QAKFUDGE ' ORNIJTM-2001/78

NATIONALLABORATORY
I MANAGED BY UT-BAlTELLE
FOR THE DEPARTMENT OF ENERGY

Characterization of Soluble
Organics in Produced
Water

D. T. Bostick
H. Luo
B. Hindmarsh

UT-BATTELLE
ORNL-27 (4.00)
 DOCUMENT AVAILABILITY

Reports produced after January 1, 1996, are generally available free via the U.S. Department
of Energy (DOE) information Bridge.

Web site http://www.osti.gov/bridge

Reports produced before January 1,1996, may be purchased by members of the public from
the following source.

National Technical information Service

5285 Port Royal Road
Springfield, VA 22161
Telephone 703-605-6000 (l-800-553-6847)
TDD 703487-4639
Fax 703-605-6900
f-mail info@ntis.fedworld.gov
Web site http://www.ntis.gov/support/ordemowabout.htm

Reports are available to DOE employees, DOE contractors, Energy Technology Data
Exchange (ETDE) representatives, and international Nuclear Information System (INIS)
representatives from the following source.

Office of Scientific and Technical Information

P.O. Box 62
Oak Ridge, TN 37831
Telephone 865-576-8401
Fax 865-576-5728 c

E-mail reports@adonis.osti.gov
Web site http://www.osti.gov/contact.html

This report was prepared as an account of work sponsored by

an agency of the United States Government. Neither the
United States Government nor any agency thereof, nor any
of their employees, makes any warranty, express or implied,
or assumes any legal liability or responsibility for the
accuracy, completeness, or usefulness of any information,
apparatus, product, or process disclosed, or represents that
its use would not infringe privately owned rights. Reference
herein to any specific commercial product, process, or
service by trade name, trademark, manufacturer, or
otherwise, does not necessarily constitute or imply its
endorsement, recommendation, or favoring by the United
States Government or any agency thereof. The views and
opinions of authors expressed herein do not necessarily state
or reflect those of the United States Government or any
agency thereof.
 ORNL/TM-200 l/78

Chemical Technology Division

CHARACTERIZATION OF SQLUBLE ORGANIC? IN PRODUCED WATER

Debra T. Bostick
H. Luo, and B. Hindmarsh

Date Published: January 2002

,a;,.

Department of Energy Project FEAB3 12

(Funded through the Natural Gas and Oil Technology Partnership Office)

Supporting Petroleum Energy Research Forum

Project 98-04

Prepared by
OAK RIDGE NATION& LABOR&TO&Y
Oak Ridge, Tennessee 3783 1-6285
operated by
UT-BATTELLE, LLC
for the
U.S. Department of Energy
under contract DE-AC05-OOOR22725
.
 TABLE OF CONTENTS
*

-5 LIST OF TABLES ..................................................................................................................................... V

LIST OF FIGURES .................................................................................................................................. vi
LIST OF ABBREVIATIONS.. ..................................................................................................................vii
EXECUTIVE SUMMARY ............................................................................................................y......... ix
/ c
1, BACKGROUND ........................................................................................................................a...; .. 1
1.1 Analytical Approach ................................................................................................................. 2
1.2 Experimental Approach ............................................................................................................. 2

2. COLLECTION AND SAMPLING OF ACTUAL OIL AND PRODUCED

WATER SAMPLES. .......................................................................................................................... 4
2.1 Field Samfiling of DEEP1 Crude Oil Received.. ...................................................................... 4
2.2 Determination of Water Content of As-received Oil Samples .................................................. 4
2.3 Laboratory Sampling of as-Received Crude Oil.. ..................................................................... 5

3. EXPERIMENTAL PROTOCOL FOR AMBIENT-PRESSURE TESTING .................................... 5

4. EXPERIMENTAL PROTOCOL FOR HIGH-PRESSURE TESTING ............................................. 5

8
5. SYNTHETIC GOM BRINE .............................................................................................................. 6
* 5.1 Preparation of Simulant ............................................................................................................ 6
5.2 Verification of GOM Brine Simulant Content .......................................................................... 7

6. ANALYTICAL METHODS.. ............................................................................................................ 8

6.1 Identification of Semi-Volatile Organic Compounds (SVOC)
in Actual Oil and Produced Water by GC/MS .......................................................................... 8
6.2 GC/FID Analysis of Total Petroleum Hydrocarbons.. .............................................................. 10
6.3 Infrared Analysis of Total Recoverable Petrochemical Hydrocarbons.. ...............-. .................. 11
6.4 Fractionation of WSO in Simulated Produced Water ............................................................... 11
6.5 Determination of pH Water Content in Crude Oil .................................................................... 15
6.6 Determination of pH and Bicarbonate/Carbonate Content in Water ........................................ 15
6.7 Determination of Chloride Content in Produced Water ............................................................ 16
6.8 Determination of Metal Content in Produced Water ................................................................ 16
6.9 Analysis of C,-C6 Organic Acids.. ........................................................................................... 18

7. ASSESSMENT OF DATA QUALITY ......... . ................................................................................... 18

7.1 Correspondence Between Analytical Methods for WSO Constituents.. ................................... 18
7.2 Surrogate Recovery And Blank Levels For Open LC Fractionation and GC/FlD Analysis .... 20
7.3 Biases Associated with LC Fractionation Protocol ................................................................... 21
* 7.4 Comparison of LC Fractionation Protocol with Infrared Spectrometric Analysis.. .................. 23

8. RESULTS ..........................................................‘...... ......................................................................... 23

8.1 Chemical Characterization of Deep-Well Crude ...................................................................... 23
i
8.2 WSO as a Function of P&cent Water Cut ................................................................................ 24
8.3 WSO Distribution Coefficients as a Function of Percent Water Cut ........................................ 24
8.4 WSO as a Function of Salinity .................................................................................................. 25
8.5 WSO as A Function of pH ........................................................................................................ 25
8.6 WSO as a Function of Temperature .......................................................................................... 25

...
111
 8.7 WSO as a Function of Pressure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
8.8 Composition of Cl-C6 Organic Acids in Produced Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
8.9 Presence of Inorganic Constituents in Crude and Produced Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
.
9. CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........
. 29

10. REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
I
11. APPENDIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

iv
 LIST OF TABLES

Table Page
5

1.1 Experimental test matrix for crude oil/GOM brine contacts.. ............................................... 3
5.1 Average concentrations of GOM brine components.. ............................ ............................... 6
5.2 Preparation of GOM brine simulant ..........................................................~.......~...! .................... 7
*
5.3 Anion concentrations in GOM brine.. ................................................................................... 8
5.4 Cation concentrations in GGM brine ......... .................................. . . ~........................................
........... 8
6.1 USEPA semivolatile organic compounds of interest to petrochemical applications ............ 8
6.2 Surrogate recovery standards for SVOC analysis ................................................................. 9
6.3 Internal standards for SVOC analysis ................................................................................... 9
6.4 Compounds included in laboratory control spike standard ..........................................~........ 9
6.5 GC/MS conditions for SVOC analysis ................................................................................. 10
.6.6 GC/PID operating conditions for TPH analysis .................................................................... 10
6.7 IC operating conditions for anion analysis ......................... :..............:.* ....................................
_ .,ll,,_ ~,_ ‘16
6.8 Inorganic analysis of GOM brine simulant, using ICP-MS .................................................. ‘17
6.9 IC Operating conditions for Cl-C6 organic acid analysis .................................................... 18
7.1 Correspondence of analytical methods in the characterization of deep-well crude .............. 19
7.2 Surrogate spike recovery compounds for LC fractionation protocol.. .................................. 21
‘7.3 Calculated blank levels assobated WitheacF LC f~@??. ..........................................................
... .......... I-i,-“,- 21
7.4 Correspondence of TPH results between,LCfractionation ... u 4&^“““./“>* and
...........IR*_spectroscopy.. . . ............. ‘23
8.1 Chemical fraction data for GOM deep-well crude.. .............................................................. 24
?
8.2 Chemical fraction data for GOM synthetic produced water ...................................:!............ 24
8.3 __ ___ ..~ .: J.r.:f;,T.‘. .... . ................
WSO distribution coefficients within chem&$l,~c,lasses~~ ............................ ..... 1. 28
8.4 WSO distribution coefficients within carbon ranges for TEM fractions .............................. 28
1

