bodies. Radial ¯uid the fault, the average ¯ow rate, the distance of ity parallel to bedding, transected by faults of ¯ow toward fault pipes requires fairly iso- fault-parallel ¯ow, and the distance perpendic- wide spacing. An isotropic permeability struc- tropic fracture permeability for ¯ow along ular to the fault over which ¯uid is focused ture of the formation and/or close fault spac- bedding and a low permeability across bed- toward the fault in the source region and dis- ing would lead to low ratios of F. The per- persed away from the fault in the sink region. meability anisotropy of the source rock may *Present address: Department of Geological and An additional parameter of interest is the be controlled by strati®cation within the for- Environmental Sciences, Stanford University, Stan- source volume, which is a function of ¯uid mation (intrinsic anisotropy) and by the for- ford, California 94305, USA, and Monterey Bay Aquarium Research Institute, Moss Landing, Cali- volume and source-rock porosity and mation boundaries to overlying and underly- fornia 95039, USA; e-mail: eichhubl@pangea. storativity. ing units (extrinsic anisotropy) (Fig. 1). stanford.edu. For the understanding of basinal hydrody- For a kilometer-scale fault system in the

GSA Bulletin; November 2000; v. 112; no. 11; p. 1667±1679; 11 ®gures; 2 tables.

1667 EICHHUBL and BOLES

Monterey Formation in southern California, we demonstrate that the distance of formation- parallel ¯uid ¯ow exceeds the distance of cross-stratigraphic ¯ow by a factor F Ն 6. The distance of cross-stratigraphic ¯uid ¯ow ⌬H will be estimated using the strontium isotopic composition of carbonate fault cement as a natural tracer. The formation-parallel ¯ow distance ⌬L is inferred based on mass-balance estimates of the ¯uid volume that has migrated along the fault system. On the basis of the estimated distances of ¯uid ¯ow and the characteristic permeability structure of Monterey rocks as observed in outcrop, we infer that the expulsion of basinal ¯uid in the Monterey Formation is highly focused toward faults by the intrinsic, layer-controlled permeability anisotropy of the formation (type A in Fig. 1).

STRUCTURAL SETTING OF Figure 1. Geometry of upward-directed basinal ¯uid expulsion as controlled by faults and CONDUCTIVE FAULTS by the permeability anisotropy of sedimentary sequences. Conductive faults can be subdivided into source and sink regions and an intermediate neutral ¯uid conduit of no or The Miocene Monterey Formation is an or- minimal ¯uid exchange with the surrounding formation. The ratio of formation-parallel ganic-rich, hemipelagic sequence of diatoma- over cross-formational ¯ow distance may be used to quantify the extent of ¯uid focusing ceous and phosphatic shale and organogenic toward faults in source units. Fluid focusing will be controlled by the permeability an- dolostone, deposited between 18 and 6 Ma in isotropy of the source unit, which can be intrinsic, i.e., controlled by bedding, or extrinsic, a series of continental-borderland basins (Pis- i.e., controlled by contacts to overlying and underlying formations of low permeability. ciotto and Garrison, 1981; Barron, 1986; Hor- Fluid ¯ow in the Monterey Formation is inferred to be strongly focused toward faults by na®us, 1994a). During burial, the organic-rich the intrinsic anisotropy of the formation (case A). diatomaceous sequence underwent a complex sequence of diagenetic alteration, dominated by the following reactions: (1) silica transfor- mation of diatom tests, from opal A via opal CT to quartz through two distinct steps of dissolution-reprecipitation (Stein and Kirkpa- trick, 1976; Isaacs 1981); (2) degradation of organic matter and maturation of hydrocarbons (Petersen and Hickey, 1987; Baskin and Peters, 1992); (3) precipitation of authigenic dolomite during early diagenesis and subse- quent dissolution and/or recrystallization (Burns and Baker, 1987; Compton, 1988; Ma- lone et al., 1994); and (4) transition of smectite to illite (Pollastro, 1990; Compton, 1991). With an initial organic matter content of up to 34% (Isaacs and Petersen, 1987), the Monte- rey Formation is both source and reservoir rock for hydrocarbons (Ogle et al., 1987; MacKinnon, 1989). Due to the low matrix permeability, typically Ͻ1 md, hydrocarbon migration and production critically depend on fractures as ¯ow pathways (Crain et al., 1985; Roehl and Weinbrandt, 1985; Isaacs and Pe- tersen, 1987; MacKinnon, 1989). Faults in the Monterey Formation are locally cemented by massive volumes of carbonate and quartz, indicative of large volumes of ¯uid ¯owing along these faults. At Jalama Beach, 100 km west of Santa Barbara (Fig. 2), two subparallel faults are cemented by dolo- Figure 2. Location map of Jalama and Arroyo Burro Beaches in south-central California.

1668 Geological Society of America Bulletin, November 2000 FOCUSED FLUID FLOW ALONG FAULTS mite and minor amounts of quartz. The southern fault forms a resistant wall of cemented fault breccia, ϳ6 m thick, 7 m tall, and ex- tending more than 35 m across the wavecut platform between mean low tide water line and the beach cliffs (Fig. 3). The second fault, ϳ300 m northwest of the ®rst one, is less well exposed in a cove. Both faults are characterized by a large amount of dolomite cement; 7±10-cm-thick layers of banded dolomite coat host rock fragments and aggregates of earlier dolomite cement (Fig. 3). Dolomite fault cement can be traced into extension veins that branch off from both faults and extend over distances of a few hundred meters (Fig. 4). The combined thickness of dolomite cement of both faults and connected veins, measured along a tra- verse across the strike of the fault-vein systems, is ϳ8 m. Fragments of cement and host rock ®ll interstitial cavities in fault cement. Sedimentary layering of in®ll and sorting of fragments by grain size suggest transport of detrital fracture ®ll in suspension by rapidly moving ¯uid along these faults (Eichhubl, 1997). The fault system at Jalama Beach is hosted by siliceous dolostone and dolomitic opal CT porcelanite and minor amounts of clay shale and chert. A fault zone similar to that of Jalama Beach is exposed in coastal cliffs at Arroyo Burro Beach in Santa Barbara (Eichhubl and Behl, 1998). Two subparallel faults and associated veins are cemented by banded calcite of cu- Figure 3. Dolomite-cemented, left-lateral strike-slip fault at Jalama Beach. The south- mulative cement thickness similar to that of eastern one of two exposed faults (Fig. 4) extends for ϳ35 m across the wavecut platform. Jalama Beach. The host rock sequence at Ar- A former slip surface forms the margin of the fault zone to the right (northwest), and a royo Burro Beach consists of dolomitic por- second parallel slip surface is exposed on the left margin, to the left of the measuring stick celanite and siliceous mudstone and only sub- (2 m long). ordinate dolostone. Coastal sections of the Monterey Formation underwent for the most part a single burial- exhumation cycle; exhumation started ca. 1 Ma (Jackson and Yeats, 1982), while offshore sections of the Monterey Formation are still parts of active basins undergoing burial. The inferred maximum burial depth for the Jalama Beach section is ϳ1000 m (Eichhubl, 1997). The exposed fault system at Jalama Beach is ϳ50 m below the top of the Monterey For- mation (Dibblee, 1988a, 1988b), which is ϳ700 m thick at this location (Grivetti, 1982). The fault system at Arroyo Burro Beach cuts across a bed of siliceous dolostone that marks the Luisian-Mohnian stage boundary (Echols, 1994; Horna®us, 1994c). On the basis of the measured thickness of the Naples Beach section (Echols, 1994; Horna®us, 1994c), located 22 km west of Arroyo Burro Beach, the fault system is ϳ300 m above the base of the Mon- Figure 4. Map of cemented faults and associated vein systems at Jalama Beach, ϳ1.6 km terey Formation, which has a total thickness northwest of state beach.

