thesis draft 2016-02-01 final formatted · this thesis has been accepted on behalf of the...
TRANSCRIPT
1
2
Copyright
By
David Yongxian Kong
2016
3
Establishing the base of underground sources of
drinking water (USDW) using geophysical and chemical
reports in the southern San Joaquin Basin, CA
By
David Yongxian Kong, B.S.
A Thesis Submitted to the Department of Geological Sciences
California State University, Bakersfield
In Partial Fulfillment of the Degree of Masters of Science in Geology
Winter 2016
Establishing the base of underground sources of drinking water (USDW)
using geophysical and chemical reports in the southern San Joaquin
Basin, CA
By David Yongxian Kong, B.S.
This thesis has been accepted on behalf of the Department of Geology
by their supervisory committee:
Dr. ice Gillespi Co mittee Chair
D:fig~ Dr. Dirk Baron
iii
5
Acknowledgements
I would like to thank my committee chair, Dr. Janice Gillespie, for her guidance through the entirety
of this research. Her expertise in well logs and the oil fields of the San Joaquin Basin helped me
greatly. She has helped me understand petroleum geology and oilfields through the courses she
taught and also as an advisor to the Imperial Barrel Award competition.
I appreciate my committee members, Dr. Dirk Baron (CSU Bakersfield) and Dr. David Shimabukuro
(CSU Sacramento), for their feedback and help in refining my thesis.
Thank you to the Department of Conservation, Oil and Gas Division for their support in helping
provide funding over the course of this research.
Thank you to everyone I have met and worked with during my years at CSUB because each and
every one of you helped me get to where I am. I have made some great friends in the geology
department.
And finally, an enormous thank you to my parents and sister for supporting me through my pursuit
of graduate studies.
6
Abstract
Recent concerns about well stimulation and oilfield disposal practices has resulted in the desire to
learn more about the distribution of usable groundwater that might be impacted by these practices.
Waters that require protection are classified by the United States Environmental Protection Agency
as USDWs (Underground Sources of Drinking Water, defined by EPA regulations at 40 CFR 144.3).
These waters have a concentration of 10,000 parts‐per‐million (ppm) total dissolved solids (TDS) and
are not within an exempt aquifer. Direct sampling and chemical analyses of the water from oil and
gas producing formations provide the most accurate values for the formation water salinities, but
the data is scarce. There were 1,017 chemical reports recorded for the southern San Joaquin Basin.
The method in this analysis uses open‐hole geophysical logs and Archie’s equation to calculate the
salinity. The two methods used in the analysis are the spontaneous potential (SP) method that uses
the spontaneous potential log and the mud and formation resistivity values to calculate a salinity,
and the resistivity‐porosity (RP) method that uses the resistivity and porosity logs. The SP method
was performed on 110 wells, and the RP method was performed on 51 wells. The low number of RP
method calculations was due to the lack of porosity logs in the interested interval. The RP method
was chosen to calculate the USDW due to its lower error (0.275) compared to the SP method (0.704)
and better correlation between the tested and calculated salinities (0.807 for the RP method
compared to 0.467 for the SP method). The USDW was calculated or 182 wells in the southern San
Joaquin Basin. A deeper USDW exists on the eastern margin (> 5,000ft below sea level) near the
Sierra Nevada Mountains that shallows westward toward the Coast Ranges where the USDW
boundary is near surface. The shallowing of the USDW is more gradual along the Bakersfield Arch,
although the USDW remains deep well into the Tejon Sub‐basin.
7
Table of Contents Table of Contents .................................................................................................................................. vi
List of Tables ........................................................................................................................................ vii
List of Figures ...................................................................................................................................... viii
Introduction ........................................................................................................................................... 1
Geologic Setting .................................................................................................................................. 3
Previous Work ........................................................................................................................................ 6
Methodology .......................................................................................................................................... 6
Spontaneous Potential (SP) Method .................................................................................................. 7
Resistivity‐Porosity (RP) Method ...................................................................................................... 17
Salinity Calculations .......................................................................................................................... 25
Depth to USDW ................................................................................................................................ 31
Resistivity Analysis ............................................................................................................................ 33
Results .................................................................................................................................................. 33
Sensitivity Analysis for RP Calculations .............................................................................................. 44
Porosity ............................................................................................................................................. 44
Resistivity .......................................................................................................................................... 47
Archie’s Constants ............................................................................................................................ 47
Conclusion ............................................................................................................................................ 47
Appendix I ............................................................................................................................................ 50
Appendix II ........................................................................................................................................... 51
Appendix III .......................................................................................................................................... 54
Appendix IV .......................................................................................................................................... 56
References ............................................................................................................................................ 61
8
List of Tables
Table 1 Equations for the spontaneous potential method from Asquith, 1982 Table 2 The range of cementation (m) and tortuosity (a) values for the Archie Equation
for various reservoir types Table 3 Density values for matrix materials and fluids in various reservoirs Table 4 Interval transit times for different reservoirs used for calculating porosity from
sonic logs
9
List of Figures
Figure 1 Location map showing the study area in the southern San Joaquin Basin, California
Figure 2 Stratigraphic column for the San Joaquin Valley, California Figure 3 Depiction of the spontaneous potential response to different drilling muds and
reservoir saturations Figure 4 Well log header showing the mud resistivity values and bottom‐hole
temperature (BHT) Figure 5 Spontaneous potential and resistivity log examples from well 74X‐35 in the
Bellevue field Figure 6 A spontaneous potential and resistivity log showing how to identify an oil‐
water contact Figure 7 A plot showing the correction factor to correct pseudo‐spontaneous potential
to static spontaneous potential Figure 8 This chart shows the relationship between mud resistivity values with various
mud weights Figure 9 This is the nomograph to convert resistivity values at different temperatures
and relates the values to total dissolved solids Figure 10 An example of a density‐neutron porosity log Figure 11 An example of the bulk density log Figure 12 Map showing the wells with available chemical reports, porosity logs, and well
calculations Figure 13 Well log showing the determination of depth to base of USDW from the
resistivity log Figure 14 Graph showing the calculated salinity values plotted against the tested salinity
values for the spontaneous potential method Figure 15 Graph showing the error from the spontaneous potential method Figure 16 Graph showing the calculated vs tested salinity values for the resistivity‐
porosity method Figure 17 Graph showing the error for the resistivity‐porosity method Figure 18 Graph showing the deep resistivity plotted against the tested salinity in
sampled zones Figure 19 Graph showing the deep resistivity plotted against the tested salinity in
adjacent 100% water‐wet zones Figure 20 Contour map showing the depth to USDW in the southern San Joaquin Basin,
California Figure 21 Graphs showing the effects on the error of the RP method by varying
resistivity, porosity, cementation (m), and tortuosity (a) Figure 22 General porosity values listed for different field and various reservoirs
10
Introduction
It is important to determine the characteristics (depth, size, salinity, etc.) of underground water
aquifers in order to safeguard the resources needed for the future of the California economy.
Protection of the water resources in California is an important issue due to the state’s large
population and prosperous agricultural industry and also the propensity of the state to fall into
multi‐year and, historically, multi‐decade droughts. In 2012, California produced nearly one‐third of
the nation’s vegetables and over two‐thirds of the fruits and nuts (USDA, www.nass.usda.gov/ca,
9/18/2014). The San Joaquin Valley produces most of the state’s agricultural bounty and is also the
source of over 70% of California’s oil production. Major oil production in the southern San Joaquin
basin began in earnest in 1899 with the discovery of oil near the west bank of the Kern River. After
more than 100 years of production, many of the area’s oil wells are in decline and the produced
water to oil ratio is increasing. This saline produced water is usually disposed of by injecting it back
into underground formations. There is concern that these saline brines could negatively impact
potentially useable groundwater, termed underground sources of drinking water (USDW).