A.1 PERP experiments associated with each oil/water. contact parameter.. ................................ 52
A.2 Starting experimental conditions for PERP ezxperiments.. .................................................... 53
A.3 Equilibrium concentrations of inorganic constituents ........................................................... 55
A.4 Equilibrium concentrations of ICP metals ... ;, l..l’“..“t .z “‘1.........
..~%...11...4 ........ .e”*“Li.E,,-:<- ,.&/ “,*)/ . . . 55
......................................................
A.5 Equilibrium concentrations of organic matter by gravimetric and lR analyses .................... “56
A.6 Equilibrium concentrations of WSO in open LC fractions TEM, HEX, and TSAT ............. 57
A.7 Equilibrium concentrations of WSO in open LC fractions TARO, TPOL(subtract),
and TPOL (GC). .................................................................................................................... 58 .
I
 LIST OF FIGURES

Figure m
.
4.1 Schematic diagram of oil/water contact pressure vessel.. ..................................................... 34
4.2 Oil/brine contact pressure vessel ........................................................................................... 35
6.1 Fractionation of WSO in simulated produced water ............................................................. 13
6.2 Open liquid chromatographic column for fractionation of TPH.. ......................................... 14 , .
7.1 GC/FID chromatogram of n-alkane standards ...................................................................... 36
7.2 GUFID chromatogram of surrogate recovery standards ...................................................... 36
7.3 GCYFID chromatogram of TEM fraction .............................................................................. 37
7.4 GC/FID chromatogram of HEX fraction .............................................................................. 37
7.5 GC/FID chromatogram of TSAT fraction.. ........................................................................... 38
7.6 GUFID chromatogram of TAR0 fraction ............................................................................ 38
7.7 GUFID chromatogram of TPOL fraction ............................................................................. 39
8.1 WSO chemical character versus percent water cut ............................................................... 40
8.2 WSO carbon content versus percent water cut ...................................................................... 40
8.3 WSO & vs percent water cut ................................................................................................ 41
8.4 WSO IQ vs GOM simulant volume ...................................................................................... 41
8.5 Total dissolved solids (as Cl-) as a function of chemical fraction ........................................ 42
8 . 6 Total dissolved solids (as Cl-) as a function of carbon content ............................................ 52
8.7 Chemical character as a function of GOM brine pH ............................................................. 43
8.8 Carbon content in TEM fractions as a function of GOM brine PH.. ..................................... 43
44 1
8.9 Chemical character as a function of GOM brine temperature.. .............................................
8.10 Carbon content in TEM fractions as a function of GOM brine temperature ........................ 44
8.11 Chemical character as a function of GOM brine pressure .................................................... 45
8.12 ‘Carbon content in TPOL as a function of GOM brine pressure ............................................ 45 v
8.13 Influence of percent water cut on organic acid content ........................................................ 46
8.14 Influence of brine pH on organic acid content ...................................................................... 46
8.15 Influence of salinity on organic acid content ........................................................................ 47
8.16 Influence of brine temperature on organic acid content ........................................................ 47
8.17 Ion chromatogram of organic acids in produced water ......................................................... 48

vi
 LIST OF ABREVIATIOFS

DRO Diesel range organic compounds

GC Gas chromatography
GOM Gulf of Mexico
FID Flame ionization detector
GRO Gasoline range organic compounds
HEX Hexane fraction
HP Hewlett Packard
IC Ion chromatography
ICP-AES Inductively coupled plasma-atomic emission spectrometry
IR Infrared
ISA Ionic adjustor solution
ISE Ion selective electrode
K-D Kudema-Danish
LC Liquid chromatography
MS Matrix spike sample
MSD Mass selective detector
ORNL Oak Ridge National Laboratory
PAH Polynuclear aromatic hydrocarbons
PERF Petroleum Environmental Research Forum
PI Principal investigator
PPm Parts per million (by weight or volume)
svocs Semi-volatile organic compounds
SW Statement of Work
TAR0 Total aromatic fraction
TDM Total dissolved material
TEM Total extractable material
TPH Total petroleum hydrocarbons
TPOL(gc) Total polar fraction determined by gas chromatographic analysis
TPOL( sub) Total polar fraction determined by subtraction from TEM fraction data
TSAT s Total saturate fraction
USEPA United States Environmental Protection Agency
WAF . Water accumulated fraction
wso Water soluble organic compounds

vii
 .

.
 EXECUTIVE SUMMARY

Soluble organics in produced water and refinery effluents represent treatment problems for the petroleum
industry. Neither the chemistry involved in the production of soluble organics nor the impact of these
chemicals on total effluent toxicity is well understood. The U.S. Department of Energy provides funding
for Oak Ridge National Laboratory (ORNL) to support a collaborative project with Shell, Chevron,
Phillips, and Statoil entitled “Petroleum and Environmental Research Forum project (PERF 9844:
“Manage Water-Soluble Organics in Produced Water”). The goal of this project, which involves
characterization and evaluation of these water-soluble compounds, is aimed at reducing the future
production of such contaminants.

To determine the effect that various drilling conditions might have on water-soluble organics (WSO)
content in produced water, a simulated brine water containing the principal inorganic components
normally found in Gulf of Mexico (GOM) brine sources was prepared. The GOM simulant was then
contacted with as-received crude oil from a deep well site to study the effects of water cut, produced-
water pH, salinity, pressure, temperature, and crude oil sources on the type and content of the WSO in
produced water. The identities of individual semivolatile organic compounds (SVOCs) were determined
in all as-received crude and actual produced water samples using standard USEPA Method (8270C)
protocol. These analyses were supplemented with the more general measurements of total petroleum
hydrocarbon (TPH) content in the gas (C&ra), diesel (C&C&, and oil (C&C& carbon ranges as
determined by both gas chromatographic (GC) and infrared (IR) analyses. An open liquid
chromatographic procedure was also used to differentiate the saturated hydrocarbon, aromatic
hydrocarbon, and polar components within the extractable TPH. Inorganic constituents in the produced
water were analyzed by ion-selective electrodes and inductively coupled plasma (ICP)-atomic emission
spectrometry @ES).

The WSO found in produced water samples was primarily polar in nature and distributed between the low
and midrange carbon ranges. Typical levels of total extractable material (TEM) was about 20 mg/L; that
associated with the aromatic fraction was present at 0.2 mg/L and that in the saturated hydrocarbon
fraction was present at less than 0.02 mg/L. Formic, acetic, and propionic acids were also found in the
produced water, occurring at a total concentration of 30 mg/L. It was estimated that the presence of 30
mg/L organic acids would artificially overstate TEM content by 2 mg/L.

Of the five tested parameters, the factor that most controlled the total WSO in produced water was that of
aqueous phase pH. Beyond a value of pH7 significant quantities of Cra-&o range material become
markedly soluble as they deprotonate in a basic aqueous phase. Both the absolute and relative volumes of
GOM brine and crude additionally affected total WSO. Produced water appeared to reach a saturation
level of WSO at a.50% water/oil ratio. Pressure slightly enhanced WSO by increasing the relative
quantity of C&lo range material. Temperature primarily altered the relative ratio of carbon ranges
within the WSO without significantly elevating the total WSO in the GOM brine. Salinity had the least
affect on the chemical character or the carbon size of WSO in produced water.

ix
 .