Geological Society of America Bulletin, November 2000 1669 EICHHUBL and BOLES

TABLE 1. 87Sr/86Sr COMPOSITION OF MONTEREY FAULT CEMENT fault. In all cases, the composition and purity Sample number Sample location Description 87Sr/86Sr Ϯ2␴ of carbonate cement were veri®ed by thin-sec- 03/01/95-7 Jalama Beach, southeastern fault Early, randomly oriented dolomite 0.708820 0.000014 tion petrography. zone cement Carbonate samples were dissolved in 5 M 03/01/95-6AB Jalama Beach, southeastern fault Late banded, acicular dolomite 0.708815 0.000019 acetic acid and the separated samples analyzed zone 04/23/96-6 Jalama Beach, eastern margin of Gray clear dolomite in fault-parallel 0.708874 0.000025 on a Finnigan MAT 261 adjustable ®ve-col- southeastern fault zone vein lector mass spectrometer with a precision of 02/26/95-2 Jalama Beach, vein 145 m west Early banded clear dolomite, vein 0.708837 0.000014 0.00003 (2 mean; Table 1). The accuracy of southeastern fault connected to fault (?) Ͻ ␴ 02/26/95-2 Jalama Beach, vein 145 m west Late acicular dolomite, vein con- 0.708753 0.000014 of the measurements was monitored with the of southeastern fault nected to fault (?) NBS 987 standard, which gave a mean value 02/11/95-5 Jalama Beach, veinlet 70 m east Isolated, bedding-con®ned dolo- 0.708829 0.000014 of southeastern fault mite veinlet of 0.710235 Ϯ 0.00002 (2␴ mean). Separa- 03/14/95-5A Arroyo Burro Beach, eastern fault Early cement band of large calcite 0.708765 0.000015 tions and analyses were performed in the lab- 600 m east State Beach botryoid within fault oratories of James Mattinson and George Til- 03/13/95-6 Arroyo Burro Beach, vein 20 m Calcite vein, connected to fault 0.708749 0.000020 east eastern fault ton at the University of California, Santa 02/24/95-3B(B) Arroyo Burro Beach, stratabound Laminated calcite vein reactivating 0.708774 0.000025 Barbara. vein 88 m west aluminum tube, earlier dolomicrite vein across In addition, four bulk rock samples of clay- 1100 m east State Beach dolostone marker bed 04/25/94-5 Arroyo Burro Beach, stratabound Baroque dolomite cement in 0.708683 0.000014 poor, siliceous dolostone host rock were ana- breccia 4 m east aluminum quartz-dolomite-calcite cement- lyzed to provide a strontium-stratigraphic ref- tube, 1100 m east State Beach ed dolostone breccia erence age of the section adjacent to the fault Note: All 87Sr/86Sr values reported relative to NBS-987 standard ϭ 0.710235. For detailed location description, at Jalama Beach. Carbonate was dissolved us- refer to Eichhubl and Behl (1998). Sample numbers are same as used for stable isotopic analyses reported in Eichhubl and Boles (1998) and Eichhubl (1997). ing 5 M acetic acid to minimize contamination by clay dissolution and the 87Sr/86Sr ratio de- termined from the leachate (Table 2). Remain- of ϳ410 m at this location. The inferred max- STRONTIUM ISOTOPIC ing carbonate in the dissolution residue was imum burial depth of the Arroyo Burro Beach COMPOSITION OF FAULT removed with 2.5 M hydrochloric acid until section is 750 m (Eichhubl and Boles, 1998). CARBONATE CEMENT AND HOST no visible reaction occurred. The acid was The faults and fault-related veins at Jalama DOLOSTONE subsequently replaced two times to minimize and Arroyo Burro Beach dip steeply and cut strontium contamination in exchangeable clay across outcrop-scale folds, indicating fault and The strontium isotopic compositions of car- layers, and the resulting HCl-insoluble residue vein formation during shortening and exhuma- bonate vein and fault cement from Jalama was analyzed for 87Sr/86Sr (Table 2). tion of the basin ¯anks. The late, postfolding Beach and Arroyo Burro Beach were analyzed Strontium concentrations of the bulk rock origin of these fault-related veins is also docu- in order to infer distances of mass transport. and leachate obtained by X-ray ¯uorescence mented through untilted strati®ed detrital frac- At Jalama Beach, microdrilled samples of spectrometry (XRF) and thermal ionization ture ®ll. Abundant hydrocarbon ¯uid inclusions banded, inclusion-poor dolomite cement from mass spectrometry (TIMS), respectively, in- and bitumen-cemented breccias indicate that the eastern fault and from veins 145 m west dicate that 5%±6% of total strontium of the fault-related carbonate vein cementation is con- and 70 m east of the eastern fault were ana- bulk dolostone host rock from Jalama Beach current with hydrocarbon migration. Striations lyzed. At Arroyo Burro Beach, calcite cement is radiogenic strontium contained in clays. and drag of bedding indicate left-lateral strike was sampled from the fault, from veins 20 and Following results by Schultz et al. (1989, Ta- slip for both fault systems. Displacements are 400 m to the east, and from dolomite ce- ble 2), we assume that the radiogenic stron- estimated to be Ͻ100 m in both cases. menting a dolostone breccia 500 m east of the tium resides in the nonexchangeable clay lay-

TABLE 2. ELEMENTAL AND STRONTIUM ISOTOPIC COMPOSITION OF SILICEOUS DOLOSTONE HOST ROCK AT JALAMA BEACH

87 86 ² 87 86 87 86 Sample Location SiO2 CaO MgO Al2O3 K2O Sr Rb HCl- Sr in residue Sr/ Sr Sr in carbonate Sr/ Sr Sr/ Sr number (wt%) (wt%) (wt%) (wt%) (wt%) (ppm) (ppm) insoluble of of of residue (ppm) (wt% of residue (ppm)* (wt% of* carbonate² bulk (wt%) total Sr)* total Sr)* rock*

04/23/96-3 88 m west of 28.34 38.97 26.38 3.19 0.43 345 14 20 84.8 5.0 0.71104 Ϯ 3 411 95.0 0.70895 Ϯ 2 0.70905 Ϯ 5 southeastern fault 01/22/98-1 180 m west of 51.86 26.35 13.68 1.97 0.39 300 13 39 43.6 5.7 0.71097 Ϯ 3 464 4.3 0.70891 Ϯ 1 0.70903 Ϯ 4 southeastern fault§ 04/23/96-8 80 m east of 41.89 31.83 18.70 2.93 0.55 244 19 35 44.3 6.3 0.71102 Ϯ 2 351 93.7 0.70899 Ϯ 2 0.70912 Ϯ 4 southeastern fault 01/22/98-3 110 m east of 31.94 36.09 24.71 3.58 0.48 241 14 24 58.0 5.8 0.71160 Ϯ 2 299 94.2 0.70886 Ϯ 2 0.70902 Ϯ 4 southeastern fault# Note: All 87Sr/86Sr values reported relative to NBS-987 standard ϭ 0.710235. *Calculated value. ²Acetate leachate. §Approximately 20 m stratigraphically below sample 04/23/96-3. #15 m stratigraphically below sample 04/23/96-8.

1670 Geological Society of America Bulletin, November 2000 FOCUSED FLUID FLOW ALONG FAULTS

ers that are largely unaffected by the acetic acid leach, whereas strontium in the exchangeable layers mimics the isotopic composition of the carbonate. This assumption is based on the similarity of our radiogenic strontium ratios of the dissolution residue with that of smectite in Schultz et al. (1989) after clay treatment with ammonium chloride, whereas the 87Sr/86Sr ratio of their untreated smectite samples com- pares to that of selectively leached carbonate pore cement. Negligible radiogenic strontium contamination during the selective acetic acid leach is also con®rmed by the generally good correspondence of the strontium stratigraphic position of selectively leached carbonate samples with their biostratigraphic or lithostratigraphic position, as demonstrated by our samples and by Miller (1995), and as discussed in the following. Strontium isotopic ratios for dolomite vein and fault cement at Jalama Beach, ranging between 0.70875 and 0.70887, are distinctly lower than dolomite (i.e., acetate leachate) Figure 5. Strontium isotopic composition of Monterey vein carbonate. Strontium values values, 0.70886 to 0.70899, of surrounding are plotted with respect to the biostratigraphic age of the host rock (Arroyo Burro Beach) host dolostone. Strontium isotopic ratios of and with respect to the strontium isotopic composition of the host dolostone on both sides calcite vein and fault cement at Arroyo Burro of the fault system (Jalama Beach). Horizontal bars on Jalama values represent the spread Beach are generally lower than vein and fault in strontium isotopic composition of the dolostone host rock. Seawater curve is after Ho- cement values from Jalama Beach, with values warth and McArthur (1997; revised second edition of database, written commun., August ranging between 0.70875 and 0.70877. The 1998). The 95% con®dence interval of seawater curve is within line thickness. Time scale 87 86 lowest Sr/ Sr ratio of 0.70868 is obtained is after Shackleton et al. (1994) for 0±7 Ma and Cande and Kent (1995) for 7±25 Ma. from a single analysis of dolomite vein ce- Vertical error bar approximates 2␴ standard deviation of vein cement analyses. All values .0.710235 ؍ ment at Arroyo Burro Beach, 500 m east of are normalized to NBS 987 the cemented fault. On the basis of crosscutting relations, this vein predates calcite fault and vein cement at Arroyo Burro Beach (Eich- ate cement precipitating during or after ¯uid burial but is inferred to progressively recrys- hubl and Boles, 1998). migration as compared to cement precipitated tallize during burial, based on its oxygen iso- prior to ¯uid migration. The strontium isoto- topic composition (Burns and Baker, 1987; DISTANCE OF CROSS- pic composition of cement may thus be used Malone et al., 1994). Miller's (1995) data in- STRATIGRAPHIC FLUID FLOW as a tracer for cross-stratigraphic ¯ow of con- dicate that, despite recrystallization, the do- nate pore ¯uids, provided that diagenetic lostone beds remain closed with respect to For carbonate cement precipitating from an changes in strontium isotopic ¯uid composi- strontium during recrystallization and oxygen aqueous solution, the carbonate cement inher- tion are negligible or predictable. isotopic resetting. its the strontium isotopic composition of the The strontium isotopic composition of the In contrast, the strontium isotopic values of aqueous solution, making the strontium iso- acetate-leached dolomite component in the calcite and dolomite vein and fault cement topic composition of carbonate cement a use- four host dolostone samples from Jalama from Jalama Beach and Arroyo Burro Beach ful natural tracer for basinal ¯uid ¯ow Beach agrees well with the upper Miocene are distinctly lower than the expected seawater (Schultz et al., 1989; Sullivan et al., 1990; stratigraphic position of the sampled section, strontium isotopic values for the particular Mountjoy et al., 1992; Chaudhuri and Clauer, ϳ50 m below the Monterey-Sisquoc contact stratigraphic position of these locations (Fig. 1993; Feldman et al., 1993). Throughout Neo- as mapped by Dibblee (1988a, 1988b). Ap- 5). For Jalama Beach, where a biostratigraphic gene and Quaternary time the seawater 87Sr/ parently, the selectively leached carbonate has reference is not available, 87Sr/86Sr ratios of 86Sr composition steadily increased to the pre- retained the primary strontium isotopic com- fault and vein dolomite are compared to the sent-day value of 0.709162 (Howarth and position of upper Miocene seawater. Similarly, measured 87Sr/86Sr composition of the sur- McArthur, 1997) (Fig. 5). The strontium iso- Miller (1995) found a good correlation be- rounding host dolostone. For Arroyo Burro topic composition of entrapped pore ¯uid tween strontium isotopic composition of ace- Beach, the 87Sr/86Sr composition of vein and within a Neogene sediment sequence is there- tic-acid leached dolostone and biostratigraphic fault cement is compared to the biostratigraph- fore expected to decrease steadily with depth, age of samples taken from a continuous Mon- ic age of the dolostone marker bed that cor- provided the strontium isotopic composition terey core from the South Elwood offshore oil responds to the Luisian-Mohnian stage bound- of the ¯uid remained unmodi®ed by diagenet- ®eld, 17 km west of Arroyo Burro Beach. Au- ary (Echols, 1994; Horna®us, 1994c) of ca. ic processes. Upward migration of pore water thigenic dolostone of the Monterey Formation 13.85 Ma (Barron, 1986; time scale revised would lead to lower 87Sr/86Sr ratios in carbon- has been shown to precipitate during early after Cande and Kent, 1995).