The United States Environmental Protection Agency (EPA) defines a USDW as an aquifer or its
portion:
1. Which supplies any public water system; or
2. Which contains a sufficient quantity of groundwater to supply a public water system; and
i. Currently supplies drinking water for human consumption; or
ii. Contains fewer than 10,000 mg/L total dissolved solids; and
3. Which is not an exempted aquifer.
(United States Environmental, 3/11/16)
11
The base of underground sources of drinking water (USDW) in the southern San Joaquin Valley is not
clearly defined. The U.S. EPA’s Underground Injection Control (UIC) program established regulations
to protect aquifers that are USDW. USDW aquifers contain waters with total dissolved solids less
than 10,000 parts‐per‐million (ppm) total dissolved solids (TDS). A boundary for the 10,000 ppm
TDS horizon needs to be established so the injection of saline produced waters will not contaminate
potentially useable water resources. Water in aquifers greater than 500 ppm TDS is not frequently
used for the drinking water supply, but protecting water under 10,000 ppm TDS will safeguard water
resources that can be made useable through current or future treatment methods (Lyle, 1988). In
this study the units of mg/L and ppm TDS are used interchangeably although in reality the units
begin to differ significantly at concentrations above 30 to 40 thousand ppm due to change in the
specific gravity of waters with higher salinities (Schlumberger, 2009).
This study uses geophysical logs in conjunction with geochemical analyses to determine the water
quality of the aquifers within the oil fields of the southern San Joaquin Basin in Kern County,
California. Direct sampling and chemical analyses of the water from oil and gas producing
formations provides the most accurate values for the salinity of the formation waters, but chemical
analyses are not common in deeper aquifers. Many chemical tests are also done on perforated or
near production intervals, and those intervals may not coincide with USDWs, therefore, log analyses
are used to bridge the gap in data. Most oil wells in the basin have open‐hole geophysical logs. The
abundance of open‐hole logs gives the opportunity to relate the water resistivity to its salinity and
determine the base of USDW from petrophysical analyses.
12
Geological Setting The San Joaquin Valley is in the southern portion of the Great Valley of California. It is separated
from the Sacramento Valley in the northern part of the Great Valley by the Stockton Arch. Figure 1
shows the San Joaquin Valley bounded to the west by the Coast Ranges, to the south by the San
Emigdio Mountains (Transverse Ranges), and to the east by the Sierra Nevada Mountain Range
(Croft, 1972). The 25,000 feet of continuous basin fill in the San Joaquin Valley records the late
Mesozoic through Cenozoic geologic history of the western margin of North America (Scheirer and
Magoon, 2007).
The San Joaquin Basin is an asymmetric structural trough which initially formed as a late Mesozoic
and early Cenozoic forearc basin (Bartow, 1991). The basin was submerged in the middle Eocene by
a basin‐wide transgressive cycle. The sedimentation was dominated by fine‐grained deposition
during this period (Scheirer and Magoon, 2007). By the Miocene epoch, much of the deep water
setting was confined to the southern San Joaquin Basin, resulting in deposition of organic rich shales
and siliceous oozes and associated turbidite sands. Figure 2 shows the stratigraphic column of the
southern San Joaquin Basin.
During periods of sea‐level transgression in the San Joaquin Basin, at least three high‐quality source
rocks were deposited (Scheirer and Magoon, 2007). This makes the San Joaquin Basin a prolific
hydrocarbon extraction area. A total of six petroleum systems are identified in the San Joaquin
Basin: the San Joaquin (?), McLure‐Tulare(!), Antelope‐Stevens(!), Tumey‐Temblor(.), Kreyenhagen‐
Temblor(!), and Moreno‐Nortonville(.) petroleum systems. The petroleum systems are named by
the source rock followed by the reservoir rock, and the certainty of the existence of these systems is
indicated by the punctuation: speculative (?), hypothetical (.), and known (!) (Magoon et al., 2007).
13
These petroleum reservoirs usually contain water too saline for most uses, although some waters
produced in conjunction with oil extraction on the eastern and western margins of the basin are
suitable for agricultural use (Croft, 1972).
Figure 1: Location map of the southern San Joaquin Valley showing the study area. The study area in the Kern County in the southern San Joaquin Basin is within the box. The Sierra Nevada Mountains, Coast Ranges, and the San Emigdio Mountains are labeled on the map.
14
Figure 2: The stratigraphic column for the southern San Joaquin Valley is shown above with red signifying a gas reservoir rock and green signifying an oil reservoir rock. The shales of the Kreyenhagen Formation represent the middle Eocene transgression. The Miocene shows the various oil reservoir rocks deposited around organic rich shales. The map to the right shows an outline of the San Joaquin basin with the stratigraphic column representing the part of the basin in the grey box (Modified Scheirer and Magoon, 2007).
15
The last major regression in the basin started in late Miocene and continued through the Pliocene as
the sedimentation in the basin axis progressed from a marine shelf (Etchegoin Formation) to fluvial
and lacustrine (Tulare Formation) deposits (Bartow, 1991). The Pleistocene and Holocene deposits
in the basin are composed of alluvial and lacustrine sediments due to the intermittent closure of the
basin to the ocean by tectonic activity (Bartow, 1991). The alluvial deposits of the Tulare Formation
form the main freshwater aquifers. Lacustrine clays separate these alluvial deposits into separate
aquifers in some parts of the basin (Croft, 1972).
Previous Work
A study by R.W. Page (1973) used a specific conductance value of 3,000 µmhos/cm in the southern
San Joaquin basin to map fresh water aquifers with a salinity of 2,000 mg/L total dissolved solids—
termed the base of fresh water in the valley. Conductivity is the inverse of resistivity, and it is
defined as Cw = 10000/Rw, therefore, 3,000 µmhos/cm is around 3.3 ohm‐m in resistivity. Page
(1973) noted that the base of fresh water is deeper on the eastern margin near the Sierra Nevada
Mountain Range and shallower on the western margin near the Coast Ranges. The data used for the
analysis in Page’s (1973) report were electrical logs from oil wells, chemical analyses of the water
wells, and previous hydrologic and geologic reports. Nearly 3,000 well logs were included in this
study to determine water salinity. The salinity of the water under the mapped fresh water increases
at different rates in different areas (Page, 1973). A similar method is used in this study to establish
the 10,000 ppm TDS boundary in the southern San Joaquin Basin in Kern County.
Methodology
The best method to determine the total dissolved solids (TDS) of formation water is to perform a
chemical analysis from water samples obtained from oil or water wells (Mitchell, 1994). These
16
analyses are limited in number in the southern San Joaquin Basin. However, the basin is a mature
petroleum producing province, and as such, there is an abundance of open‐hole electrical and other
geophysical logs. The well logs provide an indirect method to estimate TDS in formation waters
using empirical charts and equations. Two of the methods used to determine the TDS concentration
are the spontaneous potential and resistivity‐porosity methods.
Spontaneous Potential Method (SP Method) The spontaneous potential log measures the electrochemical factors that occur as a result of
differences in salinity between the mud filtrate and the formation water. The electric charge is
created by the flow of ions from concentrated solutions (with the primary ions typically being Na+
and Cl‐) to more dilute solutions such as fresh drilling mud in the borehole (Selley, 1998). This ion
flow process may reverse where the drilling mud is more saline than formation waters as the ions
flow in the opposite direction from the mud to the formation. When the formation water resistivity
(Rw) is greater than the mud resistivity (Rm), the SP curve will show a positive deflection in the sands
(Fig. 3a). If the formation waters are more saline than the drilling mud filtrate, the deflection
opposite the sands will be negative (Figure 3a). The SP curve will look featureless when the two
resistivity values are equal. Bed thickness, high formation resistivity, deep mud filtrate invasion, and
shale or clay content in the vicinity of the borehole also affects the deflection behavior of the SP
curve (Fig. 3b, Asquith, 1982). The spontaneous potential curve deflection in these areas needs to
be corrected to static spontaneous potential (SSP) which is the maximum deflection of the SP curve
from the shale baseline in a clean sand formation from the pseudo‐static deflection shown in figure
3b. The pseudo‐static deflection is the deflection in the area with a suppressed SP curve due to
shale effects.
17
The SP method is reliable in areas where there are clear alternating sand and shale sequences, but it
breaks down when carbonates or low porosity rocks are encountered (Schnoebelen et al., 1995).
This method should apply to the predominantly sand and shale lithologies of the southern San
Joaquin Valley. The information needed for the SP calculation is often found in the header of most
well logs (Figure 4): the mud resistivity (Rm), mud‐filtrate resistivity (Rmf), and the bottom‐hole
temperature (BHT) are all located within the log header information.