.
 1. BACKGROUND

Soluble organics in produced water and refinery effluents represent treatment problems for the
petroleum industry. Production insta!lations and refineries take special efforts to meet
increasingly stringent regulatory discharge requirements for dissolved organics. Early data from
the Gulf of Mexico (GOM) wells indicate that the more polar hydrocarbons from deep-water
drilling will significantly increase the dissolved hydrocarbon content in produced water from
these sites. Neither the chemistry involved in the production of soluble organics nor the impact
of these chemicals on total effluent toxicity is well understood.

The U.S. Department of Energy provides funding for Oak Ridge National Laboratory (ORNL) to
support a collaborative project with Shell, Chevron, Phillips, and Statoil, entitled “Petroleum and
Environmental Research Forum Project (PERF 98-04: “Manage Water-Soluble Organics in
Produced Water”). The goal of this project, which involves characterization and evaluation of
these water-soluble compounds, is aimed at reducing the future production of such contaminants.
The information obtained in the project will increase our understanding of the generation of
water-soluble organics (WSO) and ultimately result in strategies to minimize WSO production
and lead to the development of guidelines for effluent treatment. This, information will,be,used to
design the next generation of offshore platforms for deep-water wells; and could also help reduce
construction costs.

The first phase of the task was to acquire characterization data for produced water. A subsequent
task will be to use the data to construct a model that can be used to predict WSO content based on
the composition of a specific crude source. This report describes ‘results acquired for phase one
of the task. Analysis of produced water was performed at both O,RI$L,~an,d Statoil Company.
ORNL characterized WSO derived from sources of GOM crude, whereas Statoil accumulated
characterization data from North Sea oil sources. The objective of this report is to provide the
sponsor of the project with a description of the analytical procedures used and the data acquired
by ORNL. The format of the document follows closely the document submitted by staff at
Statoil. This was done in order to aid the customer in compiling data from both ORNL and
Statoil studies. The format of the report is designed to:

> 1) briefly describe the scope of the study (Chapter 1);

2) report the sampling of the crude by industrial partners (Chapter 2) ;
3) provide a detailed discussion of analytical procedures (Chapters 3-6),
4) validate analytical results through a discussion of QC data (Chapter 7); and, finally,
5) present WSO characterization data derived from the contact ofthe GOMcrude with
brine (Chapters 8,9 and appendix tables).

Because the primary objective of the project was to provide,a,database from which a second party
would use to design a mathematical model to predict WSd content in produced water, the
pertinent data was set aside as a separate group of tables in the appendix. All tables within the
report text summarize information concerning objectives, analytical procedures, QC performance,
results non-essential to the data base (although included as points of interest), etc. This approach
in format should facilitate modeling by obviously separating types of data and clearly defining
information that will be incorporated into the predictive model.

Finally, several limitations were placed on the format this document due to information he!d
proprietary by the PERF committee. The well site from which the crude was derived is not
specifically named; it is only stated as Deep 1 Well. This report tabulates WSO data information
specific to this well site. The Deep 1 Well data amplifies a proprietary report written for the
PERP members that summarizes general levels and character of WSO found in national and
international literature. Because the proprietary nature of the summary report and its availability
to PERP committee members, no effort was made in this document to compare Deep 1 Well
results to general trends. Comparisons were made in closed session with PERF committee .
members. The primary goal of this document is data compilation for the project.

1.1 ANALYTICAL APPROACH *

Quantitative characterization data were needed as the first step in understanding the generation of
WSO in produced water. To achieve this objective, industrial partners provided ORNL samples
of GOM deep-water crudes and their associated produced water. Together with industrial
partners, ORNL selected methods to characterize the WSO in both actual and simulated samples
of produced water. The identities of individual semi-volatile organic compounds (SVOCs) were
determined in all as-received crude and actual produced water samples using standard USEPA
Method (8270C) protocol. These analyses were supplemented with the more general
measurements of total petroleum hydrocarbon (TPH) content in the gas (C&r,& diesel
(C1&&), and oil (C,&& carbon ranges as determined by both GC (USEPA SW-846 Method
8015B) and infrared (IR) analyses. An open liquid chromatographic procedure was used to
differentiate the saturated hydrocarbon, aromatic hydrocarbon, and polar WSO components
contained in the TPH extract. Inorganic constituents in the produced water were analyzed by ion-
selective electrodes and inductively coupled plasma (ICP)-atomic emission spectrometry @ES).

Since the methods primarily follow those established by the U.S. Environmental Protection
Agency (USEPA), the terminology used in this report follows that of USEPA regulatory
nomenclature. The stated carbon ranges (C&&, Ci,,+o, C&&s) refer to sample components
eluting from a gas chromatograph between n-alkane standards within each of these carbon size
ranges. The terminology does not define the identity of the sample component eluting, merely
that it has a volatility within the given saturated alkane range. The regulatory nomenclature is
specific to the USEPA method used, in this case USEPA SW-846 Method 8015B. A description
of the analytical procedures is presented in chapters 3 through 6 to fully explain under what
condition the data was acquired.

TPH is also regulatory nomenclature, referring to all components extracted using either USEPA
SW-846 Method 8015B or USEPA SW-846 Method 1664. Since polar WSO is also extracted in
these procedures, TPH results include some mass that is actually not petroleum hydrocarbons.
One of the goals of this study was to estimate the positive bias introduced into TPH results from
these non-TPH components.

1.2 EXPERIMENTAL APPROACH

Only a limited volume of produced water was present in the crude oil samples shipped from
drilling sites. To determine the effect that various drilling conditions might have on WSO content
in produced water, a simulated brine water containing the principal inorganic components
normally found in GOM brine sources was prepared. The GOM simulant was then contacted with
as-received crude oil to study the effects of water cut, produced-water pH, salinity, pressure,
temperature, and crude oil sources on the type and content of the WSO in produced water. The
test plan involved the analysis of the WSO content of simulated produced-water samples derived
from oil/brine contacts covering the following operating ranges:

2
 l pressures of l-100 bar;
l temperatures of 90-2OOpF (25-100°C);
* pH level of 6.5 z!z 2;
* water cuts of 20,50, and 80%; and
0 salinities of 35,000-150,000 ppm.

The test matrix presented in Table 1.1 below was used for the Deep1 crude ail source. The
specific name for each experiment associated with a given experimental variable is denoted as
“PEW-“ in the table. A complete description of each PEW experiment and its associated data
are summarized in the appendix of this document.

Table 1.1. Experimental test matrix fqr crude oiVGQh!J brine contacts
Temperature Produced Hz0 Salinity Water cut
Experimental variable Pressure (“Cl source PH (ppm> (%)

As received Ambient 25 GOM simulant 7 63,000 80

(PERIL8, 16,20)
, . _
Water cut Ambient 25 GOM simulant 7 63,000 67
(PERP-6)

Water cut Ambient 2 5 GOM simulant 7 63,000 50

t (PEW-7, 10)

Water cut Ambient 25 GOM simulant 7 63,000 20

(PERK9)

Ambient 25 GOM simulant 8.1 63,000 80

(PEE-13)

Ambient 2 5 GGM simulant 9.0 63,000 80

(PEE-1 1)

PH Ambient 2 5 GOM simulant 4.7 63,000 80

(PERP-12)

PH Ambient 2 5 GOM simulant 6.0 63,000. 80

(PER%19)

Salinity Ambient 25 GGM simulant 7 41,000 80

(PERP-14)
Ambient ’ .‘. 2j ‘- : -’ G6G si...ui;nt ’ 7‘
Sal&y 115,000 80
(PERK15)

Temperature Ambient 47 GGM simulant 7 63,000 80

c (PERI-17)

Temperature Ambient 75 GOM simulant 7 63,000 80

(PER&l 8)
~

3
 Table 1.1. Experimental test matrix for crude oil/GOM brine contacts (cont.)

Pressure 14 bar 50 GOM simulant 7 63,000 80

(PERF-26) *
Pressure 35 bar 50 GGM simulant 7 63,000 80
(PERF-27)
1
Pressure 60 bar 50 GOM simulant 7 63,000 80
(PEW-28)

The distribution and partitioning of WSO components were then calculated for the operating
ranges. The resulting information was forwarded to industrial partners for incorporation into a
model to predict the production of water-soluble organics as a function of crude composition and
formation characteristics.