Geological Society of America Bulletin, November 2000 1671 EICHHUBL and BOLES

The average 87Sr/86Sr ratio of fault and vein affect the strontium isotopic composition of some stratigraphic lower source layer with the dolomite at Jalama Beach of 0.70882 is pore ¯uid on a regional scale. Carbonate, distance of strontium transport is tantamount 0.00011 lower than the mean 87Sr/86Sr value however, is abundant as authigenic dolomite to assuming that all the strontium is derived of the surrounding host dolostone (0.70893). and biogenic calcite (Isaacs, 1980; Burns and from a single layer rather than from a depth For a formation thickness at Jalama Beach of Baker, 1987; Compton, 1988). range, and no exchange of strontium has taken ϳ700 m and assuming uniform sedimentation Due to their higher abundance, clays have place during upward ¯ow. These assumptions rate throughout the section, the difference sug- a potentially larger in¯uence on the strontium are justi®ed if a distinct stratigraphic horizon gests upward transport of strontium by ϳ245 isotopic composition of Monterey Formation serves as an aquifer supplying ¯uid to the m(Ϯ100 m for 87Sr/86Sr ratios Ϯ0.00002 2␴). water than feldspar. If we take (lacking a more fault zone, and provided that carbonate dis- The inferred distance of upward strontium systematic data set) the four host dolostone solution or precipitation during upward ¯ow transport at Arroyo Burro Beach is less and samples from Jalama Beach as a guide, com- does not cause signi®cant exchange of stron- close to the resolution limit of the method. plete dissolution of clay contained in these do- tium with the country rock. Because no single Fault-related calcite cement at Arroyo Burro lostones would raise the pore-water 87Sr/86Sr distinct aquifer can be identi®ed in the sec- Beach has an average 87Sr/86Sr composition of ratio by ϳ0.00013 as compared to pore ¯uid tions at Jalama Beach and Arroyo Burro 87 86 0.708757 Ϯ 0.000020 that is 0.000031 Ϯ whose Sr/ Sr is controlled by carbonate dis- Beach, we assume that ¯uid is drawn into the 0.000023 below the expected value of solution only. Addition of radiogenic stron- fault uniformly from the underlying section 0.708788 Ϯ 0.000003 for this particular strati- tium would lead to an apparent decrease in over a ®nite depth interval and with uniform graphic position. For a thickness of the un- inferred ¯uid migration distance. Assuming strontium concentration. Assuming a uniform derlying Monterey section of ϳ300 m, this complete clay dissolution, addition of radio- strontium concentration of ¯uid entering the genic strontium would thus double the migra- difference in isotopic composition is equiva- fault will be better justi®ed by the uniformity tion distance of 245 m for Jalama Beach and lent to 70 m (Ϯ50 m) of upward strontium of the dolostone-rich section at Jalama Beach quadruple the distance of 70 m for Arroyo transport. At both Jalama and Arroyo Burro (Grivetti, 1982) than for the more variable Burro Beach. While the inferred ¯uid migra- Beach, the strontium isotopic composition of lithologic composition of the Arroyo Burro tion distances would be signi®cantly longer, vein cement points toward a ¯uid source be- Beach section. both estimates would still be within the thick- low the surrounding host rock but within the Assuming uniform ¯uid in¯ux with uni- ness of the Monterey Formation at both lo- formation rather than within deeper sections cations, leaving the basic conclusion of an in- form strontium concentration over the depth of the basin. traformational ¯uid source as derived earlier interval transected by the fault, the strontium The interpretation of these numbers as dis- unchanged. Dissolution of smectite, the pre- isotopic composition of the ¯uid and thus of tance of upward ¯uid ¯ow is contingent upon dominant clay phase in the Monterey Forma- the cement at the top of the fault is the inte- the following three assumptions. tion, has been observed at burial temperatures grated strontium isotopic composition over the above 80 ЊC only (Pollastro, 1990; Compton, depth interval. If dissolved strontium enters 1. The Strontium Isotopic Composition of 1991). Thus, strontium release into the ¯uid the fault and fracture system over the depth the Pore Fluid has not Signi®cantly Been due to clay dissolution is unlikely to be sig- interval htop±hbottom at equal amounts per unit Changed by Dissolution of Detrital ni®cant except within the deepest parts of the depth interval ⌬h, then the strontium isotopic Minerals formation. The lack of a signi®cant radiogenic ratio of the ¯uid at the top of the depth inter- strontium component in Monterey Formation val is Mineral dissolution may change the stron- water is also suggested by the strontium iso- 87/86 tium isotopic ¯uid composition. Dissolution topic composition of present-day Monterey Srfluid of clays and potassium feldspar tends to in- Formation water. Analyzed formation waters top 87 86 1 crease the Sr/ Sr ratio, whereas dissolution from three different Monterey reservoirs in the 87/86 ϭ Srformation dh. (2) of plagioclase tends to decrease the ratio of Santa Barbara and Santa Maria Basins, in- htopϪ h bottom ͵ bottom 87Sr/86Sr. The extent to which these reactions cluding the South Elwood offshore oil ®eld alter the pore ¯uid strontium isotopic com- studied by Miller (1995) with respect to do- In equation 2, 87/86Sr stands for the strontium position depends on the water-rock ratio, the lomite strontium isotopes, range between isotopic ratio and htop and hbottom correspond to strontium concentration in the mineral phase, 0.70863 and 0.70826 (Eichhubl, 1997), and the depth at top and bottom of the fault sys- and the mineral abundance. Potassium feld- are even lower than the strontium isotopic tem, respectively. The isotopic composition of spar and plagioclase may contain 400±700 range predicted for connate Monterey For- ¯uid entering the fault zone can be assumed ppm Sr2ϩ (Schultz et al., 1989; Feldman et al., mation water based on the seawater curve. to correspond to the strontium isotopic ratio 1993), somewhat higher than Sr2ϩ concentra- Lacking any indication of a signi®cant ra- of the formation at this particular depth h. diogenic strontium component in Monterey 87/86 tions in dolomite of 300±460 ppm measured With Sr¯uid inferred through the measured at Jalama Beach (Table 2) and similar values Formation ¯uids, the following discussion on isotopic composition of the vein cement and ¯uid migration distance is based on the ®rst 87/86 reported previously (Faure, 1986; Burns et al., d Srformation/dh approximated through the 1988). Potassium feldspar and calcic plagio- estimates of 245 m for Jalama Beach and 70 seawater strontium isotopic curve in Figure 5, clase form only a minor detrital component in m for Arroyo Burro Beach. equation 2 can be solved numerically to obtain Monterey rocks, of 2% and 0.5% , re- h , the depth at the base of the depth in- ϳ weight 2. The Strontium Composition of the bottom spectively (averages of 41 analyses; Compton, Fluid Corresponds to the Strontium terval drained by the fault system. The me- 1991). Volcanic ash beds, which occasionally Composition of a Distinct Source Layer dium distance of mass transport thus obtained contain Sr2ϩ in excess of 1000 ppm (Horna- is 535 m (Ϯ120 m, Ϯ2␴) for Jalama Beach ®us, 1994b), make up only ϳ1% of the total Equating the difference in strontium isoto- and 140 m (Ϯ80 m, Ϯ2␴) for Arroyo Burro Monterey section and are therefore unlikely to pic composition between the vein cement and Beach. For Jalama Beach in particular, and po-