The mud resistivity values are needed to calculate the formation water resistivity (Rw) which is
related to the salinity. Equation 1 (Lyle, 1988) shows the relationship of the SSP to the resistivity of
the formation water and the mud filtrate for pure NaCl solutions.
(Equation 1)
Variables:
SSP = static spontaneous potential
K = temperature constant
Rmfe = Equivalent mud‐filtrate resistivity
Rwe = Equivalent formation water resistivity
Equivalent resistivity is the value of the resistivity in a pure NaCl solution. The temperature
constant, K, is calculated from equation 2 (Lyle, 1988), and it is dependent on the temperature of
the formation.
61 0.133 (Equation 2)
Variable:
Tf = formation temperature (°F)
18
Figure 3: In figure 3a, the deflection of the spontaneous potential curve is shown with varying levels of mud filtrate resistivity (Rmf) compared to formation water resistivity (Rw). When the mud filtrate resistivity is lower than the formation water resistivity, the SP curve will deflect positive. When Rmf is greater than Rw, the SP curve will deflect negative. Figure 3b shows the SP response to a constant salinity difference between Rmf and Rw but with various types of fluid and reservoir combinations (Asquith, 1982). The SSP represents the maximum deflection in a clean sand, water‐wet sandstone. The PSP represents the SP response if shale is present.
a) b)
19
Figure 4: This header from the well Johnson 64X‐32 (API: 02965147) in the West Bellevue field shows the information needed for the spontaneous potential method calculations. The values needed are the mud resistivity (Rm), the mud‐filtrate resistivity (Rmf), and the bottom‐hole temperature (BHT). The formation temperature is dependent on the bottom‐hole temperature (BHT), and equation 3
shows how to obtain Tf using a linear gradient and the BHT recorded in the well header.
20
(Equation 3)
Variables:
Tsurf = average surface temperature (⁰F)
BHT = bottom‐hole temperature
TD = total depth of the log curve where BHT is measured
Df = depth of the formation of interest
The SSP value is best obtained from a water‐wet and clean, non‐shaly sand formation. A water‐wet
zone is characterized by a large negative SP deflection (assuming that the mud filtrate is fresher than
formation waters) and a large separation between shallow and deep resistivity curves (saline
formation waters) where the shallow resistivity reads higher than the deep resistivity (Figure 5).
Another way to find a clean, wet zone is to look for the oil‐water contact (OWC) which is
characterized by a sharp drop in deep resistivity in the lower part of a sand interval (Figure 6). The
sharp drop in resistivity occurs because the saline formation water below the oil zone has a much
higher conductivity (lower resistivity) than the overlying oil bearing interval. The deep resistivity (Rt)
measures the uninvaded zones of the formations (areas not affected by the borehole fluids).
Finding a water‐wet zone is important because the presence of hydrocarbons will suppress the SP
curve deflection and affect the calculation. Finding a clean sand is equally important because clays
within the sand matrix will also suppress the negative deflection of the SP curve.
Thick sand beds allow the SP values to reach static conditions and achieve maximum deflection (SSP)
from the shale baseline (Lyle, 1988). In thin beds (< 10 feet), the shales above and below the sand
suppress the SP signal of the sand due to the averaging effects of the logging tool. In this case, the
21
SP deflection value needs to be corrected to static SP using the chart in figure 7. The next steps to
determine Rw using the SP method is done with empirical charts shown in Asquith (1982) or the
equations shown in table 1. The equations 4 to 11 allow for automation of the method through
programming languages or spreadsheet software. In summary, these equations uses the initial
mud‐filtrate resistivity from the log header and converts it to equivalent formation water resistivity,
which is then used to find the apparent water resistivity that will yield a relationship to the
formation salinity.
Table 1: Equations for the SP method (from Asquith, 1982)
75 6.7781.77
Equation (4)
10 Equation (5)
If Rmf at 75oF < 0.1, 146 5337 77
Equation (6)
If Rmf at 75oF > 0.1, 0.85
Equation (7)
Equation (8)
If Rwe < 0.12,
7577 5
146 377
Equation (9)
If Rwe > 0.12,
75 0.58 10 . .
Equation (10)
75 81.776.77
Equation (11)
22
Figure 5: Spontaneous potential and resistivity logs are shown for well 74X‐35 in the Bellevue oil field. The left is the spontaneous potential curve and the right is the resistivity curve. The resistivity has units of ohm‐m. The O.S. curve on the resistivity track represents the off scale resistivity values. The brown line marks the shale baseline. The SP deflection of the sand at 3565 feet (PSP, upper blue arrow) is less than the total deflection of the sand at 3710 feet (SSP, lower blue arrow) due to the thin bed effects. The resistivity curve for the sand at 3660‐3720 shows a large separation between the deep (dashed curve) and shallow (solid curve) resistivity values indicating the presence of a permeable, saline, wet zone.
Deep Resistivity
Shallow Resistivity
Spontaneous Potential Resistivity Curve
SSP
PSP
23
Figure 6: The deep resistivity is shown by the dotted curve and the shallow resistivity is shown by the solid curve. In the sand at 1150 feet, the deep resistivity is initially high, representing a hydrocarbon source. Deeper into the sand, at a depth of approximately 1175 feet, the deep resistivity value drops to 2 ohm‐m representing a zone filled with saline water. The Rmf at the measured temperature is converted to Rmf at 75°F using equation 4. If the well header
does not have an Rmf value, then the mud resistivity (Rm) can be used to estimate Rmf using figure 8.
Following this, equation 5 calculates the ratio between equivalent mud‐filtrate and formation water
resistivity using the temperature constant (Equation 2) and the SSP (from SP curve). Equivalent
resistivities are resistivities corrected to assume 100% NaCl as ions in the fluid. Equations 6 and 7
calculate the equivalent mud‐filtrate resistivity. Equation 6 is used when Rmf is less than 0.1 and
equation 7 is used when Rmf is greater than 0.1. The equivalent formation water resistivity, Rwe, in
equation 8 is calculated by dividing the Rmfe value by the Rmfe/Rwe ratio from equation 5. Equations 9
and 10 are used to calculate the formation water resistivity at 75°F. The final step is to convert the
resistivity from surface temperature at 75°F to the formation temperature using equation 11. The
resulting formation water resistivity value is used to calculate NaCl salinity for the formation by
Oil‐Water Contact
0 200‐200 mV OHM‐M
Spontaneous Potential Resistivity Curve
24
using the nomograph in figure 9 which relates temperature and resistivity to salinity in parts‐per‐
million TDS.
Figure 7: This plot determines the SP correction factor to correct PSP to SSP according to the thickness of the study zone and the ratio between the shallow and mud resistivity (Ri and Rm respectively) (Asquith, 1982).
25
Figure 8: This chart shows the relationship between the mud resistivity values with different mud densities in the event that only mud resistivity (Rm) is given and mud‐filtrate resistivity (Rmf) is needed. The dotted line shows the relationship between Rm and the mud‐cake resistivity (Rmc). The solid lines show the relationship between Rm and Rmf (Weatherford Ltd., 2009).
26
Figure 9: The nomograph allows the determination of salinity in k ppm TDS or grains/gal to the temperature (oF) to resistivity (ohm‐m) of formation waters. Resistivity Porosity Method (RP Method) The concept of the resistivity log assumes that the rock grains in a formation are non‐conductive,
and the only material conducting the electric current in the formation are fluids in the pores of the
rock‐‐ especially saline waters. The most common tool for measuring resistivity is the electric log
27
tool, which uses an alternating current to create an alternating magnetic field in the formation. The
flowing current is read by a receiver that measures the conductivity, and the inverse of the
conductivity is the resistivity, usually measured in ohm‐meters (Asquith, 1982).
The resistivity porosity method uses the deep resistivity of the formation along with porosity data
derived from geophysical logs or core samples to determine the formation water resistivity using the
Archie Equation in equation 12 (Archie, 1942). The logs that measure the deep resistivity are
typically the deep induction log (ILD), the deep laterolog (LLD), or the long normal (Asquith, 1982).