2. COLLECTION AND SAMPLING OF ACTUAL OIL AND

PRODUCED WATER SAMPLES

2.1 FIELD SAMPLING OF DEEP1 CRUDE OIL

Two 5-gallon samples were collected directly from the Deep1 Well flowline; these samples were
transferred directly into 5-gal DOT-approved containers. According to field notes, the crude oil
had a 22.2 API gravity at a collection temperature of 75°F. Each sample was collected slowly
over a l-h interval to allow the oil to degas. While the samples were being collected, the
operators performed a shakeout test, which gave a 10% water cut when heated with solvent at
140°F.

2.2 DETERMINATION OF WATER CONTENT IN AS-RECEIVED OIL SAMPLES

Crude oil samples were shipped from the drilling site in 5-gal barrels equipped with two bung-
hole ports in the lid. The ports were opened slowly to vent the slight overpressure of the barrels.
Any actual produced water was collected by dropping weighted tubing (Tygon@ F40440-A)
down to the bottom of the slightly tipped barrel. A peristaltic pump, operating in the forward
direction, was used to create a slight positive pressure to displace any oil from the tubing prior to
withdrawing the water phase. Pumping was continued until all water had been collected; the
volume of water was then estimated relative to the initial volume of sample present in the barrel.

In the case of Deep 1 crude, a separate aqueous phase could not be removed directly, so 0.5 to 1 L
of oil was withdrawn from the bottom of the. barrel and allowed to settle in a glass bottle. Again a
separate aqueous phase did not form. Ten-mL aliquots of withdrawn oil were then transferred to
15-mL tapered centrifuge tubes. The oil was then acidified to 0.001 M HCl by the addition of 10
& of 1 M HCI to separate any produced water that might have emulsified during shipment.
When an aqueous phase was not observed at the base of the tube after 15 min, the acidified oil
samples were centrifuged at 2000 r-pm for 20 min to enhance phase separation. No distinct water
.phase was noted at the base of the tube, indicating the water content was too low to visually
measure. Results acquired using an automated Karl Fischer iodometric titrator indicated 8.9f 0.4
% water content, as opposed to field testing results of 10-12 %.

4
2.3 LABORATORY SAMPLING ‘OF AS-RECEIVED CRUDE OIL.
._ >.._ /’ .” ,... 4
Consistent testing of the WSO contents in the oil/water mixing experiments was predicated on the
use of homogeneous oil samples. Toward this end, a 5-gal barrel of oil was placed horizontally
on a barrel roller and rotated at the rate of 8 rpm for 2 h. The barrel was then set upright and oil
was withdrawn using submersed tubing and a peristaltic pump. No entrained air bubbles were
observed in the withdrawn oil.

3. EXPERiMENTAL PROTO& FOR Alt&IE~ PRESSlJ@$ .TlZfj’I’I,WG _ _.

,
Simulated produced water samples were prepared under ambient pressure conditions using a
Water Accumulated Fraction (WAF) vessel. The WAF procedure is typically used for the aquatic
toxicity testing of acute and chronic exposures of marine life to lubricants in ,contact with
^ ^water
.,.
sources. The glass vessel was fabricated according to dimensions stipulated in ASTM D6081, in
which the height:diameter ratio of the liquid phases in the 5-L vessel can be varied from 1: 1 to
2: 1. Testing was initiated by first adding up to 3 L of GOM simulant to the vessel. Up to 4 L of
oil were slowly drained down the side of the vesselsothat it.,w,as-layered over the water without
undo turbulence. Typical water/oil ratios used in testing varied from 2O-80%. A baseline value
of 80% was used for testing all experimental variables except water cut. A TeflonB-coated
stirring bar submersed in the GOM simulant was used to stir the,oil/water interface such that the
vortex of the oil phase descended to a depth of approximately 15-30% into the aqueous phase.
Although the ASTM’protocol indicates that a 24-h mixing time is sufficient to reach equilibrium
extraction of WSO, a mixing time of 4 days was typically used to duplicate the experimental
setup employed by Statoil in the testing of North Sea oil samples. Produced water was collected
at the end of the equilibration time by opening a Teflon@ stopcock located at the base of the
WAF vessel. If necessary, any dispersed oil in the withdrawn water was removed using a
separatory funnel. Several 20-mL aliquots of the produced water were stored in glass bottles at
-4°C for subsequent analysis of carbonate/bicarbonate, inorganic metals, chloride, and low-
molecular weight organic acids. One liter of produced water was then acidified with concentrated
hydrochloric acid (HCI) to bring the pH level to less than 2. Surrogate recovery standards were
added to the produced water fraction prior to the solvent extraction of WSO.

4. EXPERIMENTAL PitOTOCOL ~~~~~IQZ.J’&EjWJRE TESTING

_:,
A factory-assembled pressure vessel incorporating armored sight glasses on opposing walls, was
used to generate produced water samples at pressures up to 1000 psi and 50°C. This vessel
having an internal volume of 300 mL, is fabricated of mild steel, that has been certified as to
chemical composition, hardness, and tensile strength. Additional hardware supplied by Pressure
Products Company, Inc. (4540 W. Washington St., Charleston, W. VA 25313) included a 1200
psi pressure relief valve, a l-1000 psi liquid-filled pressure gauge, an inlet for a digital
thermometer, an inlet to pressurize the vessel with helium, and tubing and needle valves to allow
filling and draining of solution from the vessel (Figs. 4.1 and 4.2). The 60-lb pressure vessel was
clamped onto a RoTTap Model B Sieve Shaker (C. M. Tyler Co.), which rotated the system
horizontally at approximately 60 rpm to produce a slight vortex at the oil/water solution interface.
During mixing, the temperature was maintained at 50°C by controlling the voltage output to a
heating tape wrapped around the circumference of the vessel.

Two hundred mL of brine simulant were,added to the heated vessel; the system was then
pressurized with helium, heated, and mixing for 4 days to determine blank concentration levels
for TPH in the vessel. A HokeB sample chamber was connected to the gas sampling port of the

5
vessel. at the end of the equilibration time. Pressure was vented into the evacuated gas sample
vessel by opening the gas sampling valve located at the top of the pressure chamber. Once the
chamber had been brought to ambient pressure, the HokeB vessel was sealed. The exit valve
located at the base of the pressure chamber was subsequently opened to collect the water phase.

The gases that collected in the HokeB sampling vessel were vented through an -5 cm length of
V&inch stainless steel tube packed with sorbents to retain hydrocarbon components and then
through a mass flow meter. The tube contained Carbotrap C to capture high-molecular-weight
material, followed by Carbotrap for the majority of components and, finally, ‘Carbosieve SIII for
low (C&s) molecular-weight organics. These sample tubes have been sealed and archived
should the need for their analysis exist.

Once the TPH blank levels of pressure vessel were suitably low, actual oil/brine experiments
were conducted by loading the heated chamber with 200 mL of GOM brine simulant and 30 mL
of crude oil. The first experiment was performed at ambient pressure and 50°C for comparison
with data derived from oil/water contacts in the WAF vessel. The remaining experiments were
conducted at pressures of 14,35, and 60 bar and a temperature of 50°C. Again, the gases in the
chamber were vented into a gas sampling vessel to depressurize the pressure chamber at the end
of a 4-day equilibration time. Immediately after being depressurized, the produced water phase
was collected at the drain port for subsequent analysis using the open LC fractionation protocol
(see Sect. 6.4).