1672 Geological Society of America Bulletin, November 2000 FOCUSED FLUID FLOW ALONG FAULTS tentially for Arroyo Burro Beach, these distances are signi®cant fractions of the thick- nesses of the underlying formation, which is ϳ650 m at Jalama Beach (Fig. 6) and ϳ300 m at Arroyo Burro Beach. However, the inferred distances still do not suggest that strontium mass transport originates signi®cantly below the base of the Monterey Formation. 87/86 Inferring d Srformation/dh from the seawater strontium isotopic curve requires knowledge of variations in sedimentation rate within the particular stratigraphic interval. The transport distance obtained for Arroyo Burro Beach is corrected for sedimentation rates given by Echols (1994) for Naples Beach. For Jalama Beach, where a detailed biostratigraphic analysis is not available, we assumed a uniform sedimentation rate throughout the formation thickness. Because 535 m (Ϯ120 m) consti- Figure 6. Inferred distances of cross-stratigraphic mass transport, based on the 87Sr/86Sr tutes a large fraction of the total stratigraphic ratio of vein cement, plotted against the stratigraphic section at Jalama Beach. The lower thickness of 700 m at Jalama Beach, varia- estimate assumes a single source bed of ¯uid, and the higher estimate assumes that ¯uid tions in sedimentation rate throughout the sec- is drawn into the fault uniformly over the depth interval. In both cases, a uniform sedi- tion are likely to cancel out when integrated mentation rate throughout the section is assumed, allowing substitution of the strontium over the depth range. isotopic sequence of the section by the 87Sr/86Sr seawater curve. If part of the strontium is continuously removed from solution by precipitation during upward ¯ow in the fracture system, then the mates of mass transport distance, dissolution position re¯ecting the seawater strontium iso- ®nal strontium ¯uid composition in the frac- of host carbonate along the conduit wall will topic evolution during deposition will provide ture system can be estimated through the ex- have an equally small effect on the strontium a proxy only for ¯uid ¯ow across bedding, but pression isotopic composition of the ¯uid. Fluid would not for bedding-parallel ¯ow such as ¯ow up- have to enter and react with the host rock for dip along the tilted ¯anks of the basin. The 87/86Sr fluid (3) tens to hundreds of meters adjacent to the ¯uid ¯ow distance parallel to bedding is estimated conduits for any signi®cant shift in ¯uid stron- in the following section based on a mass-bal- top 87/86Sr (1 Ϫ␾h) dh tium composition to occur. Host rock alter- ance calculation that estimates the ¯uid vol- ͵ formation ation visible in thin section, such as dolomite ume required for carbonate precipitation and bottom ϭ top , recrystallization, is restricted to within 1 m by comparing the estimated ¯uid volume to a ͵ (1 Ϫ␾h) dh adjacent to fault-related veins at Jalama likely source volume. bottom Beach. If dolomite recrystallization and ac- companying strontium isotopic resetting of the VOLUME OF FLUID EXPELLED AND where ␾ is the fraction of strontium removed host dolostone would have been more perva- THE DISTANCE OF FORMATION- from solution through cement precipitation sive, the composition of the host rock samples PARALLEL FLUID FLOW per unit ¯ow distance. on which the transport distance estimates are The mass-balance calculation in the following based would have been equally affected. Sig- The large volumes of carbonate cement section suggests that ϳ1.9 ϫ 10±4 cm3 dolomite ni®cant exchange in strontium isotopes be- contained in the fault zones at Jalama Beach per liter ¯uid per meter of ¯ow distance precip- tween ¯uid and host rock during upward ¯ow and Arroyo Burro Beach are indicative of itates during upward ¯uid ¯ow. With an average can thus be excluded. large ¯uid volumes and potentially effective concentration of 300 ppm Sr2ϩ in dolomite, ce- focusing of ¯uid ¯ow along these faults. In mentation removes ϳ1.6 ϫ 10±4 mg/L strontium 3. The Distance of Strontium Transport conjunction with the cross-stratigraphic ¯ow per meter ¯ow distance. Given an average stron- Corresponds to the Distance of Fluid distance derived herein, an estimate of the ¯u- tium concentration of ϳ12 mg/L in Monterey Transport id volume necessary for precipitation of the Formation water (Eichhubl, 1997), the factor ␾ observed volume of carbonate cement may becomes ϳ1.3 ϫ 10Ϫ5 mϪ1. With the factor ␾ The distances calculated herein refer to the provide an estimate of the distance of bed- that small, the ¯uid strontium isotopic ratio ob- transport of strontium, corresponding to dis- ding-parallel ¯ow and thus the extent of ¯uid tained through equation 3 differs from that ob- tances of ¯uid ¯ow only if strontium exchange focusing. The following discussion focuses on tained through equation 2 by Ͻ1 ϫ 10±6, which between solution and country rock or earlier dolomite cementation at Jalama Beach but ap- is within the precision of the strontium isotopic vein cement during upward ¯ow is negligible. plies similarly to calcite cementation at Ar- analysesÐon average Ϯ2 ϫ 10±5 (2␴)Ðand The calculation suggests that this is the case, royo Burro Beach. therefore within the uncertainties of the distance the distances thus representing true distances A mass-balance calculation of ¯uid involved estimates given herein. of ¯uid ¯ow. In sedimentary sequences, how- in precipitation of the observed volume of do- As little as precipitation affects the esti- ever, stratigraphic differences in 87Sr/86Sr com- lomite requires information about the concentra-

Geological Society of America Bulletin, November 2000 1673 EICHHUBL and BOLES

formation water composition may therefore be considered the ®nal ¯uid composition after rapid upward ¯ow rather than the initial composition before ¯ow.

Carbonate-Limited Model

A more viable approach to carbonate mass balancing is based on the assumption that the formation ¯uid is always saturated in Ca2ϩ and/or Mg2ϩ with respect to calcite or dolomite and that precipitation is controlled by

changes in CO2 partial pressure as ¯uid moves upward and decompresses. Initial saturation with respect to dolomite or calcite appears reasonable for ¯uid that has been in contact with dolostone and calcite shell tests for several 2± 2ϩ million years. In a pure H2O-CO3 -Me system, carbonate cementation is controlled by

changes in temperature and CO2 partial pres- Figure 7. The variation of CO2aq concentration in an aqueous solution as a function of sure, with the tendency for carbonate disso- depth and mole fraction m of CO2gas.ACO2gas mole fraction of 0.07 is measured in gas lution with decreasing temperature under con- produced from the Hondo oil ®eld (Exxon, Thousand Oaks, 1997, oral commun.), whereas stantPCO2 , but for precipitation with a mole fraction of 0.01 is more characteristic for the temperature range of 80±100 ؇C decreasingPCO2 under constant temperature -Smith and Ehrenberg, 1989). Curves are calculated for thermobaric gradients of 50 ؇C/ (Ellis, 1959). Under common geobaric gradi) -atm (hydrostatic), 50 ؇C/240 atm (lithostatic), and 50 ؇C/130 atm, which is an upper ents, the pressure effect outweighs the tem 100 limit of characteristic gradients in the Monterey Formation. The mole fraction of CO2aq perature effect and carbonate precipitates dur- and using Henry's Law coef®cients ing upward ¯ow, to the saturation 1 ؍ is calculated assuming a fugacity coef®cient m after Ellis and Golding (1963) for 1 M NaCl solutions. concentration of available divalent metal cations, primarily Ca2ϩ and Mg2ϩ.

The CO2 concentration in solution can be tions, or rather activities, of Ca2ϩ,Mg2ϩ, and carbonate is available in excess. Concentra- calculated using Henry's Law 2Ϫ 2ϩ 2ϩ CO3 , and of pH, pressure, and temperature pri- tions of Ca and Mg in present-day Mon- or to and after upward ¯uid ¯ow. Because these terey Formation water in the Santa Barbara fCO K ϭ 2, (4) parameters are not known to the desirable extent, and Santa Maria basins average ϳ60 mg/L h x order of magnitude estimates need to be ob- and 20 mg/L, respectively (Eichhubl, 1997). tained by limiting the number of variables. The If all the magnesium were consumed to form where Kh is Henry's Law coef®cient and x is

®rst estimate of ¯uid mass will be derived by dolomite, these concentrations correspond to a mole fraction of CO2aq in solution. Under di- assuming that cementation is controlled by the maximum of 153 mg or 0.05 cm3 of dolomite lute conditions, the fugacity,fCO2 , approaches amount of Ca2ϩ and Mg2ϩ in solution. The sec- per liter solution. The entire fault system, hav- the partial pressurePCO2 , which is equal to the ond model assumes that Ca2ϩ and Mg2ϩ are not ing a cumulative cement thickness of 8 m and total pressure times the mole fraction of limiting factors, but that precipitation is con- an estimated depth, based on the strontium CO2gas. The change in CO2aq concentration trolled by changes in CO2 partial pressure, other transport, of 500 m, would thus require 8 ϫ with depth is shown in Figure 7 for a constant 7 3 factors affecting pH being constant. In addition, 10 m ¯uid per meter fault length. Assuming mole fraction CO2gas of 0.07, which is the reactions between a solid and an aqueous solu- that the cemented zone extends 35 m along mole fraction of CO2gas in the gas phase pro- tion may not reach equilibrium if the solution is strike of the fault, a total volume of 2.8 ϫ 109 duced from the Monterey Formation in the moving rapidly. A mass-balance calculation m3 or 2.8 km3 ¯uid is required for dolomite offshore Hondo oil ®eld 40 km west of Santa based on the assumption of thermodynamic precipitation. Barbara (Exxon, Thousand Oaks, 1997, oral equilibrium may thus provide only a minimum Because only a fraction of Mg2ϩ is likely to commun.). Henry's Law coef®cients are taken estimate of the mass of ¯uid involved. A third be removed from solution during upward ¯ow, from Ellis and Golding (1963) for an ionic estimate is therefore obtained through observed this estimate is unrealistic. The pumping of for- strength of 1 m and gas activity coef®cients precipitation rates in a moving solution under mation water to the Earth's surface during oil have been approximated by unity. similar hydrodynamic conditions. production and before sampling may resemble Equilibrium concentrations of dissolved

upward ¯uid ¯ow in fault zones. During pump- CO2aq are calculated in Figure 7 for three ther- Cation-Limited Model ing, the formation water decompresses, thus de- mobaric gradients, hydrostatic (50 ЊC/100

gasses CO2, and possibly precipitates carbonate atm), lithostatic (50 ЊC/240 atm), and in be- A conservative estimate can be obtained as scale in wells (Boles and Ramseyer, 1987). tween (50 ЊC/130 atm), the latter correspond- based on the assumption that dolomite precip- Ca2ϩ and Mg2ϩ concentrations are therefore ing to a pressure gradient of 130 atm/km or itation is limited by the concentration of Ca2ϩ likely to be higher in the subsurface than their 0.6 psi/ft (12.7 MPa/km). Pore-¯uid pressure and Mg2ϩ in solution only, and that dissolved values analyzed at the surface. The analyzed gradients in Monterey hydrocarbon reservoirs