ɸ (Equation 12)
Variables:
Sw = water saturation (percent)
a = tortuosity constant
m = cementation constant
n = saturation exponent
ɸ = formation porosity (percent)
Rw = formation water resistivity (ohm‐m)
Rt = true formation resistivity (deep resistivity read from the electric log) (ohm‐m)
In summary, equation 12 takes reservoir values of porosity, deep resistivity, and empirical constants
to calculate the water saturation. But with the saturation known, the equation can be simplified
and rearranged to find the formation water resistivity. The tortuosity factor (a) and cementation
factor (m) change in different types of reservoirs. The value of ‘m’ can be determined from the slope
of the cross‐plot of porosity and the resistivity in a 100% water saturated zone (Lyle, 1988). Typical
28
ranges of the cementation factor and tortuosity are shown in table 2. The cementation constant
varies in different formations and even in different samples, but the value usually sits between 1.3
and 3.0 (Salem and Chilingarian, 1999).
Table 2: Range of cementation (m) and tortuosity (a) values (from Lyle, 1988)
Rock Type Cementation factor (m)
Highly cemented: limestone, dolomite, quartzite 2.0 – 2.2
Moderately cemented: consolidated sands 1.8 – 2.0
Poorly cemented: friable, crumbly sands 1.4 – 1.7
Unconsolidated sand 1.3
Tortuosity constant (a)
Carbonates 1.0
Unconsolidated sands 0.62
Consolidated sands 0.81
The standard form of the Archie’s equation can be used or variants, such as the Humble and Tixier
equations can be used. The standard Archie’s constants are set at a = 1, and m = n = 2. The Humble
equation is a generalized form of Archie’s Equation used for unconsolidated sand formations
(Winsauer et al., 1952). This equation uses 0.62 for tortuosity (a) and 2.15 for cementation (m). The
Tixier equation (a = 0.81 and m = 2) (Lyle, 1988) is another form of the Humble equation where the
fractional cementation constant is eliminated from the exponent on the porosity (Doveton, 2014),
making calculations simpler before computers took over most of the calculating work. Special core
analysis is required to determine the cementation and tortuosity constant in a formation, but these
analyses are seldom undertaken and none were present in the online data files.
The deep resistivity value (Rt) is related to the salinity of the water in the formation beyond the zone
of invasion by mud filtrate during drilling. This relationship is also dependent upon temperature,
porosity and the values used for the Archie exponents ‘a’ and ‘m’. The formation water resistivity
(Rw) is related to the salinity through the formation temperature using the nomograph in figure 9.
29
Both of these resistivity values were plotted against the measured salinity in the wells with chemical
analyses.
In a water wet formation, the water saturation, Sw, is assumed to be 1, and raising both sides to the
power of n eliminates the saturation exponent from the equation. Rearranging the equation to
solve for formation water resistivity, Rw, gives equation 13.
ɸ
(Equation 13)
The porosity data required for the calculations come from either core samples or geophysical logs.
Most of the porosity data in the southern San Joaquin Valley is from porosity logs such as the
density, neutron, and sonic logs. Many core descriptions were made in wells of the study area, but,
while core descriptions are common in the well histories, core analyses with measured porosity data
are sparse. This data restriction guided the study to depend extensively on geophysical logs such as
the sonic, neutron and density logs.
A type of neutron log measures the hydrogen ion concentration in a formation by using neutron‐
emitting elements of americium and beryllium to continuously bombard the formation with
neutrons. The neutrons are roughly the same mass as a hydrogen atoms nucleus. Collision of the
emitted neutrons with the nuclei of hydrogen atoms of similar mass causes emission of gamma rays
from the atomic nucleus. The number of gamma rays emitted are proportional to the hydrogen
content (Selley, 1998). This method relates the scattered gamma rays to porosity because, in the
relative absence of clays, the majority of the hydrogen nuclei are contained in the formation fluids
occupying the pore space (H2O for water, and CxHx for hydrocarbons). The exception to this
assumption is in shale where the OH‐ groups of the clay minerals as well as bound water common in
30
some clay minerals cause the neutron log values to read a higher porosity (Selley, 1998). Another
neutron logging tool (thermal neutron devices) detects the energy levels of the neutrons emitted
into the formation. In this type of neutron tool, the difference in energy of neutrons at the emitter
and the detector is related to the porosity of the formation (Ellis and Singer, 2007). The porosity
values can be read directly from the neutron log because the curve shows the percent porosity in
the formation (Figure 10).
Another geophysical log that can be used to obtain porosity values for the RP method is the bulk
density from the density log. The density log measures the electron density of a formation by
emitting gamma rays into the formation from a gamma ray source such as Colbalt‐60 or Cesium‐137
(Asquith, 1982). The density log is often plotted as bulk density units in grams per cubic centimeter
(g/cc) as shown in figure 11, and equation 14 shows how to convert from bulk density to density
porosity.
ɸ (Equation 14)
The logged density ( ) is the bulk density (RHOB) from the log curve and measured in g/cc. The
matrix density ( ) and fluid density ( ) are values relating to the type of formation and
fluid modeled. For sandstones, which are the typical reservoir lithology encountered in the San
Joaquin Valley, the matrix density is 2.648 g/cc (Table 3). More accurate values for matrix density
are obtained from special core analyses. Standard values for common fluid and matrix densities are
shown in table 3. In the example in figure 10, the density is displayed as density porosity (DPHI). In
this case, the density porosity is calculated using a specified matrix density representing the
formation, and the porosity can be read directly from the log.
31
Figure 10: An example of a density‐neutron log is shown. The neutron porosity curve (NPHI) is the dashed curve and the density porosity is the solid curve (DPHI). In this case, both the density and neutron curves have been converted to formation porosity using a specified matrix density. Another type of a density‐neutron log has the density curve in bulk density instead of percentage porosity.
32
Figure 11: In this example of a density curve, the bulk density of the formation is displayed with a scale usually ranging from 2 to 3 grams/cc (cubic centimeter). The density porosity of the formation will have to be calculated from the bulk density curve. Equation 14 converts bulk density to density porosity using an appropriate matrix density.
33
Table 3: Density Values for Matrix Materials and Fluids (Asquith, 1982)
Material Density (g/cc)
Sandstone 2.648
Limestone 2.710
Dolomite 2.876
Anhydrite 2.977
Salt 2.032
Salt Water 1.100
Fresh Water 1.000
Gas 0.700 The final type of geophysical log used for the RP method is the sonic log. This type of geophysical
log is the earliest type of porosity log, preceding the neutron and density logging methods. The
sonic tool consists of a sound transmitter and a set of receivers. The tool measures the speed of an
acoustic wave through one foot of formation rock, and the log curve is plotted in µs/ft. The
equation for calculating the sonic porosity is shown in equation 15 (Wyllie Method).
ɸ∆ ∆
∆ ∆ (Equation 15)
In this equation, ∆Tlog represents the interval travel time and is read from the sonic curve. The
∆Tmatrix is the travel time for the type of matrix material present, and ∆Tfluid is the interval travel time
for the fluid present. The values of commonly used interval transit times are listed in table 4.
Table 4: Interval transit times for different materials (Asquith, 1982)
Material Interval Transit Time (µsec/ft)
Sandstone 55.5 – 51.0
Limestone 47.6 – 43.5
Dolomite 43.5 – 38.5
Anhydrite 50.0
Salt 66.7
Iron Casing 57.0
Freshwater 189
Saltwater 185
34
Salinity Calculations The log calculated ppm TDS from the SP and RP methods are compared to the geochemical analysis
of the well to calibrate the calculations. The water chemical analyses are found on the Department
of Conservation Oil and Gas Division FTP site
(ftp://ftp.consrv.ca.gov/pub/oil/D4%20Chemical%20Analysis/, 1/27/2015). Figure 12 shows the
distribution of wells with a chemical analysis found on the DOGGR FTP site along with the wells
containing a calculated TDS value. The TDS values are calculated when both a chemical analysis and
porosity log are present for the RP method. The salinity in ppm TDS is calculated for 110 wells using
the SP method and 51 for the RP method.
Figure 12: This map shows the chemical (tested wells) and wells with porosity information available in the study area. The tested wells have a good representation throughout the area, but these wells also have poor porosity coverage. The blue triangles show the wells with salinity calculations from logs. These wells have both chemical analyses and porosity logs.