5. SYNTHETIC GOM BRINE

1
5.1 PREPARATIOrJ OF SIMULANT

The staff of the Phillips Company provided the recipe for the preparation of GOM brine simulant.
Table 5.1 summarizes the average concentrations of actual GOM brine components as derived
from data members of Phillips staff members have collected from 300 past and present GOM
wells. These accumulated data provided the basis for the recipe formulation.

Table 5.1. Average contientrations of

GOM brine componentsa
GOM brine Concentration
Component (mg/L)
cl- 63,000
so4= 3
HCOs- 220
Na+ 35,354
Ca’+ 3,260
Mg2+ 9 1 0
Ba2’ 95
Sr2+ 131
Fe3+ 23
Dissolved CO2 68
________________________________________---
Temperature 90°F
pH 6.7
aData provided by D. Bourg, Philips Co.,
May 9,200O.

.
Table 5.2 summarizes the results of a spreadsheet devised by Phillips staff that delineates the
chemical components and their quantities required to prepare 1- and 4-L batches of produced
water simulant. Once prepared, the simulant was stored in a refrigerator and then brought to the
required experimental temperature just before use.
.__. . ,.
5.2 VERIFICATION OF GOM BRINE SIMULANT CONTENT
,. .
Ion chromatography (IC) and ion-selective electrodes were used to determine.anion
concentrations in the prepared simulant brine. ICP-AES was used to verify dissolved metal
components. Complete descriptions of the analytical procedures used for cation/anion
determinations can be found in Sects. 6.6 and 6.7 of this document. IC was initially selected for
the analysis of brine salinity; however, results based on IC appeared to negatively bias the total
chloride concentration in GOM brine simulant. The high sodium concentration in the simulant
mandated that the brine samples be chemically treated with a cation exchange resin column to
reduce the level of sodium ion before the sample was introduced into IC equipment. As can be
seen in Table 5.3, significant error was introduced by the pretreatment procedure since the
electrochemical balance of the cation content in the simulant (Table 5.4) indicated that the
chloride content should be approximately 65,000 ppm. Chloride analysis based on the use of a
chloride selective electrode was considered more accurate for this application because this
analysis is typically performed in the presence of 5 M NaN03, a matrix having an in ionic
strength similar to that of the simulant brine.

Ferrous ion was initially included at a concentration of 23 ppm to reflect the concentration
present in actual GOM brine. However, precipitation of ferric oxide at near neutral pH required
filtration of the simulant. prior to use. The final soluble iron concentrat,io,n.in the filtrate,was less
than the detection’&& of the ICP. Therefore, ferrous sulfate was not included in simulant
preparation for any of the oil/water contact experiments in this study.

Table 5.2. Preparation of Gulf of Mexico brine. simulant .*

l-L Prepircii6ti“-4iL prepi-fi.6i
^.,‘. - ” .I. I.
Component
requires (g) Requires (g)
_^_
NaCl 89.708 358.832
CaCl2 l 2H20 11.959 47.835 ’
MgClz .6HzO 7.613 30.453
BaC12l 2HzO 0 . 1 6 9 0.676
SrCl2 l 6H& 0 . 3 9 9 1.595
FeS04 l 7H20 0.000 ’ 0.000
NaHC03 0 . 3 0 3 1.212
Na2SOd 0.063 .- 0.252
,._
Hz0 889.740 3,559.0

Final solution (g) 1,000.0 3,999.8

Total dissolved material 102,971 102,97 1
(TDM ppm)
Salinity as Cl- (ppm) 62,975 ’ 62,975
._

7
 Table 5.3. Anion concentrations in GOM brine

.
Simulant Analysis Results (ppm)
Actual GOM brine Ion Ion Selective
Anion concentration (ppm) Chromatography Electrode .
so4 2- 3 <lO NAa
Cl- 62,975 41,500 62,600 f 1,700
HC03- 220 NA 153 Z!Z 1
aNA = not analyzed.

Table 5.4. Cation concentrations (mg/L) in GOM brine

Cation (mg/L)
Na+ Ca2’ K+ Sr2’ Ba2+ Fe3+
Actual brine 35,400 3,260 0.0 0.131 0.095 0.023
Simulant by ICP 35,400 3,270 <0.004 0.139 0 . 0 4 4 <0.020

6. ANALYTICAL METHODS

6.1 IDENTIFICATION OF SEMI-VOLATILE ORGANIC COMPOUNDS (SVOC) IN

ACTUAL OIL AND PRODUCED WATER BY GC/MS

Specific chemical compounds were identified and quantitated in as-received crude oils and actual
produced water using USEPA SW-846 Method 8270 C. (This method is also identified as
USEPA Method 525.2 of the “500 series” in the USEPA compendium for the analysis of
contaminants is drinking water.) The protocol uses GC-MS to identify a core group of over 45
semi-volatile organic target compounds in environmental samples. Within the list of target
compounds, those that are of particular interest to petrochemical applications are included in
Table 6.1:

Table 6.1. USEPA Semi-volatile organic compounds

of interest to petrochemical applications
Acenaphthene Fluoranthene
Acenaphthylene Fluorene
Anthracene Indeno( 1,2,3-c,d)pyrene
Benzo(a)anthacene 2-Methylnaphthalene
Benzo(a)pyrene 2-Methyl phenol
Benzo(b)fluoranthene Naphthalene
Benzo(e)pyrene Phenanthrene
Benzo(k)fluoranthene P yrene
Chrysene

Although polynuclear aromatic hydrocarbons (PAHs) are included in the list due to the toxicity
imparted by these compounds to environmental samples, the PERF activity reported here did not

8
address toxicity concerns as part of its scope of work. Therefore, PAH data were not acquired in
samples derived from GOM wells.

Crude oil and actual produced water samples were analyzed by Mountain States Analytical, Inc.
(1645 West 2200 South, ,Alt Lake City, Utah 84119). The protocol from this off-site testing
facility includes the analyses of over 120 semi-volatile compounds pertinent to the petroleum
industry. Protocol requirements include the addition of surrogate recovery standards (see Table
6.2) to samples prior to their extraction with methylene chloride. These standards are used to
monitor the variability in extraction efficiency of the solvent for individual target compounds.

Tgble 6.2. Surrogate recovery

standards for SVOC analysis
2,4,6-Tribromophenol
2-Fluorophenol
2-Fluorobiphenyl
Nitrobenzene-d5
Phenol-d6
Terphenyl-dl4
..^ . . . . ,.

Neat oil samples were initially diluted by a factor of 25 with methylene chloride prior to GC-MS
analysis. Produced water samples were acidified to pH 2 prior to extraction of WSOs.. Three 30-
mL contacts with methylene chloride were then used to serially extract organic compounds from
200 to 1000 mL of produced water in a separatory funnel. Internal standards were also added
after solvent extraction as a means of checking the GC-MS instrument sensitivity and linearity.
These compounds are listed in Table 6.3.
Table 6.3. Internal st&@rds
for SVO compound analysis
1,4-Dichlorobenzene
Naphthalene-d8
Acenaphthene-dl0
,‘,I Phenanthrene-dl0
Chrysene-dl2
Perylene-dl2

The combined extracts were dried with sodium sulfate and concentrated
.‘:” ..r.it*.,,iwI. L. I.) ,*,*,.I-, I .i ..i$Cj.. to1
. . . . . a_mL. A laboratory
.&Qr”*i *134*pi C‘, “‘4
control spike standard provided an additional quality control check. The spike solution
containing the compounds listed in Table 6.4, was added to an aliquot of the concentrated
methylene chloride extract to check for the interference of sample matrix components with the
quantitation of individual target compounds.