1674 Geological Society of America Bulletin, November 2000 FOCUSED FLUID FLOW ALONG FAULTS range between hydrostatic and 0.6 psi/ft (U.S. id pressure conditions close to the hydrostatic concentrations of coastal Monterey Formation Geological Survey, 1983; Crain et al., 1985). gradient unlikely. Evidence for lithostatic ¯uid waters are generally below this threshold, with Upward ¯ow of 500 m to the depth of interest, pressures, however, is not observed. Extensive an average concentration of 2500 mg/L (S.G. ϳ1000 m, and following a hydrostatic pres- bedding-parallel veins that would indicate lith- Franks, 1996, written commun.). Temporary sure gradient, leads to a drop in CO2aq con- ostatic ¯uid pressures are not found in the ¯uctuations in organic acid concentration above centration of ϳ30 mg/L, corresponding to Monterey Formation, with the exception of this threshold may lead to stages of carbonate ϳ0.02 cm3/L of dolomite precipitation. If the some water sills inferred to form during early dissolution intervening precipitation of vein ce- ¯uid were initially at a ¯uid pressure corre- burial based on crosscutting relations with folds ment. Stages of dissolution in fault carbonate ce- sponding to a pressure gradient of 130 atm/ and faults. On the basis of these constraints and ment were observed through cathodoluminesc- km and were to drop to a hydrostatic pressure, present-day ¯uid pressure gradients, an initial ence at Jalama and Arroyo Burro Beach then CO2aq concentration would drop by ϳ130 ¯uid pressure corresponding to an intermediate (Eichhubl, 1997). In any case, due to the pres- 3 mg/L, precipitating ϳ0.10 cm /L dolomite. ¯uid pressure gradient of 130 atm/km may be ence of organic acid in Monterey Formation wa- With the latter number providing a mini- considered a reasonable estimate of the ¯uid ter, the ¯uid volume estimates and the inferred mum ¯uid volume, the entire fault system pressure before upward ¯uid ¯ow. ¯ow distances calculated here for systems free with a cumulative vein thickness of ϳ8 m and In addition to ¯uid pressure, ¯uid volume of organic acid must be considered minimum an estimated depth of 500 m requires a min- estimates depend critically on the CO partial 2 estimates. imum of 4 ϫ 107 m3 of ¯uid per meter fault pressure, which is a function of the mole frac- Two further assumptions are the pressure length. Assuming this ¯uid volume corre- tion CO .ForaCO mole fraction of 0.01, 2gas 2gas pro®le over the vertical extent of the fault sponds to 5% of the source rock volume, the a more common value for hydrocarbon res- zone, and the ratio of expelled ¯uid to source fault would drain a rock volume of at least 0.8 ervoirs in the 80±100 ЊC temperature range km3 per meter fault length. Because the faults than 0.07 used herein (Smith and Ehrenberg, rock volume. The preceding calculation as- at Jalama Beach are exposed only over a dis- 1989), the ¯uid volume becomes 10 km3, and sumes that the pressure and temperature drop tance of ϳ35 m along strike, these numbers the radius of radial drainage becomes 11 km. is gradually distributed over 500 m. This assumption agrees with pore-pressure pro®les can be interpreted in two ways. If the ce- The CO2gas mole fraction typically increases in mented zone is only 35 m long, the required hydrocarbon reservoirs with increasing burial with depth for the Monterey Formation (U.S. ¯uid volume is 1.2 ϫ 109 m3, corresponding temperature and depth (Smith and Ehrenberg, Geological Survey, 1983) that suggest a grad- 3 ual pressure increase with depth. In addition, to a source volume of ϳ23 km and a distance 1989). Curves of dissolved CO2gas plotted of radial ¯uid ¯ow into the fault of 4.3 km. If against burial depth would therefore have a it is expected that, after initial rupture of a the fault cementation were continuous along lower slope than the curves plotted for con- seal, the largest gradient in pore ¯uid pressure strike, focusing ¯uid bilaterally over a 500-m- stant mole fraction CO2gas, as shown in Figure is presumably where the ¯uid leaves the for- thick depth interval, then the fault would have 7. During rapid upward ¯ow along fractures mation and enters the fracture network. Fol- to drain an 800-km-wide zone to each side of in a formation of low matrix permeability, ¯u- lowing the assumption made earlier, that ¯uid the fault, which is wider than the basin. Clear- id may not exchange CO2 easily with the for- seeps into the fault zone over its entire depth ly, cementation along the fault must be dis- mation water at shallower depth. The ¯uid range, the ¯uid pressure drop would be equal- continuous along strike and the fault focused may therefore be expected to follow the ly distributed over the estimated ¯ow distance ¯uid radially. closed-system trajectories shown in Figure 7. thickness of the cemented zone, which is as- To see how sensitive the ¯uid volume es- Upward ¯uid ¯ow through the matrix of per- sumed to be 500 m. timate is to the assumed pressure gradient, we meable formations, however, may follow tra- It is dif®cult to assess the ratio of expelled may assume that upward ¯uid ¯ow followed jectories for a decreasing mole fraction CO2gas ¯uid to the volume of the source rock. This a hydrostatic pressure gradient throughout. In and correspondingly precipitate more carbon- ratio estimate is necessary to relate the cal- this case, the total ¯uid volume required to ate cement per unit ¯uid volume and ¯ow dis- culated ¯uid volume to the ¯ow distance. The 3 precipitate an 8 ϫ 35 ϫ 500 m dolomite- tance than would be expected for rapid ¯ow expelled ¯uid volume may correspond to a cemented zone is 6.3 109 m3, four times along faults. ϳ ϫ change in pore volume during compaction. more than obtained for the superhydrostatic These calculations were obtained assuming, Because of the abundance of hydrocarbon in- case herein. The drainage radius becomes 9 for the sake of convenience, a pure H O-CO 2±- 2 3 clusions in Jalama Beach vein cement, ¯uid km, twice the value, all other parameters being Me2ϩ system. Pore water of the Monterey For- ¯ow along the fault is likely to be the conse- held constant. mation, however, is characterized by high con- quence of hydrocarbon maturation in the ad- A ¯uid pressure gradient close to hydrostatic centrations of organic acid (S.G. Franks, 1996, is not very likely during ¯uid expulsion at Ja- written commun.). The dissociation of organic jacent synclines of the basin. Ungerer et al. ↔ (1983) estimated a 57% volume gain for type lama Beach, however, because ¯uid migration acids, such as the reaction CH3COOH ± ϩ II kerogen during hydrocarbon maturation, al- with respect to the formation under hydrostatic CH3COO ϩ H for the acetate-acetic acid pair, lowing for selective escape of CO2 and H2S. conditions is possible only when the section may control pH in addition toPCO2 . Solutions compacts through the stagnant water column with acetate concentrations below ϳ0.06 M For an initial average organic matter content during burial. One of us (Eichhubl, 1997) de- (ϳ3600 mg/L) at 100 ЊC would still precipitate of 8% of the rock volume (Isaacs and Peter- sen, 1984), the possible volume gain amounts rived ¯ow velocities of as much as 0.1 m/s carbonate through CO2 exsolution, although to a based on sedimentary structures of detrital frac- lesser extent than in a solution whose pH is sole- to ϳ5% of the initial source rock volume. Be- ture ®ll. These high ¯ow velocities and the in- ly controlled by the carbonate system (Lunde- cause organic matter will be lost prior to the ferred timing of ¯uid ¯ow during basin margin gard and Land, 1989). At higher acetate concen- activity of the fault system, 5% yield may be exhumation require above-hydrostatic ¯uid trations, loss of CO2 leads to carbonate considered a maximum estimate, also leading pressures to induce ¯uid ¯ow, thus making ¯u- dissolution rather than precipitation. Acetate to a minimum estimate in ¯ow distance.

Geological Society of America Bulletin, November 2000 1675 EICHHUBL and BOLES

Kinetic Model

The preceding approach is based on the assumption that ¯uid is in thermodynamic equilibrium with wall-rock carbonate prior to and during ¯ow. While this assumption is justi®able as an initial condition with a long residence time prior to ¯uid expulsion, sluggish reaction kinetics may prevent the solution from reaching equilibrium during upward ¯ow. Long-term average ¯ow velocities for Jalama Beach are ϳ1 m/day, based on inferred rates of heat loss during upward ¯ow (Eichhubl, 1987). Experimental kinetic data for dolomite precipitation at the temperatures of interest are not available. A crude estimate of the ¯uid volume required for precipitation of dolomite cement at Jalama Beach can be obtained, how- Figure 8. Fault-fracture mesh in alternating sequence of opal CT porcelanite and organic- ever, from the observed rate of calcite scale rich, phosphatic opal A mudstone at Arroyo Burro Beach. Porcelanite fails brittlely by formation in well casings. Boles and Ramsey- forming extension fractures. Extension is accommodated in overlying and underlying mud- er (1987) reported calcite scale with a cross- stone by normal faults. sectional area of 21 cm2 being precipitated from 89.1 ϫ 103 m3 of ¯uid over 3342 days (ϳ9 yr) of hydrocarbon production at a depth neled along beds of high matrix permeability of 2591 m in a Miocene sandstone reservoir. that alternate with beds of low permeability, Precipitation is accompanied by a pressure as in an alternating sequence of sandstone and drop from 4900 psi (33.8 MPa) to 400 psi (2.8 shale. Rock types of the Monterey Formation MPa) and degassing of the ¯uid. Mole percent are characterized by equally low matrix per-