Legend
SJV_Oilfields
CA_Basins
CA_Counties
! Tested Wells
#* Calculated Wells
35
Well information for this study is obtained from the geophysical logs and well summaries located on
the website of the California Division of Oil, Gas, and Geothermal Resources (CA DOGGR,
http://www.conservation.ca.gov/dog/Pages/Index.aspx, 6/04/2014). Spontaneous potential and
resistivity logs are common, but porosity logs are not. Porosity logs did not come into wide use until
later in the 1960s and 1970s; so many wells drilled prior to those decades only have SP and electric
logs. Porosity logs are often not available for the wells with chemical tests; therefore, porosity logs
over the sampled interval from a nearby well must be used or the porosity must be estimated,
increasing the possibility of error.
Most of the analyses were obtained from produced waters or from drill stem tests during drilling.
The chemical analyses used date from the oldest in 1927 to a few tested in the 2000s. The chemical
analyses used are mainly from pre‐waterflood and pre‐injection dates. Analyses after injection or
secondary recovery may affect the original state of the reservoir. The formations tested or within
the production interval may not be useful for the RP calculations due to the possibility of
hydrocarbons present. If a formation has hydrocarbon saturation, resistivity values from an
overlying or underlying water‐wet sand formation was used for the calculation. The depth
information of the underlying or overlying sand is noted in the calculations for future reference.
The formations are determined to be water‐wet through core descriptions, core analysis reports, or
a mudlog. The core descriptions contain mostly qualitative data such as stain, fluorescence or gas
shows representing the presence of hydrocarbons, and do not provide numerical values for the
saturation data. Information from the core description may relay information about oil shows in the
formation, apparent wet sand, or fluorescence and odor. These descriptions help determine if the
36
sand contains hydrocarbons. A water‐wet formation will have no oil or gas show and no
fluorescence or odor. A sand formation with this description is assumed to be 100% water‐wet.
A core analysis gives a quantitative value of the hydrocarbon saturation in the formation as
measured in laboratory tests. Water‐wet formations in core analysis reports are identified as
sandstones with an oil or gas saturation value of 0.00%.
The Archie’s constants used also differed between different formations. The shallower formations
near the surface with soft, loose sands and poor consolidation used the Humble formula (a = 0.62
and m = 2.15) for Rw calculations (Van Meir and Lebbe, 2003). This is usually the case on the
western edge of the basin where the formations of interest are only a few hundred feet deep. The
more consolidated formations, such as the deeper Stevens and Vedder sands, used the standard
Archie’s constants (a = 1 and m = 2). A nomograph such as that shown in figure 9 or equation 16
(Baker Hughes Inc., 2002) is used to relate formation temperature and formation water resistivity to
salinity.
10 . . / . (Equation 16)
Variables:
Rw75 = formation water resistivity at 75⁰F
Salinity = formation NaCl salinity in parts‐per‐million TDS
.
. (Equation 17)
Variables:
R1 = initial resistivity at 75⁰F
T1 = initial temperature (75⁰F)
37
R2 = final resistivity at 10,000 ppm TDS boundary
T2 = final temperature (in ⁰F) at formation depth
Shown below is a sample calculation using the SP method for well Johnson 64X‐32 (API: 02965147)
in the West Bellevue oilfield.
1. Gather mud resistivity and temperature information from the log header.
a. Rmf = 2.67 ohm‐m @ 73⁰F
b. BHT = 184⁰F
c. Log total depth (TD) = 9744 feet
2. Find a water‐wet sand near tested interval and record SP deflection.
a. Perforated interval:
b. Depth of water‐wet formation: 9050 feet
c. SP deflection: ‐70 mV
3. Calculate the temperature constant and formation temperature using equations 2 and 3.
a. Surface temperature estimated at 70⁰F
b. Using equation 3, 70 9050; Tf = 175.8⁰F
c. Using equation 2, 61 0.133 175.8 ; K = 84.4
4. Use equation 4 to calculate Rmf at 75⁰F and at formation temperature
a. @75⁰F 2.67. .
.; Rmf@75⁰F = 2.605 ohm‐m
b. Use figure 9 to get Rmf at 175.8⁰F; Rmf = 1.167 ohm‐m
38
5. Use equation 5 to find the ratio of mud‐filtrate and formation water resistivity
a. 10 . 6.754
6. Calculate Rmfe with equation 6 if Rmf@75⁰F is less than 0.1 ohm‐m, if not, use equation 7.
a. Use Rmf at formation temperature; 0.85 1.167 0.992ohm‐m
7. Calculate Rwe with equation 8.
a. 0.9926.754 0.147 ohm‐m
8. Calculate Rw at 75⁰F. Use equation 9 if Rwe < 0.12, if not, then use equation 10.
a. @75 F 0.58 10 . . . 0.147 ohm‐m
9. Use figure 9 or equation 16 to find corresponding ppm TDS for Rw at 75⁰F
a. TDS = 43,937 ppm
10. Calculate error from the TDS recorded in the chemical report.
a. Chemical report TDS: 37,960 ppm
b. Error = (37960 – 43937)/37960 = ‐0.157
Below is an example using the RP method to calculate salinity for the same well (Johnson 64X‐32) in
the West Bellevue oilfield.
1. Find a water‐wet formation near tested interval (9050 feet).
39
a. Formation temperature = 175.8⁰F
2. Read off resistivity and calculate porosity of the water‐wet formation.
a. Deep resistivity from the deep induction log (ILD) reads 1.5 ohm‐m.
b. The neutron‐density porosity log shows a neutron porosity = 0.27 and density
porosity = 0.21.
c. Average the porosity between the neutron and density = 0.24.
3. Use equation 13 to calculate for Rw.
a. Use standard Archie constants of a = 1 and m = 2.
b. .
1.5 0.0864 ohm‐m (this is at formation temperature)
4. Use figure 9 or equation 16 to get salinity in ppm TDS
a. Use equation 17 to convert Rw at 175.8⁰F to Rw at 75⁰F (Equation 16 is calculated for
Rw values at 75⁰F)
b. Rw at 75⁰F = 0.180 ohm‐m
c. TDS = 34,794 ppm
5. Calculate error
a. Chemical report TDS: 37,960 ppm
b. Error = (37960 – 34794)/37960 = 0.083
Depth to USDW
40
The depth to USDW (10,000 ppm TDS) is determined by an initial analysis of the deep resistivity
curve. This calculation is performed on 182 wells in the southern San Joaquin basin. The formation
water resistivity at 75⁰F for a 10,000 ppm TDS sample is calculated using equation 16 by rearranging
the equation and solving for Rw,75. The formation water resistivity (Rw,75) is used to calculate the
deep resistivity (Rt) at 75⁰F for a 10,000 ppm TDS water sample by solving equation 13 for Rt. This
resistivity value is corrected to formation temperatures using equation 17 which is the generalized
form of equation 4.
The depth to base of USDW and, therefore, the formation temperature is initially unknown, so an
initial value is used to start an iterative process. The initial depth is picked by examining the electric
log for a water‐wet formation within a transition zone between high resistivity to low resistivity
sands. This transition zone marks the change from fresher to more saline waters where the 10,000
ppm TDS boundary may occur. The formation temperature is used to calculate the deep resistivity,
for a 10,000 ppm TDS water saturated formation at this depth. The Humble equation is used more
along the western side of the San Joaquin Basin because of the unconsolidated nature of saline
sands at shallow depths. The standard Archie’s equation is used for the aquifers found in deeper,
more consolidated formations in the center and southeastern areas of the basin. The calculated Rt
value is compared to the log Rt value to find the 10,000 ppm TDS water saturated formation. If the
calculated and log resistivity values do not match, the depth is varied until a match occurs. In this
iterative process, the porosity also adjusts to the varied depths if a porosity log is available. Figure
13 shows an example of the determination of the 10,000 ppm TDS boundary when the calculated
deep resistivity value for the water wet formation matches the electric log value. The Rt values from
the deep resistivity curve that are lower than the calculated value for 10,000 ppm TDS would be
more saline, and the higher values would indicate fresher waters.