Ttiblc 6:4. Compounds included in

the laboratory control spike standard
Acenaphthene N-Nitroso&N-propylamine
4-Chloro-3-methylphenol Pentachlorophenol
2-Chlorophenol Phenol
1,4-Dichlorobenzene P yrene
2,4-Dinitrotoluene 4-Nitrophenol
1,2,4-Trichlorobenzene

9
Prepared methylene chloride extracts were then injected into a GC-MS system (operated under
the conditions listed in Table 6.5) to identify and quantitate individual SVOCs. -

Table 6.5. GC-MS conditions for SVOC analysis

Instrumentation: Hewlett Packard (HP) 6890 plus Gas Chromatograph with a 5973 MSD
GC column: 30 m X 0.25 mm (or 0.32 mm) ID 1-ym-film-thickness silicone-coated
fused-silica capillary column
Software: HP MSD ChemStation with EnviroQuant
Ionization: Electron impact, 70 eV
Ion source 230°C
temperature:
Carrier gas Helium, 30 cm/s
Initial pressure: 16 psi
Inlet: Pulsed splitless, initial temperature 270 “C
Oven program: 50 ‘C (4 mm), 10 “C/mm to 350 OC, hold until dibenz(a,e)pyrene elutes
Sample Volume: 1 P-L

6.2 GC/FID ANALYSIS OF TOTAL PETROLEUM HYDROCARBONS

A modification of SW-846 Method 8015B was used to determine the concentrations of

nonhalogenated volatile organic compounds and semivolatile organic compounds by GC using
flame ionization detection (FID). This method analyzes total petroleum hydrocarbons (TPH) in
specific carbon ranges defined as

l TPH-Gasoline (GRO): C&10,

l TPH-Diesel (DRO): Cl&~, and
l TPH-Oil: C&& .

TPH-GRO represent alkanes having a boiling point range of 60-170°C; the remaining
compounds correspond to alkanes having a boiling point range of 170-43O”C. Prior to solvent
extraction, the surrogate recovery standards (n-dodecane-d26, naphthalene-d8, and n-phenol-d6
or tribromophenol) were added at a level of 40 pg each to produced water in order to estimate the
variability in extraction efficiency of alkanes, aromatic, and polar compounds, respectively. The
concentrated solvent extracts containing WSO were then injected into a gas chromatograph using
the operating conditions summarized in Table 6.6.

Table 6.6. GC/FID operating conditions for TPH analysis

Instrumentation: HP 5890 Series II Gas Chromatograph with an FID
GC column: 30 m X 0.53 mm ID fused silica capillary column bonded with 5%
methyl silicone (HP-5), 0.88~pm film thickness
Software: HP ChemStation version 3365 Series II
Carrier gas flow rate: Helium, 4 mL/min
Makeup gas flow rate: Helium, 26 nL/min
Injector temperature: 200°C
FID temperature: 340°c
Hydrogen flow rate: 30 r&/mm
Air flow rate: 375 mL/min

10
 /

Initial temperature: 45°C (hold 3 mm)

Program: 45’C to 275°C at 12°CImin
Final temperature: Hold 12 mm
Sample size: 0.5-3 pL .,.. .l...“~j ..,. *>, ‘e.*(..*. ^I *, ,“‘, a..*.. ,. .w*+n “,*W^r. r.*/.*w .4.m*,.,I” .“4.,, i , 8..
Detector response was externally standardized using a six-point calibration curve. The
calibration data set was constructed from dilutions of a certified standard (Absolute Standards,
catalog no. 908 14, lot 030700) containing 2000 ug/mL each of n-alkanes spanning the range Ce-
Czs. The FID response was linear over the range of 0.01 - 0.1 pg injected mass of each alkane,
corresponding to a TPH range of 0.2-l .4 ~..@rnL for each alkane in produced water. HP Chem
Station@ software was used to integrate peak area over the three carbon ranges: C~-cta, C&C&,
and C20-C28 to obtain the TPH mass in each of the solvent extraction concentrates.
.?..,
6.3 INFRARED ANALYSIS OF-TOTAL RECOVERABLE PETROCHEMICAL
HYDROCARBONS

The analysis of WSO in produced water samples was also supplemented using USEPA Method
418.1. The infrared (lR) method is more commonly used in the field or as a screening technique
due to the simplicity of the procedure. As modified for this task, the protocol uses 10 p,L of 4 M
HCI to acidify 2 mL of produced water to a value less than pH 2. The oil-and-grease components
are then extracted from the treated sample with 3 mL of LB-grade tetrachloroethylene. The
organic phase is subsequently transferred to a l-cm path length IR cell. The absorbance at a
wavelength of 3.4 pm is measured on a single-beam IR spectrometer. Sets of oil-and-grease
standards are prepared in the working range of 4 to 40 mg in tetrachloroethylene using either
GOM’crude oil or a synthetic oil 1375(v/v) % n-hexadecane, 37.5 (v/v) % isooctane, and 25
(v/v) % chlorobenzene). A three milliliter volume of oil standard is acidified; 2 mL of 100 g/L
NaCl is then added to match the salinity content of GGM brine simulant. After the shaking and
phase separation steps, the solvent phase i‘s withdrawn as the working‘oil-and-grease standard.

6.4 FRi+CTIONATION OF WSO IN SIMULATED PRODUCED WATER

A more definitive approach to characterizing the WSO content in produced water samples is to
fractionate compounds into primary classes of saturated hydrocarbon, aromatic hydrocarbon, and
polar components. The TPH content of each fraction is then determined by GC/FID analysis. An
open liquid chromatographic (LC) column was used to perform the fractionation of extractable
TPH using a procedure similar to that described by Mills et al. In this procedure, a gross
measurement of organic content is made by extracting total TPH from an aqueous sample with
methylene chloride. The extract is then applied to a multi-sorbent column to separate the
chemical fractions. The advantage of such an analytical approach is that the chemical character
of WSO can be determined inexpensively as a function of general classes for a sample containing
a myriad- of organic constituents.

The procedure of Mills et al. was modified to accommodate the analytical requirements of the
PERF 98-04 project. A primary change in the protocol was to replace the gravimetric analysis of
column fractions with that of GC/FID analysis based on USEPA SW-846 Method 8015B.
Gravimetric analysis did not provide the necessary detection limits and compositional
information considered relevant to the project. A secondary procedural change was to estimate
polar hydrocarbon content by eluting the bulk of the material from the open LC column with a

11
final methanol wash. The distribution of carbon content within molecular size ranges could then
be estimated for this fraction by GC/FID analysis. The GUFID data for the polar fraction were
then compared the method by Mills et al. The organic mass of the polar fraction can also be
estimated by subtracting the combined mass of saturated and aromatic fractions from the total
TPH content originally present in the methylene chloride extract. Finally, the solvent exchange
procedure used to remove methylene chloride and transfer total TPH to a hexane matrix was
altered midway through the study to minimize the loss of more volatile WSO components.
Rather than evaporation at 65’C in Kudema-Danish glassware, the methylene chloride solvent
was heated to 35°C under a light flow of nitrogen gas to reduce the sample volume. Two
sequential additions of hexane were then added and concentrated in the same manner.

An outline of the modified procedure is reproduced in diagrammatic form in Fig. 6.1. Prior to
TPH extraction, aliquots of produced water were withdrawn for analyses of inorganic constituents
and low molecular weight organic acids. Surrogate recovery standards (n-hexanoic acid, iz-
dodecane-d26, naphthalene-d8, and n-phenol-d6 or tribromophenol) are then added to 1 L of
produced water. Finally, WSOs were then serially extracted with three 30-mL methylene
chloride contacts.