CO2gas averaged 1.7% over this period, in- meability (Crain et al., 1985; Roehl and Wein- creasing from ϳ0.13% to 2.16% (data by brandt, 1985; Isaacs and Petersen, 1987; Arco, Bakers®eld, provided to J.R. Boles, 1993). MacKinnon, 1989), with the exception of Under the same conditions, 280 m2 of cal- some dolostone beds that contain appreciable cite, equivalent to the surface area of dolomite secondary porosity we observed in core from at Jalama Beach, would require 12 ϫ 109 m3 the offshore Hondo oil ®eld. Dolostone from or 12 km3 of ¯uid, which is one order of mag- Jalama and Arroyo Burro Beach, however, ap- nitude higher than the best equilibrium esti- pears to be tightly cemented throughout. mate given herein. The corresponding radius 2. Fluid ¯ow may follow partings parallel to of ¯uid drainage is 12 km. Although the re- bedding planes. Although siliceous rock types of action kinetics may be slower for dolomite, the Monterey Formation typically part parallel Figure 9. Schematic view of fault-fracture this estimate is based, in contrast to the equi- to bedding in outcrop, it is not certain if bed- mesh, modi®ed after Eichhubl and Behl librium thermodynamic calculation here, on ding-parallel fractures are conductive for ¯uid (1998) and inspired by Sibson (1996) and actual precipitation rates in a moving ¯uid. ¯ow in the subsurface. Mineralization along Gross (1995). The average calculated ¯ow velocity in the bedding planes is infrequent and, where cross- well casing is 0.3 L/s or ϳ20 cm/s. cutting relations are observed, appears to predate folding (Bartlett, 1994). tem of normal faults and extension fractures DISCUSSION 3. Fluid ¯ow may follow bedding-con®ned was termed fault-fracture mesh by Sibson fractures. Extension fractures oriented at high (1996), following Hill (1977). Fluid ¯ow is The estimated distances of ¯uid ¯ow, to angles to bedding and con®ned to the more likely to occur preferentially along the exten- 650 m across bedding and a minimum of 4± competent units of porcelanite and dolostone sion fractures, parallel to the bedding. Slip 12 km parallel to the formation, is equivalent in sequences alternating with clay shale of low along the normal faults would lead to contin- to ¯uid focusing by a factor F Ն 6 as de®ned competence are characteristic of the Monterey ued opening of the extension veins, possibly in equation 1. This high degree of lateral ¯uid Formation (Narr, 1991; Hickman and Dun- preventing complete sealing of extension focusing indicates that the permeability of the ham, 1992; Gross, 1995; Gross et al., 1998). veins by cementation. The permeability and formation away from faults is highly aniso- Hydrocarbon ®ll and carbonate and silica ce- connectivity of bedding-con®ned fractures tropic, the highest permeability being parallel mentation of these bedding-con®ned veins in- may also be enhanced due to shear reactiva- to bedding. dicate that they serve as ¯uid pathways in the tion of fractures initially formed in extension Fluid ¯ow may preferentially follow bed- subsurface. Frequently, extension fractures in (Dholakia et al., 1998). ding due to the following types of permeabil- the more competent beds lead into normal 4. Potential pathways for preferred ¯uid ¯ow ity anisotropy. faults in the overlying and underlying mud- parallel to bedding are bedding-parallel breccia 1. Bedding-parallel ¯uid ¯ow may be chan- stone beds (Figs. 8 and 9). The connected sys- zones (Redwine, 1981; Roehl, 1981; Bel®eld et

1676 Geological Society of America Bulletin, November 2000 FOCUSED FLUID FLOW ALONG FAULTS

The mass-balance estimates require that ¯uid enters the faults in a radial rather than uni- or bilateral symmetry. It follows that faults are preferred pathways over short segments, forming pipes rather than planar ducts within these fault zones (Fig. 11). Possible pathways for ¯uid ¯ow are releasing bends, extensional stepovers of fault segments, and fault intersections. Radial ¯ow into these ¯uid pathways requires that ¯uid ¯ow along bedding is fairly isotropic within the bedding plane. A single dominant joint set would likely result in a strong ¯ow anisotropy within bedding planes (Fig. 11B). Radial and thus isotropic ¯ow requires at least two or more crosscutting and mutually connected joint sets (Fig. 11C), as described by Gross (1993) from the Monterey Formation. Isotropic ¯ow along bedding may be best accommodated by bedding-parallel breccia zones. Bel®eld et al. (1983) and MacKinnon (1989) reported that oil production is most proli®c from wells intersecting the regional, north-northeast±south-southwest±trend- ing joint set at high angles, a ®nding not supported by others (Jon Schwalbach, Exxon, Thousand Oaks, 1996, oral commun.) and in- consistent with the radial ¯ow geometry inferred from this study. This ambivalence may indicate a regional variability in the permeability anisotropy for ¯ow parallel to bedding. Within the tilted sequence on the basin ¯anks, ¯uid will be drawn preferentially from the center of the basin or from adjacent synforms. This semicircle or arc of a circle geometry will extend the minimum distances of formation-parallel ¯ow of 4 to Ͼ12 km by at least a factor of 2. An updip migration path of hydrocarbons from synforms undergoing burial and catagenesis into Figure 10. Brittlely folded chert bed forming bedding-parallel breccia body. Hydrocarbon uplifted structural highs was also inferred by matrix of chert breccia indicates that breccia serves as ¯uid migration pathway. Photo- Ogle et al. (1987) based on the charge history graph courtesy of Richard Behl, California State University, Long Beach. of Monterey reservoirs. The effect of faults on basinal ¯uid migration is thus governed by three geometric constraints: al., 1983). Dholakia et al. (1998) documented folds. Brecciated chert layers are frequently (1) the permeability structure of the fault zone the formation of bedding-parallel brecciation in found saturated with hydrocarbons. such as the presence of zones of local extension the Monterey Formation due to differential slip The observed control of bedding on the along the fault; (2) the permeability anisotropy between bedding planes, presumably caused by fracture permeability structure suggests that of the surrounding formation as controlled by ¯exural slip in folds. Prominent breccia bodies the inferred permeability anisotropy of the for- strati®cation, jointing, and lateral stratigraphic in excess of 1 m thickness are also observed as mation and resulting ¯uid focusing are intrin- variations; and (3) the orientation of the regional the consequence of brittle folding of chert layers sic to the formation (Fig. 1A) rather than con- ¯ow ®eld as controlled by differences in hy- (Snyder et al., 1983; Hickman and Dunham, trolled by formation boundaries (Fig. 1C). draulic head. An analysis of all three compo- 1992; Behl and Garrison, 1994) (Fig. 10). Typ- However, the degree to which the permeability nents is required for the prediction of basinal ically, diatomite and porcelanite layers overlying anisotropy is intrinsic may vary on a regional ¯uid ¯ow in faulted sequences, as done routinely and underlying folded chert beds are unfolded, scale and with depth. Porcelanite and chert in by numerical modeling. Although the combina- giving chert folds the resemblance of synsedi- the quartz diagenetic stage may be suf®ciently tion of low matrix and high fracture permeability mentary folds. Behl and Garrison (1994) sug- fractured to allow ¯ow across bedding. The is unique to the siliceous shale of the Monterey gested that brittle chert folds are tectonic, with permeability anisotropy of the formation may Formation, bedding or layering, lateral strati- the shortening accommodated in the overlying then be controlled by diagenetic boundaries graphic variations, and the preferred orientation and underlying diatomite and porcelanite layers and by variations in silica and clay content of of regional joint sets will create a permeability by bulk shortening and porosity collapse, where- the formation with depth on the scale of strati- anisotropy in most sedimentary sequences as as low-porosity chert buckles, brecciates, and graphic members of the formation. well as in basement rocks, thus affecting local-

Geological Society of America Bulletin, November 2000 1677 EICHHUBL and BOLES

detailed suggestions for clari®cation. This project was funded in part by National Science Foundation grant EAR-93-04581 and a grant from the Geolog- ical Society of America.