41
Figure 13: The deep resistivity is calculated for a 10,000 ppm water saturated formation at surface temperature and corrected to formation temperature using an iterative process. The formation temperature is unknown initially and a depth is picked at which the deep resistivity changes. This depth is a potential candidate for the base of USDW. Shown above (Well Brandt 36X‐27 in Bellevue oilfield, API: 03018662), the resistivity drops to 3 ohm‐m or less at depths greater than 3550 feet and is over 4 ohm‐m in formations near 3480 feet. The porosity at 3480 feet is 0.30, and the calculated deep resistivity for 10,000 ppm at the temperature for 3480 feet is 3.5 ohm‐m. The log reads 4 ohm‐m so the resistivities do not match and the boundary should be deeper. At 3560 feet, the porosity is 0.29 and the calculated deep resistivity is 3.7, but the log reads 3 ohm‐m so the boundary should be shallower. The deep resistivity calculated at 3532 feet with a porosity of 0.30 is 3.5 ohm‐m (red line), and there is a match between the resistivity log and calculated deep resistivity so the boundary of the 10,000 ppm is placed there (blue line). Resistivity Analysis
0
1.65 2.65
0.60 0Deep Resistivity
Base of USDW
42
The resistivity of the formation is related to the salinity of the formation waters through the Archie
equation. Therefore, the deep resistivity value (true formation resistivity) should be proportional to
the salinity in the absence of hydrocarbons, clays or thin bed effects. The deep formation resistivity
can be gathered in two ways. The first method is to read the deep resistivity value from the log
opposite the perforated or tested intervals. However, the resistivity values obtained in this way may
not represent the true formation waters due to the likely presence of hydrocarbons in these zones.
Another method is to look for a water‐wet sand formation near the tested interval and record the
deep resistivity of a wet sand at a nearby depth. This will give a better relationship between the
deep resistivity and the formation water resistivity because of the absence of the hydrocarbon
effect. A wet sand can be indicated by a mudlog indicating no show or fluorescence, a core
description, or a deep resistivity value much lower than that of the producing sands.
Results
Water salinity calculations using the SP method were performed on 110 wells with chemical
analyses in the Department of Conservation FTP site. The data includes wells from 24 oil fields in
the southern San Joaquin Valley within Kern County. The salinity of the water ranges from 570 ppm
TDS (very fresh) to 46,200 ppm (very saline) TDS. Equations 4 through 11 were entered into a
spreadsheet to calculate the formation water resistivity (Rw) using the SP curve deflection and the
mud‐filtrate resistivity. The formation water resistivity is converted into an equivalent NaCl salinity
using the nomograph in figure 9 and compared with the salinity from the chemical test of the water
sample. The calculated salinity using the SP method is plotted against the salinity from the chemical
analyses in figure 14.
43
Figure 14: The SP calculated salinity plotted against the chemical analysis values (N = 110). The bold line indicates a perfect match between the tested and calculated values. The results show weak to no correlation between the measured and calculated salinities (R2 = 0.4686) as the calculated values show large deviation to the 1‐to‐1 line through the entire range of tested salinities. The calculated versus tested salinity plot for the SP method shows a weak correlation in figure 14.
The bold line on the plot shows a 1‐to‐1 relationship between the calculated salinity and tested
salinity. Deviation from this line shows the error between the SP method and the chemical tests.
There is a general trend showing that the calculated salinity increases as the tested salinity
increases, but the calculated values show a large deviation from the measured values throughout
the range of salinities. A linear regression of the data points with the intercept set to zero shows a
coefficient of determination, R2, of 0.467. This reveals that there is weak to no correlation between
the SP calculation and the tested salinity value, and the error is large for many of the data points.
R² = 0.4686
0
20000
40000
60000
80000
100000
120000
140000
0 10000 20000 30000 40000 50000
Calculated Salinity (PPM)
Tested Salinity (PPM)
SP Method
Series1
Linear (Series1)
44
The error from the SP method is plotted against tested salinity in figure 15. The error for the SP
method salinity calculation shows that many of the data points lie outside of the 20% error bars.
Only 25 out of the 110 data points for the SP method lay within the 20% error margin. The average
error for this method is 70.4%, indicating large inaccuracies when using the SP method as a
representation of salinity.
Figure 15: This is the plot of errors calculated from the SP method. The scatter of calculated errors across the range of salinity shows little correlation between salinity and error. The black lines show the ±20% error boundary with most of the data points for the SP method plotting errors outside of the boundary. The RP method was performed on 51 wells that had both chemical tests and porosity logs. The
porosity logs consisted of neutron, density, and sonic logs. These wells represent 14 different
oilfields in Kern County in the southern San Joaquin Valley. The neutron porosity was read directly
from the neutron curve while the density and sonic porosities are calculated using equations 14 and
‐1.000
‐0.800
‐0.600
‐0.400
‐0.200
0.000
0.200
0.400
0.600
0.800
1.000
0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000
Calculated Error (SP M
ethod)
Tested Salinity (PPM)
SP Method Error
Series1
45
15. The formation water resistivity, Rw, is calculated using equation 13 and related to the salinity in
ppm TDS by the nomograph in figure 9.
The tested versus calculated salinity plot in figure 16 shows the RP method has a strong trend
relating the measured and calculated salinities. The R2 value for the linear regression of the RP data
points with the intercept set at y = 0 is 0.807. There is limited scatter through the range of salinities
for the RP method, and the measured and calculated values show a strong correlation.
Figure 16: The resistivity‐porosity calculated salinity is plotted against the salinity from the chemical analyses (N = 51). The bold line indicates a perfect match between the tested and calculated values. There is a stronger correlation between the tested and calculated values in the RP method (R2 = 0.807) compared to the same calculations for the SP method (R2 = 0.467).
y = 0.8326xR² = 0.807
0
5000
10000
15000
20000
25000
30000
35000
40000
45000
50000
0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000
Calculated Salinity (ppm)
Tested Salinity (ppm)
RP Method Analysis
RP Method
Linear (RP Method)
46
Figure 17: The error from the RP method is plotted against the tested salinity. The bold lines represent the ±20% error boundary for the data. The RP method is more accurate in the lower (around 5000 ppm) and higher (around 40,000 ppm) salinities.
‐1.000
‐0.900
‐0.800
‐0.700
‐0.600
‐0.500
‐0.400
‐0.300
‐0.200
‐0.100
0.000
0.100
0.200
0.300
0.400
0.500
0.600
0.700
0.800
0.900
1.000
0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000
Calculated Salinity (ppm)
Tested Salinity (ppm)
Error Analysis
Errors
47
Figure 17 shows the error for the RP method plotted against tested salinity. This plot shows more
data points from the RP method lie within the 20% error margin for the entire range of tested
salinities. A total of 29 out of the 51 wells calculated with the RP method fell within the 20% error
margin. The average error for the RP method is 0.275.
Figure 18 shows the formation deep resistivity value from the electric log, Rt, plotted against the
salinity from the chemical analysis. The plot shows no relationship (R2 = 0.0115) between the deep
resistivity and the measured salinity of the formation water for the 86 wells with chemical analyses.
Most of the samples analyzed in the lab were from waters produced along with hydrocarbons. The
presence of hydrocarbons in the sampled formations affects the deep resistivity value on the
electrical log, causing higher resistivity readings due to the non‐conductive nature of hydrocarbons.
Figure 19 shows the deep resistivity, Rt, plotted against the salinity in ppm TDS using overlying or
underlying formations near the tested interval that are 100% water‐wet. The correlation in this plot
is much better than plotting deep resistivity of the tested zone against the tested salinity. This
correlation corresponds to a power function shown in figure 19. Using a power regression, the
coefficient of determination, R2, is 0.8059. This indicates the importance of using 100% water‐wet
formations in order to limit errors arising from the indiscriminate use of deep resistivity values.
Indications of the presence of hydrocarbons in the tested zones from mudlogs and/or conventional
or sidewall core analyses preclude the use of log resistivity values in determining water salinities
using either the SP or RP methods.