The methylene chloride solvent was concentrated to 2 mL at 65°C using Kudema-Danish (K-D)
evaporation glassware to yield the total extractable material (TEM) fraction. A 100~@ aliquot of
the TEM fraction was dried on a tared glass-fiber filter to estimate the initial’TPH mass in the
fraction gravimetrically (generally a few milligrams). A secondlOO-& aliquot of the TEM
fraction was placed’in a sealed ampoule for subsequent GUFID analysis.

In the first 14 oil/water contact experiments, 10 mL of hexane was added to 1-mL TEM aliquot
and evaporated at 65’C to a final volume of 2 mL in order to transfer TPH into a hexane matrix
(HEX fraction). This step was repeated to ensure that all methylene chloride had been removed
from the sample. Beginning with experiment PERF-15, solvent substitution was modified to
reduce organic losses at elevated temperatures. Instead, 2 mL of hexane was added to a l-r&
TEM fraction and evaporated to near dryness at 35’C under a light flow of nitrogen gas. This
step was repeated twice for complete solvent substitution. The residue was then brought to a 2-
mL volume with hexane. The hexane fraction, denoted as HEX, was refrigerated overnight to
precipitate asphaltenes. A loo-& aliquot of the HEX fraction was dried on a tared glass-fiber
filter to estimate the initial TPH mass in the fraction gravimetrically (generally a few milligrams).
A secondlOO+L aliquot of the HEX fraction was placed in a sealed ampoule for subsequent
GCIFID analysis.

The remainder of the clarified HEX fraction was added to the open LC column. The multilayered
column was packed sequentially with 0.5 cm of fired glass wool, 5 g of fired sand prewashed
with methanol, 7 g of fired activated alumina, 13.5 g of fired activated silica, 2 g of anhydrous
sodium sulfate, and a final top layer of 0.5 cm of fired glass wool (Fig. 6.2). The packed column
was prewashed with 50 nil., of methanol, 200 mL of methylene chloride, and finally 40 mL of
pentane before the HEX fraction was added. The saturated hydrocarbons from the applied HEX
fraction were eluted with 40 mL of pentane. This fraction was concentrated to 1 mL by
evaporation at 65’C in the K-D glassware, and is denoted as the TSAT fraction. The concentrated
TSAT fraction was transferred to a sealed ampoule for GC/FID analysis.

A 200 mL volume of 1: 1 pentane:methylene chloride was used to elute the aromatic compounds
from the LC column. This fraction, denoted as TARO, was also concentrated to 1 mL by

12
evaporation at 65°C. The concentrated TAR0 fraction was transferred to a sealed vial for
GCFID analysis.

The LC column was finally washed with 150~mLof methanol to displace the majority of the polar
compounds. This fraction, denoted as the TPOL, was then evaporated to 1 mL at 85°C. Each LC
fraction was analyzed for organic mass content by GUFID.
Produced
water
A liquots for
Acidify to pH 2 b inorganic
w i t h HCI b analysis I

Add ) A liquots for

surrogate recovery IR a n a l y s i s
standards ‘I
M ethylene
chloride
extraction

GC/FID a n a l y s i s o f Gravimetric analysis of

total extractable Concentrate total extractable
material (TEM) material (TEM)

Convert to
hexane matrix

I
+
G r a v i m e t r i c a n d GC/FID 6 R e f r i g e r a t e + Asphaltene
analysis of (HEX) 1 mL content

Separate on open liquid -

chrom atographic column

Elute total saturated

Elute total aromatic petroleum ____) petroleum hydrocarbons
hydrocarbons with I:1 w i t h p e n t a n e (TSAT)
pentane:methylene c h l o r i d e (TAR01

Elute polar materials with

m e t h a n o l (TPOL)

T o t a l p e t r o l e u m h y d r o c a r b o n s ( T P H ) = Total saturated (TSAT) + total aromatic (TARO)

+ total polar (TPOL) fractions
True TPH = Total saturated (TSAT) + total aromatic (TARO) fractions

Fig. 6.1. Fractionabon of”%& in &m&ted produced water.

Thus, the produced water sample is characterized’by total TPH content as determined from the
TEM fraction. Again, it is important to recognize that total TPH mass in actuality is a summation
of real petroleum hydrocarbons and water soluble organic constituents that are also extracted
using methylene chloride solvent. The HEX fraction contains total TpH that was transferred
from the methylene chloride matrix to the non-polar hexane fraction. Actual TPH is a
combination of saturated hydrocarbon mass found in the TSAT fraction and the aromatic
hydrocarbon mass found in the TAR0 fraction. The TPOL fraction contains the balance of
organic mass from the produced water that is not petroleum hydrocarbons, but was coextracted
when the water was extracted with methylene chloride.
 13.5 g activated silica

1 7 g activated alumina

7 5gfiredsand

.-I 0.5 cm glass wool

Fig. 6.2. Open liquid chromatography column for fractionation of TPH .

Column Preparation and Sample Elution

1. Pack the sorbents using a methanol solvent.
2. Once packed, elute the column successively with 50 mL of methanol, 200 mL of
methylene chloride, and 40 mL of pentane.
3. Add the HEX sample to the top of the column. Rinse with several
1 mL pentane additions to transfer the sample to the head of the column.
4. Elute the TSAT fraction with 40 mL of pentane at 1 mL/min.
5. Rinse the K-D tube with 1: 1 methylene chloride:pentane and transfer rinsates to the
head of the column. Elute the TAR0 fraction with 200 mL 1: 1 methylene chloride
:pentane.
6. Rinse the K-D tube with methanol and transfer the rinsates to the head of the
column. Elute the TPOL fraction with 150 mL of methanol.

14
 6.5 DETERMINiTIOii OF WiiiiR tibi’iiti’i Iii CRUDE OIL

The water content of as-received crude oil was determined by the standard Karl Fischer titration,
using a Brinkman Model 652 KF-Cot&meter, ,,In this procedure, water reacts with iodine in the
presence of an alkyl sulfurous acid and imidazole base to form the corresponding alkyl sulfuric
f acid. Iodine is generated at a constant rate by anodic oxidation of iodide present in the titration
medium. When all the water present in the sample has reacted, excess iodine is indicated
coulometrically. The total current multiplied by the time required to reach the titration end. point
is proportional to the water content of the sample. The anode compartment of the coulometric
cell was filled with AQUASTAR@ Coulomat A (EM Science) anode generator solution. The
cathode compartment was filled to a level of 2-3 mm below that of the anode generator solution
with AQUASTARe Coulomat C (EM Science) counter solution. The instrument was then turned
on, and the cell matrix was conditioned to remove the trace water present in cell reagents. When
the instrument response no longer drifted, lo-50 FL of diluted oil was introduced into the
coulometric cell. Equivalent water introduced to the cell from the diluted oil samples was lOO-
300 ug. Instrument response was calibrated using a 1 mg/mL water standard
(HYDRANAL@Water Standard 1.00, Allied Signal).

6.6 DETERMINATION QF pH AND BICARBONATE/CARBONATE CONTENT IN

WATER :

A Metrohm 717,DMS Titrino automatic titrator ,rri41”r.~-s”a”Lu.-i,i*^.,

_<-- _ .a+, e. .l*-*.“*“dla-~e was used to determine the pH and the
.~*rr,,i&&~ka..z‘&
hydroxide/bicarbonate/carbonate content of produced water samples (Franson, 1992). The
instrument was standardized with two NIST-traceable buffer .., . I solutions
,*. I .* ,.._ (pH 7 and pH 10,
respectively). The temperature of the solution was entered digitally before the pH of the sample
l

was measured. The OH, COs*-, and HCO3 concentrations were determined by titrating 20 mL
of produced water with stanclard.o.o-~~N_HC End points measured at pH 8.3 (i.e., volume end
point A) and 3.7 (i.e., volume end point B) were then entered into the following calculations:

If 2A > B, the solution contains OH and COs*-,

(2A - B) x normality of HCl x l7.?073/(sample volume) = ppm OH,

2(B - A) x normality of HCl x 30.0046 /(sample volume, mL) = ppm COs*-.