REFERENCES CITED

Barron, J.A., 1986, Updated diatom biostratigraphy for the Monterey Formation of California, in Casey, R.E., and Barron, J.A., eds., Siliceous microfossil and micro- plankton studies of the Monterey Formation and modern analogs: Paci®c Section, Society of Economic Pa- leontologists and Mineralogists Special Publication 45, p. 105±119. Bartlett, W.L., 1994, Field guide to the Monterey Formation at Ellwood Beach, Santa Barbara County, California, in Horna®us, S.J., ed., Field guide to the Monterey Formation between Santa Barbara and Gaviota, Cali- Figure 11. Control of fault structure and source bed permeability anisotropy on basinal fornia: Paci®c Section, American Association of Pe- ¯uid-¯ow geometry in fractured reservoirs. Fluid is depicted to ¯ow updip within frac- troleum Geologists Guide Book, v. 72, p. 95±106. tured source unit into conductive fault and leaves the fault into a fractured reservoir unit Baskin, D.K., and Peters, K.E., 1992, Early generation characteristics of a sulfur-rich Monterey kerogen: at a structurally higher level. A: Fault is uniformly conductive along strike; single fracture American Association of Petroleum Geologists Bul- set directs ¯uid ¯ow in single direction. Fault is highly effective in draining the source letin, v. 76, p. 1±13. Behl, R.J., and Garrison, R.E., 1994, The origin of chert in bed. B: Same as A, but fault is conductive only within stepover region. Fault is poorly the Monterey Formation of California (USA), in Iiji- effective in draining the source bed. C: Fault is conductive only within stepover region, ma, A., ed., Proceedings of the 29th International but two sets of fractures in source and reservoir units allow radial ¯ow pattern into and Geological Congress, Part C, p. 101±132. Bel®eld, W.C., Helwig, J., La Pointe, P., and Dahleen, W.K., out of the fault ¯uid conduit. Fault is highly effective in draining source bed. Fluid ¯ow 1983, South Ellwood Oil Field, Santa Barbara Chan- in Monterey faults is inferred to follow case C. Radial ¯ow in the Monterey Formation nel, California, a Monterey Formation fractured res- may be accommodated by multiple fracture sets or by bedding-parallel brecciation. ervoir, in Isaacs, C.M., and Garrison, R.E., eds., Pe- troleum generation and occurrence in the Miocene Monterey Formation, California: Los Angeles, Paci®c Section, Society of Economic Paleontologists and ization of cementation and ore mineralization in leasing bends or stepover regions of overlap- Mineralogists, p. 213±221. any crustal environment. ping fault segments. Outcrop observations in- Boles, J.R., and Ramseyer, K., 1987, Diagenetic carbonate in Miocene sandstone reservoir, San Joaquin Basin, dicate that likely conduits for ¯uid ¯ow along California: American Association of Petroleum Geol- CONCLUSION bedding are layer-con®ned sets of extension ogists Bulletin, v. 71, p. 1475±1487. Burns, S.J., and Baker, P.A., 1987, A geochemical study of fractures and bedding-parallel breccia bodies, dolomite in the Monterey Formation, California: Jour- Reservoir-scale faults in the Monterey For- leading to a permeability anisotropy that is in- nal of Sedimentary Petrology, v. 57, p. 128±139. mation contain locally large volumes of car- trinsic to the formation rather than controlled Burns, S.J., Baker, P.A., and Showers, W.J., 1988, The factors controlling the formation and chemistry of dolo- bonate and minor quartz cement, indicative of by formation boundaries. Fluid ¯ow parallel mite in organic-rich sediments: Miocene Drakes Bay large volumes of ¯uid migrating along these to bedding into the fault is inferred to follow Formation, California, in Shukla, V., and Baker, P.A., faults. Using the strontium isotopic composi- a radial rather than uni- or bilateral ¯ow sym- eds., Sedimentology and geochemistry of dolostones: Society of Economic Paleontologists and Mineralo- tion of carbonate cement as a natural tracer metry. The inferred radial ¯ow geometry to- gists Special Publication 43, p. 41±52. for ¯uid ¯ow, the inferred distance of upward ward fault channels or pipes requires that the Cande, S.C., and Kent, D.V., 1995, Revised calibration of cross-stratigraphic ¯ow approaches the thick- the geomagnetic polarity timescale for the Late Cre- permeability structure for ¯ow parallel to bed- taceous and Cenozoic: Journal of Geophysical Re- ness of the Monterey Formation at Jalama ding is fairly isotropic, whereas ¯ow across search, v. 100, p. 6093±6095. Beach of 700 m and appears to constitute a bedding is restricted. The isotropic permeabil- Chaudhuri, S., and Clauer, N., 1993, Strontium isotopic compositions and potassium and rubidium contents of signi®cant fraction of the underlying part of ity of bedding planes is likely to be caused by formation waters in sedimentary basins: Clues to the the Monterey Formation at Arroyo Burro multiple sets of extension fractures at high an- origin of the solutes: Geochimica et Cosmochimica Beach. At both locations, the inferred distanc- Acta, v. 57, p. 429±437. gles to bedding and by bedding-parallel brec- Compton, J.S., 1988, Sediment composition and precipita- es suggest that ¯uid is derived from within the ciation. The inferred ¯ow geometry illustrates tion of dolomite and pyrite in the Neogene Monterey Monterey Formation rather than from under- the combined effect of fault permeability and Sisquoc Formations, Santa Maria Basin area, Cal- lying older units. Mass-balance estimates of ifornia, in Shukla, V., and Baker, P.A., eds., Sedimen- structure, permeability anisotropy of the sur- tology and geochemistry of dolostones: Society of the volume of ¯uid required for fault cemen- rounding formation, and hydraulic head dis- Economic Paleontologists and Mineralogists Special tation at Jalama Beach indicate a minimum tribution in controlling basinal ¯uid ¯ow in Publication 43, p. 53±64. Compton, J.S., 1991, Origin and diagenesis of clay minerals in distance of formation-parallel ¯ow into the faulted sequences. the Monterey Formation, Santa Maria Basin area, Cali- fault zone of 4 to Ͼ12 km. Fluid ¯ow is thus fornia: Clay and Clay Minerals, v. 39, p. 449±466. highly focused toward this fault zone, the in- Crain, W.E., Mero, W.E., and Patterson, D., 1985, Geology of ACKNOWLEDGMENTS the Point Arguello discovery: American Association of ferred distance of ¯ow parallel to the forma- Petroleum Geologists Bulletin, v. 69, p. 537±545. tion into the fault exceeding the distance of This study bene®ted from discussions with Jim DePaolo, D.J., and Finger, K.L., 1991, High-resolution cross-formational upward ¯ow along this fault Mattinson, George Tilton, Dave Parkinson, Julie strontium-isotope stratigraphy and biostratigraphy of Bryce, Rick Behl, Jon Schwalbach, Rick Sibson, the Miocene Monterey Formation, central California: by at least a factor of six. The mass-balance Geological Society of America Bulletin, v. 103, p. Atilla Aydin, and Rob Knipe. An earlier version of estimates also suggest that ¯uid ¯ow along the 112±124. this manuscript was reviewed by Art Sylvester. Dholakia, S.K., Aydin, A., Pollard, D.D., and Zoback, fault is localized to pipe-shaped channels of GSA Associate Editor Elizabeth Burton and review- M.D., 1998, Development of fault-controlled hydro- limited extent along fault strike, such as re- ers Alden Carpenter and Isabel Montanez provided carbon migration pathways in the Monterey Forma-

1678 Geological Society of America Bulletin, November 2000 FOCUSED FLUID FLOW ALONG FAULTS

tion, California: American Association of Petroleum Santa Barbara and Gaviota, California: Paci®c Section Paleontologists and Mineralogists Special Publication. Geologists Bulletin, v. 82, p. 1551±1574. American Association of Petroleum Geologists Guide 15, p. 97±122. Dibblee, T.W., Jr., 1988a, Geologic map of the Lompoc Book, v. 72, p. 1±15. Pollastro, R.M., 1990, Geothermometry from smectite and Hills and Point Conception Quadrangles, Santa Bar- Horna®us, S.J., 1994b, Correlation of volcanic ashes in the silica diagenesis in the diatomaceous Monterey and bara County, California: Santa Barbara, California, Monterey Formation between Naples Beach and Gav- Sisquoc Formations, Santa Maria Basin, California: Dibblee Foundation, scale 1:24 000. iota Beach, California, in Horna®us, S.J., ed., Field American Association of Petroleum Geologists Bul- Dibblee, T.W., Jr., 1988b, Geologic map of the Tranquillon guide to the Monterey Formation between Santa Bar- letin, v. 74, p. 742. Mtn. and Point Arguello Quadrangles, Santa Barbara bara and Gaviota, California: Paci®c Section, Ameri- Redwine, L.E., 1981, Hypothesis combining dilation, nat- County, California: Santa Barbara, California, Dibblee can Association of Petroleum Geologists Guide Book, ural hydraulic fracturing, and dolomitization to ex- Foundation, scale 1:24 000. v. 72, p. 45±58. plain petroleum reservoirs in Monterey Shale, Santa Echols, R.J., 1994, Biostratigraphy of the Luisian and Mohnian Horna®us, S.J., 1994c, Field trip log to the Monterey For- Maria area, California, in Garrison, R.E., and Douglas, Stages between Santa Barbara and Gaviota, California, in mation between Santa Barbara and Gaviota, Califor- R.G., eds., The Monterey Formation and related sili- Horna®us, S.J., ed., Field guide to the Monterey Forma- nia, in Horna®us, S.J., ed., Field guide to the Mon- ceous rocks of California: Paci®c Section, Society of tion between Santa Barbara and Gaviota, California: Pa- terey Formation between Santa Barbara and Gaviota, Economic Paleontologists and Mineralogists Special ci®c Section, American Association of Petroleum Geol- California: Paci®c Section, American Association of Publication 15, p. 221±248. ogists Guide Book, v. 72, p. 29±44. Petroleum Geologists Guide Book, v. 72, p. 107±123. Roehl, P.O., 1981, Dilation brecciationÐA proposed mecha- Eichhubl, P., 1997, Scale, rates, and timing of fracture-re- Howarth, R.J., and McArthur, J.M., 1997, Statistics for nism of fracturing, petroleum expulsion and dolomitiza- lated ¯uid ¯ow in the Miocene Monterey Formation, strontium isotope stratigraphy: A robust LOWESS ®t tion in the Monterey Formation, California, in Garrison, Coastal California [Ph.D. dissert.]: Santa Barbara, to the marine Sr-isotope curve for 0 to 206 Ma, with R.E., and Douglas, R.G., eds., The Monterey Formation University of California, 298 p. look-up table for derivation of numeric age: Journal and related siliceous rocks of California: Paci®c Section, Eichhubl, P., and Behl, R.J., 1998, Field guide to ``Diagen- of Geology, v. 105, p. 441±456. Society of Economic Paleontologists and Mineralogists esis, deformation, and ¯uid ¯ow in the Miocene Mon- Isaacs, C.M., 1980, Diagenesis in the Monterey Formation Special Publication 15, p. 285±315. terey Formation'': Ventura±Santa Barbara±Jalama examined laterally along the coast near Santa Barbara, Roehl, P.O., and Weinbrandt, R.M., 1985, Geology and pro- Beach±Grefco Quarry/Lompoc, in Eichhubl, P., ed., California [Ph.D. dissert.]: Stanford, California, Stan- duction characteristics of fractured reservoirs in the Diagenesis, deformation, and ¯uid ¯ow in the Mio- ford University, 329 p. Miocene Monterey Formation, West Cat Canyon Oil- cene Monterey Formation: Paci®c Section, SEPM (So- Isaacs, C.M., 1981, Outline of diagenesis in the Monterey ®eld, Santa Maria Valley, California, in Roehl, P.O., ciety for Sedimentary Geology) Special Publication Formation examined laterally along the Santa Barbara and Choquette, P.W., eds., Carbonate petroleum res- 83, p. 85±98. coast, California, in Isaacs, C.M., ed., Guide to the ervoirs: New York, Springer, p. 526±545. Eichhubl, P., and Boles, J.R., 1998, Vein formation in re- Monterey Formation in the California coastal area, Schultz, J.L., Boles, J.R., and Tilton, G.R., 1989, Tracking lation to burial diagenesis in the Miocene Monterey Ventura to San Luis Obispo: Paci®c Section, American calcium in the San Joaquin basin, California: A stron- Formation, Arroyo Burro Beach, Santa Barbara, Cal- Association of Petroleum Geologists, v. 15, p. 25±38. tium isotopic study of carbonate cements at North ifornia, in Eichhubl, P., ed., Diagenesis, deformation, Isaacs, C.M., and Petersen, N.F., 1987, Petroleum in the Mio- Coles Levee: Geochimica et Cosmochimica Acta, v. and ¯uid ¯ow in the Miocene Monterey Formation: cene Monterey Formation, California, in Hein, J.R., ed., 53, p. 1991±1999. Paci®c Section, SEPM (Society for Sedimentary Ge- Siliceous sedimentary rock-hosted ores and petroleum: Shackleton, N.G., Crowhurst, S., Hagelburg, T., Pisias, N., ology) Special Publication 83, p. 15±36. New York, Van Nostrand Reinhold, p. 83±116. and Schneider, D., 1994, A new late Neogene time Ellis, A.J., 1959, The solubility of calcite in carbon dioxide Jackson, P.A., and Yeats, R.S., 1982, Structural evolution scale: Application to Leg 138 Sites, in Proceedings of solutions: American Journal of Science, v. 257, p. of Carpinteria Basin, Western Transverse Ranges, Cal- the Ocean Drilling Program, Scienti®c results, Volume 354±365. ifornia: American Association of Petroleum Geolo- 138: College Station, Texas, Ocean Drilling Program, Ellis, A.J., and Golding, R.M., 1963, The solubility of car- gists Bulletin, v. 66, p. 805±829. p. 73±101. bon dioxide above 100ЊC in water and in sodium chlo- Lundegard, P.D., and Land, L.S., 1989, Carbonate equilibria Sibson, R.H., 1996, Structural permeability of ¯uid-driven ride solutions: American Journal of Science, v. 261, and pH buffering by organic acidsÐResponse to fault-fracture meshes: Journal of Structural Geology,