48
Figure 18: The deep resistivity is plotted against the tested salinity in the zones sampled for chemical analysis. The deep resistivity does not show a trend or correlation (R2 = 0.0115) to the tested salinity as it varies because the sampled zones may contain hydrocarbons. The presence of hydrocarbons causes the resistivity of the formation to increase so that it no longer represents the salinity of the water.
y = 9.138x‐0.085
R² = 0.0115
1
10
100
1000 10000 100000
Deep Resistivity, R
t (ohm‐m
)
Tested Salinity (ppm)
Deep Resitivity Analysis
Series1
Power (Series1)
y = 817.48x‐0.583
R² = 0.8059
1
10
100
800 8000
Form
ation Deep Resistivity, R
t (ohm‐m
)
Tested Salinity (ppm)
Formation deep resistivity vs Tested Salinity
Series1
Power (Series1)
49
Figure 19: The figure shows the formation deep resistivity in nearby wet sands, Rt, plotted against the formation water salinity measured by laboratory analysis of sampled water in logarithmic scale. There is a strong correlation (R2 = 0.8059) using a power function regression between the resistivity and salinity. The RP method has a better correlation between tested and calculated salinities, and also, it has a
lower average error than the SP method (0.275 for the RP compared to 0.704 for the SP). Due to
this, the RP method was used to calculate the depth to the 10,000 ppm TDS horizon in underground
aquifers in the southern San Joaquin Basin. To create the map, equation 13 is rearranged to solve
for deep resistivity, Rt, with porosity, the Archie’s constants, and the formation water resistivity
known. The standard Archie equation variables of a = 1 and m = 2 are used to determine salinity in
deeper (> 2,000 feet) zones. The deeper formations tend to be more consolidated than the
shallower formations. The Humble formula gives larger errors in the deeper, more consolidated
formations than the standard Archie formula when compared to the salinity measurements from
the chemical reports. Shallower formations ((depth < 2,000 feet) are better represented by the
Humble equation constants of a = 0.62 and m = 2.15 because these formations tend to contain soft,
unconsolidated sands. The iterative process to calculate this is detailed in figure 13. The formation
water resistivity is calculated to be 0.564 ohm‐m at 75 degrees F for a 10,000 ppm salinity source of
water. This calculation is performed by calculating for Rw in equation 16 and setting the salinity at
10,000 ppm. The water resistivity at 75 degrees F is corrected to formation temperatures for the Rt
calculation using equation 11.
Using the Rt calculated for the expected depth to 10,000 ppm TDS horizon, the electric log is
examined to determine the depth where the Rt is equal to the calculated value. Aquifers with Rt
greater than the calculated Rt are fresher, hence the greater resistivity. Resistivity values lower than
the calculated deep resistivity correspond to formation waters that are more saline than 10,000
ppm TDS.
50
Across the Bakersfield Arch area, the depth to 10,000 ppm TDS becomes increasingly shallow from
east to west (Figure 20). The depth to 10,000 ppm TDS in the eastern side of the southern San
Joaquin Basin near Edison and Mountain View oilfields extends to approximately 5,000 feet subsea.
In Mount Poso field north of the Edison and Mountain View oilfields, the USDW extends to
basement rock and there are no indicators for the base of freshwater which is defined at 2,000 mg/L
(ppm TDS) (Page, 1973). Since freshwater is much lower than 10,000 ppm TDS, it is safe to assume
the depth to the USDW boundary is non‐existent here. DOGGR datasheet volume 1 for Central
California indicates that the average depth to basement in Mount Poso is around 3,000 feet where
some drilling logs report encountering granite.
There are two trends to the change in depth for the 10,000 ppm TDS horizon across the basin from
east to west (Figure 20). At the center of the Bakersfield Arch n ear the Rio‐Bravo and Greeley
oilfields, the depth to 10,000 ppm TDS is 2,500 feet subsea but shallows to 250 feet subsea on the
eastern edge of the Elk Hills oilfield. This indicates that there is a significant increase in salinity
gradient westward from the region on the Bakersfield Arch near the Rio‐Bravo‐Greeley oilfields to
the area on the eastern flank of the Elk Hills oilfield. Further south in the Tejon sub‐Basin, the depth
to 10,000 ppm TDS is at approximately 5,000 feet along the basin axis. Farther west, at Yowlumne
oil field, the depth to 10,000 ppm TDS decreases to 3,000 feet subsea before shallowing to near the
surface still farther west at the Midway‐Sunset field.
Areas in the southeastern part of the San Joaquin Valley, especially in Edison oilfield, contain a non‐
USDW (saline) reservoir bounded above and below by USDW reservoirs. The Olcese sandstone
(depth around 3550 feet measured depth) in this area contains waters for which chemical analyses
51
noted salinities above 10,000 ppm TDS while shallower units (the Chanac and Santa Margarita
formations) and deeper units (Jewett and Pyramid Hills formations) are USDW’s. Well logs in the
area show higher resistivity values in the sands above and below the Olcese, although hydrocarbons
may affect the log values. The resistivity near the base of the Olcese reads between 3 and 6 ohm‐m,
but the resistivity increases to around 10 ohm‐m at the top of the Jewett Sand matching the higher
values in the Santa Margarita formation.
52
Figure 20: This map shows the subsea depth to the calculated 10,000 ppm TDS boundary in the
southern San Joaquin Basin. The boundary is deeper in the east and shallows toward the west.
The depth to the base of USDW is shallower over the Bakersfield Arch than in the Tejon Sub‐basin
to the south. Data points are more dense along the Bakersfield Arch and sparse in the south near
the Tejon Sub‐basin.
Tejon Sub‐basin
Kern River
Kern Front
Poso Creek
Rio Bravo
Greeley
Wheeler Ridge
0
‐1000 ‐2000 ‐3000 ‐4000 ‐5000 ‐6000
Oil Well
Bakersfield
Taft
Mount Poso
53
Sensitivity Analysis for the RP Calculations
Porosity The charts in figure 21 show the different variables used in the RP method and how the uncertainty
in their values affects the error of the salinity calculation using the well KM 23XA‐33 in the West
Bellevue oilfield. The main concern in the method is the poorly constrained porosity data. A 7‐10%
error in the porosity may cause a 20% error in the salinity (Figure 21c) calculation. Many wells with
chemical data were sampled as early as the 1920’s and 1930’s, a time period in which porosity logs
were unavailable. Only a small fraction of the wells with chemical analyses have porosity logs
needed to perform the RP method. In some cases, a nearby well has a porosity log from which a
porosity value can be established with some confidence. However, errors may arise from using
porosity logs from wells near the sampled well. This introduces uncertainty in the porosity value
because of the potential for facies changes between wells. A way to mitigate this is to use offset
porosity logs as close to the tested well as possible.
CA DOGGR lists the average porosity values for each formation in volume I of the Central California
datasheet publication (Figure 22). This value can be used to estimate salinity using the RP method
when logs are not available in nearby wells. Since, these values are reservoir averages, they
increase the likelihood of error in the salinity calculations for individual wells.
54
Figure 21: These plots show the effect of adjusting the variables in Equation 13 used to determine the formation water resistivity (Rw) in order to calculate salinity. The standard Archie equation constants of a = 1 and m = 2 are used in this analysis. The error is plotted for each value as it varies from the optimal value, calibrated to the tested formation water resistivity (Rw), showing the error effects of each adjustment. These values are obtained from the calculation for well 35‐38 (API‐02981159) in the Rosedale Ranch Oilfield.
a) b)
c) d)
55
Figure 22: CA DOGGR (1998) datasheets provides an estimate for formation porosity when log data are unavailable. These datasheets contain average values for each producing reservoir in each oilfield.
56
Resistivity Varying the deep resistivity (Rt) by ±17% creates a 20% error in the salinity calculation (Figure 21d)—
a significant amount. However, the resistivity variable is the best constrained of all of the variables
in the RP equation because electrical log data is widely available so this is seldom cause for concern.
Archie’s constants The tortuosity (a), and cementation factor (m) was varied over a likely range of values while holding
the other factors constant to determine the effect of each on the overall error in the calculated Rw
value and related salinity. The chart for the tortuosity factor (a) (Figure 21a) shows that a wide
range of values can be used and salinity can still be under the 20% error limit. The cementation
constant (m) shows a narrow range of values within the 20% error limit (Figure 21b), meaning the
Archie’s equation is more sensitive to the constant m than to a.