p. 47±60. changes inpCO2 : Chemical Geology, v. 74, p. 277±287. v. 18, p. 1031±1042. Faure, G., 1986, Principles of isotope geology (second edi- MacKinnon, T.C., 1989, Petroleum geology of the Monte- Smith, J.T., and Ehrenberg, S.N., 1989, Correlation of car- tion): New York, Wiley, 589 p. rey Formation in the Santa Maria and Santa Barbara bon dioxide abundance with temperature in clastic hy- Feldman, M.D., Kwon, S.-T., Boles, J.R., and Tilton, G.R., coastal and offshore areas, in MacKinnon, T.C., ed., drocarbon reservoirs: Relationship to inorganic chem- 1993, Diagenetic mass transport in the southern San Oil in the California Monterey Formation: American ical equilibrium: Marine and Petroleum Geology, v. 6, Joaquin basin, California, U.S.A.: Implications from Geophysical Union, Field Trip Guide Book T311, p. p. 129±135. the strontium isotopic composition of modern pore ¯u- 11±27. Snyder, W.S., Brueckner, H.K., and Schweickert, R.A., ids: Chemical Geology, v. 110, p. 329±343. Malone, M.J., Baker, P.A., and Burns, S.J., 1994, Recrys- 1983, Deformational styles in the Monterey Formation Grivetti, M.C., 1982, Aspects of stratigraphy, diagenesis, tallization of dolomite: Evidence from the Monterey and other siliceous sedimentary rocks, in Isaacs, C.M., and deformation in the Monterey Formation near San- Formation (Miocene), California: Sedimentology, v. and Garrison, R.E., eds., Petroleum generation and oc- ta Maria±Lompoc, California [Master's thesis]: Santa 41, p. 1223±1239. currence in the Miocene Monterey Formation, Cali- Barbara, University of California, 155 p. Miller, N.R., 1995, Lithostratigraphic signal evaluation of fornia: Los Angeles, Paci®c Section, Society of Eco- Gross, M.R., 1993, The origin and spacing of cross-joints: the Miocene Monterey Formation, South Elwood nomic Paleontologists and Mineralogists, p. 151±170. Examples from the Monterey Formation, Santa Bar- Field, Santa Barbara±Ventura Basin, California [Ph.D. Stein, C.L., and Kirkpatrick, R.J., 1976, Experimental por- bara coastline, California: Journal of Structural Geol- dissert.]: Dallas, Texas, University of Dallas, 606 p. celanite recrystallization kinetics: A nucleation and ogy, v. 15, p. 737±751. Mountjoy, E.W., Qing, H., and McNutt, R.H., 1992, Stron- growth model: Journal of Sedimentary Petrology, v. Gross, M.R., 1995, Fracture partitioning: Failure mode as tium isotopic composition of Devonian dolomites, 46, p. 430±435. a function of lithology in the Monterey Formation of Western Canada sedimentary basin: Signi®cance of Sullivan, M.D., Haszeldine, R.S., and Fallick, A.E., 1990, coastal California: Geological Society of America sources of dolomitizing ¯uids: Applied Geochemistry, Linear coupling of carbon and strontium isotopes in Bulletin, v. 107, p. 779±792. v. 7, p. 59±75. Rotliegend Sandstone, North Sea: Evidence for cross- Gross, M.R., GutieÂrrez-Alonso, G., and Bartlett, W.L., Narr, W., 1991, Fracture density in the deep subsurface: formational ¯uid ¯ow: Geology, v. 18, p. 1215±1218. 1998, Fold-related fractures in coastal outcrops of the Techniques with application to Point Arguello Oil Ungerer, P., Behar, E., and Discamps, D., 1983, Tentative Monterey Formation: Effects on structural style, me- Field: American Association of Petroleum Geologists calculation of the overall volume expansion of organic chanical stratigraphy, and scale at Arroyo Burro Bulletin, v. 75, p. 1300±1323. matter during hydrocarbon genesis from geochemistry Beach, in Eichhubl, P., ed., Diagenesis, deformation, Ogle, B.A., Wallis, W.S., Heck, R.G., and Edwards, E.B., data. Implications for primary migration: Proceedings and ¯uid ¯ow in the Miocene Monterey Formation: 1987, Petroleum geology of the Monterey Formation of the International Meeting on Organic Geochemistry Paci®c Section, SEPM (Society for Sedimentary Ge- in the offshore Santa Maria/Santa Barbara areas, in (Advances in Organic Geochemistry 1981), v. 10, p. ology) Special Publication 83, p. 37±65. Ingersoll, R.V., and Ernst, W.G., eds., Cenozoic basin 129±135. Hickman, R.G., and Dunham, J.B., 1992, Controls on the development of coastal California: Englewood Cliffs, U.S. Geological Survey, 1983, Geology and reservoir char- development of fractured reservoirs in the Monterey New Jersey, Prentice-Hall, p. 382±406. acteristics of the Santa Ynez Unit, Santa Barbara Formation of central California, in Larsen, R.M., et Petersen, N.F., and Hickey, P.J., 1987, California Plio-Mio- Channel, off California, in Isaacs, C.M., and Garrison, al., eds., Structural and tectonic modeling and its ap- cene oils; evidence of early generation, in Meyer, R.F., C.M., eds., Petroleum generation and occurrence in plication to petroleum geology: Amsterdam, Elsevier, ed., Exploration for heavy crude oil and natural bitu- the Miocene Monterey Formation, California: Los An- p. 343±353. men: American Association of Petroleum Geologists geles, Paci®c Section, Society of Economic Paleon- tologists and Mineralogists, p. 111±130. Hill, D.P., 1977, A model for earthquake swarms: Journal Studies in Geology, v. 25, p. 351±359. of Geophysical Research, v. 82, p. 347±352. Pisciotto, K.A., and Garrison, R.E., 1981, Lithofacies and MANUSCRIPT RECEIVED BY THE SOCIETY AUGUST 20, 1998 Horna®us, S.J., 1994a, Overview of the stratigraphy of the depositional environments of the Monterey Formation, REVISED MANUSCRIPT RECEIVED JULY 22, 1999 Monterey Formation along the coastline between San- California, in Garrison, R.E., and Douglas, R.G., eds., MANUSCRIPT ACCEPTED NOVEMBER 18, 1999 ta Barbara and Gaviota, California, in Horna®us, S.J., The Monterey Formation and related siliceous rocks ed., Field guide to the Monterey Formation between of California: Paci®c Section, Society of Economic Printed in the USA

Geological Society of America Bulletin, November 2000 1679