Conclusion
The RP method can be used to determine the base of USDW due to its strong correlation between
the calculated and tested values with a correlation coefficient value of 0.807. The 0.467 correlation
coefficient for the SP method shows a poor match between the calculated and tested salinity for the
SP method. The RP method has drawbacks in that it relies on porosity information from geophysical
logs. These logs are rare before the 1970s when most of the secondary recovery methods
commenced in the area. Therefore, the wells with chemical analyses have few accompanying
porosity logs necessary to compare salinity values derived from the RP method to measured salinity
values. Figure 12 shows the distribution of wells with chemical analyses and porosity logs.
The measured salinity versus deep resistivity analysis shows that the deep resistivity value is not a
good indicator of the formation water salinity (Figure 19). Most of the water samples in the
chemical analyses come from water produced in association with hydrocarbons and, therefore, the
57
aquifers contain both water and hydrocarbons. The presence of hydrocarbons increases the
resistivity value of the formation so that it is no longer representative of the formation water
salinity. Therefore, it is important to select hydrocarbon free (100% water‐wet) formations for
analysis. A better fit is obtained by plotting the deep resistivity vs measured salinity in a wet sand
(Figure 19a). This shows a strong correlation between the deep resistivity of the formation to the
measured salinity with an R2 value of 0.8059.
The sensitivity analysis for this study indicates that the main uncertainty controlling the error in the
RP calculation is the porosity value. A relatively small percentage change in the porosity can create
errors greater than 20% for the salinity calculation using the RP method. Therefore, it is important
to obtain good porosity data on the formations of interest. The resistivity is another variable that
has great effect on the final salinity calculation, but resistivity data is abundant and more reliable
than the available porosity data. The ‘a’ and ‘m’ constants of the Archie Equation have less of an
effect than the resistivity and porosity and therefore they can have greater range of uncertainty
without causing large errors in the calculation.
The map in figure 20 shows the base of USDW to be as much as 5,000 feet deep on the eastern and
southeastern boundary of the San Joaquin Basin. This may due to the abundance of freshwater
recharge from the west slope of the Sierra Nevada Mountains. Wells in many eastside fields, such
as Mount Poso and Round Mountain oilfields, contain USDW down to metamorphic and igneous
basement rock and, therefore, the base of USDW is non‐existent in the area. On the western side of
the basin, the depth to 10,000 ppm becomes increasingly shallow—and in some areas of the
Midway Sunset oilfield may be near surface. Recharge from streams emanating from the Coast
58
Ranges is insufficient to provide appreciable fresh water recharge to aquifers on the western margin
of the San Joaquin Basin.
59
Appendix I
Nomenclature Field Oilfield containing well API Well Identifier Wellname Name of well Formation Tested/perforated formation Perf Interval Interval perforated in feet TDS Tested salinity of the formation (ppm) Log_TD Total depth of well log (feet) BHT (F) Bottom‐hole temperature in ⁰F Fm_depth Depth of water‐wet formation near tested zone (feet) Fm_DRES Deep resistivity from water‐wet formation (ohm‐m) T_surf Average surface temperature in ⁰F T_fm Temperature of water‐wet formation in ⁰F Rmf @ T0 Mud filtrate resistivity at recorded temperature (ohm‐m) T0 (F) Temperature for recorded mud filtrate resistivity in ⁰F Rmf @ T_fm Mud filtrate resistivity at formation temperature (ohm‐m) Rmf @ 75F Mud filtrate resistivity at 75⁰F (ohm‐m) Rmfe Equivalent mud filtrate resistivity (ohm‐m) Rmfe/Rwe Ratio of effective mud filtrate resistivity to formation water resistivity K Temperature constant for the SP method SSP Static spontaneous potential Rw @ 75F Formation water resistivity at 75⁰F (ohm‐m) Rw @ fm Formation water resistivity at formation temperature (ohm‐m) TDS_Calc Calculated TDS from equation 16 (ppm) Error Error between calculated ppm TDS and tested ppm TDS DPHI Density porosity (decimal) SPHI Sonic porosity (decimal) NPHI Neutron porosity (decimal) PHIA Average porosity (decimal) A Archie tortuosity constant M Archie cementation constant T, R, S Township, section, and range of the well USDW Depth to the 10,000 ppm TDS boundary (feet) Rw_10K Calculated formation water resistivity for the USDW boundary (ohm‐m) Rt_10K Calculated formation deep resistivity for the USDW boundary (ohm‐m)
60
Appendix II
61
62
63
Appendix III
64
65
Appendix IV
66
67
68
69
70
References
Archie, G.E., 1942, The electrical resistivity log as an aid in determining some reservoir characteristics: American Institute of Mining and Metallurgical Engineer Transactions v. 146. Asquith, G., 1982, Basic Well Log Analysis for Geologists: American Association of Petroleum Geologists, Tulsa, OK, 216 p. Baker Hughes Inc., 2002, Introduction to Wireline Log Analysis: Baker Hughes, Inc. Bartow, J.A., 1991, The Cenozoic evolution of the San Joaquin Valley, California: U.S. Geological Survey Professional Paper 1501, 40 p. California Department of Gas and Geothermal Resources, 1998, California Oil and Gas Fields, Volume 1 – Central California: Sacramento, ftp://ftp.consrv.ca.gov/pub/oil/publications/Datasheets/Dtasheet_vol_1.pdf., accessed 1/26/2015 Croft, M.G., 1972, Subsurface geology of the Late Tertiary and Quaternary water bearing deposits of the southern part of the San Joaquin Valley, California: U.S. Geological Survey Water Supply Paper 1999‐H. Doveton, J.H., 2014, Principles of Mathematical Petrophysics: New York, NY, Oxford Press, 245 p. Ellis, D.V., and Singer, J.M., 2007, Well Logging for Earth Scientists, Second Edition: AA Dordrecht, The Netherlands, Springer, 692 p. Lyle, R., 1988, Survey of methods to determine total dissolved solid concentrations: U.S. Environmental Protection Agency KEDA Project No. 30‐956. Magoon, L.B., Lillis, P.G., and Peters, K.E., 2007, Petroleum systems used to determine the assessment units in the San Joaquin Basin Province, California, Chapter 8, in Scheirer, A.H., ed., Petroleum Systems of Geologic Assessment of Oil and Gas in the San Joaquin Basin Province, California: U.S. Geological Survey Professional Paper 1713, 65 p. Mitchell, W.M., 1994, Assessment of the 3,000 ppm and 10,000 ppm Total Dissolved Solids Boundaries in the Mississippian and Pennsylvanian Aquifers of Southwestern Indiana: Indiana University, Indiana Geological Survey Open‐File Report 94‐2. Page, R.W., 1973, Base of fresh ground water (approximately 3,000 microhms/cm) in the San Joaquin Valley, California. U.S. Geological Survey Hydrologic Investigations Atlas HA‐489. Salem, H.S. and Chilingarian, G.V., 1999, The cementation factor of Archie’s equation for shaly sandstone reservoirs: Journal of Petroleum Science and Engineering 23, 83‐93. Scheirer, A.H., and Magoon, L.B., 2007, Age, distribution, and stratigraphic relationship of rock units in the San Joaquin Basin Province, California, Chapter 5, in Scheirer, A.H., ed., Petroleum systems and geologic assessment of oil and gas in the San Joaquin Basin, California: U.S. Geological Survey Professional Paper 1713, p.107
71
Schlumberger, 2009, Log Interpretation Charts (2009 Edition): Schlumberger Schnoebelen, D.J., Bugliosi, E.F., and Krothe, N.C., 1995, Delineation of a saline groundwater boundary from borehole geophysical data: Ground Water, v. 33, no. 6, p 965‐976. Selley, R.C., 1998, Elements of Petroleum Geology: Gulf Professional Publishing, Academic Press, 470 p. United States Environmental Protection Agency, Aquifer Exemptions in the Underground Injection Control Program: https://www.epa.gov/uic/aquifer‐exemptions‐underground‐injection‐control‐program, last accessed 3/11/2016. Van Meir, N. and Lebbe, L., 2003, Deriving TDS values in coarse sediment from long normal and electromagnetic logs: Ground Water, v. 41, no. 1, p. 33‐40. Weatherford Ltd., 2009, Wireline Services Log Interpretation Chart Book: Weatherford Ltd. Winsauer, W.O., Shearin, H.M., Masson, P.H., and Williams, M., 1952, Resistivity of brine‐saturated sands in relation to pore geometry: Bulletin American Association of Petroleum Geologists 36, 253‐277.