If B > 2A, the solution contains COs*- and HCOJ-,

2A x normality of HCl x 30.0046 /(sample volume, mL) = ppm COs*-,

(B - 2A) x normality of HCl x 61.0171 /(sample volume, mL) = ppm HCO<.

Total alkalinity is determined from the total volume of acid required to achieve a pH of 3.7. It is
calculated as:

(A + B) x normalitv of HCl x 1000 = Alkalinity to pH 3.7, mg CaC03/L

sample volume, mL

15
6.7 DETERMINATION OF CHLORIDE CONTENT IN PRODUCED WATER

Ion chromatography (IC) and chloride ion-selective electrodes (ISE) were used to determine the
salinity in produced water. Performance of IC protocol followed that of USEPA Method 9056
“Determination of Inorganic Anions in Ion Chromatography”. Chloride analysis was performed
by the sequential elution of 0.01 .to 50 ppm of anions from an IonPac AJl 1 analytical column
using a gradient elution. The eluting anion concentration was detected with a temperature-
compensated electrical conductivity cell. The instrument was calibrated by the introduction of
standard anion solutions using IC operating conditions summarized in Table’6.7.

Table 6.7 IC operating conditions for anion analysis

Sample loop volume: 4mmID; lo&
Trap column: ATC-1
Guard column: IonPac@ AGl 1
Analytical Column: IonPac@ AJ 11
Eluents: El: Type I deionized water
E2: 5.0 mM NaOH
E3: 100 n&l NaOH
Eluent flow rate: 2.0 n-L/mm
Suppressor: Anion self-regenerating suppressor
Autosuppression recycle mode

The majority of the sodium in brine samples had to be removed prior to anion analysis by IC. To
accomplish this, a portion of the sample was passed through a hydrogen-form cation-exchange
resin. The acidified sample was then introduced to the IC.

Determination of the chloride concentration was more commonly achieved using an Orion Model
94-17B chloride electrode and an Orion Model 90-02 Double-Junction Reference Electrode in
combination with an Orion 520A pH/ISE meter. Response of the electrode system was calibrated
using NaCl standards (100-3000 ppm Cl-) in which an Ionic Adjustor Solution (ISA) of 5 M
NaN03 was added to modify the ionic strength of the solution. ISA was similarly added to a
1: 100 dilution of produced water to measure the chloride content.

6.8 DETERMINATION OF METAL CONTENT IN PRODUCED WATER

ICP-ABS was used to determine the trace-element content of the GOM brine simulant and the
produced water. All of the brine simulant samples were acidified to 10% HNO3 concentration
prior to analysis. In the case of both actual and simulated produced water samples, the dissolved
organic constituents were first wet-ashed using microwave-assisted acid digestion (USEPA
Method 3015). In this preparation technique, of 50 mL sample was digested in 5 mL
concentrated HN03 in a Teflon@ digestion vessel for 20 min using microwave heating. After the
sample has cooled, it was filtered prior to analysis.

A Thermo-Jan-e1 Ash 61E ICP-AES system was used to measure trace metal concentration
according to USEPA Method 601OB. Table 6.8 shows an example of instrument output for the
analysis of GOM brine simulant, which lists the metals that are quantified by the equipment, as
well as the instrument detection levels. Instrumental response and elemental emission correction
factors were established by analyzing blanks and certified mixed-element standards.

16
 Table 6.8. ICP inorganic analysis of GOM brine simulant using ICP-AES
Quantification “Raw data ” Sample
limit cont. cont. Analytical error
Metal
Silver CG&
0.002744 (mg/L)
0:008 (ma) (mk) (mg/L
().@jc)3’5 -.‘-~3;()0 (wU
l()o(-)()f& . .f0;87
.-‘I
Aluminum 0.046404 0.06 -0.01643 0.01023 60.0 k 25.4
Arsenic 0.008304 0.02 -0.00649
*‘) O.(rOQ92 20.0 CL 2.3
Boron 0.021937 0.05 -0.03 163 0.00b94 50.0‘. ” .- f 2.3
Barium 9.26842E 0.003 0.04387 0.00125 43.9 It 3.1
Beryllium 5.32885E 0.002 -0.Obo46 Q.00013 2.00 “. rt 0.32
Calcium 0.025085 0.075 3.3131 0.07225 3313, zk 179
Cadmium 6.34906E 0.002’ -0:00004 0 . 0 0 0 1 2.00 Itr 0.25
Chromium 0.00 1673 0.005 -0.00051 0.00015 5.00 zk 0.37
Cesium” 17.10689 20 -&6i237 3.15486 20,000 + 7838
Copper 0.005883 0.02 -0.00186 O.OpO76 20.0 f 1.9
Iron 0.113441 0.2 0 . 0 4 3 5 5 0.01663 200 +41
Potassium 0.023321 0.04 -0.06447 0.00535 40 +13
Magnesium 0.040085 0.05 0.883 16 0.02262 883 +56
Manganese 0.001003 0.003 -0.00038 0.00007 3.00 zk 0.17
Molybdenum 0.00407 1 0.01 -0.00042 0.00078 10.0 f 1.9
Sodium 0.092114 0.15 35.14379 0.84618 35,144 Ik 2102
Niobium 0.00273 1 0.008 d.0018 0 . 0 0 4 2 4 8.0 I!I 10.5
Nickel 0.006758 0.02 -0.00475 0.00034 20.0 2 0.8
Lead 0.002946 0.009 -0.00114 0.00136 9.0 f 3.4
Selenium 0.010864 0.03 -0.00189 0.00169 30.0 zk 4.2
Antimony 0.017710 0.04 -0.00373 0.00158 40.0 f 3.9
Silicon 0.029039 0.06 -0.0279 0.00489 60.0 I!z 12.1
Strontium 3.90035E 0.001 ,0.13402 0.00322 ,134.0 f8.0
Thorium 0.553 176 0.8 0.01249 0.09979 800.. ‘. +‘248
Titanium 6.18243E 0.002 -0.0013 .. 0.00002 zJ.60 kO.05’
Thallium 0.008663 0;025 ,.<..r;l..~.. 0 . 0 0 3 3 7 go
9~02 f8.4
o*073$58 ‘. o;ls”‘: ‘.j
Uranium 0.12958 ($06027 -‘180 ;.‘*iso
Vanadium 0.007891 0.02 1 ,;0:00026 0,00013 2o.y,~ 3j.q
Zinc 0.004324 0.01 -0.00797 o.ooo,lf3 10.00 +0.40
Zirconium ..0.03!5.18 . % _. .!A07 ___.~ .._ l!%!!Y ._, .P!?Wik _, .,,. .--._ 7 0 0 0 It13.9
a Cesium concentration overestimated due to the’interference of large quantities of sodium ion in
the sample.
,..

17
6.9 ANALYSIS OF C1-Cs ORGANIC ACIDS

Concentrations of organic acids were determined by injecting produced water directly into a
Dionex LC20 ion chromatograph fitted with an IonPac@ ICE-AS6 (Dionex Corp.) separation
column. The acids were separated from the high chloride sample matrix using a 0.4 mM
heptafluorobutyric acid eluent; a 1-ppm detection limit was typical. Instrument response was
calibrated using formic, acetic, propionic, malonic, and oxalic acids in the concentration range of
lo-50 ppm. Specific operating conditions are listed below.

Table 6.9 IC operating conditions for C&6 organic acid analysis

Sample loop volume: 0.037~mm ID X 9.75 in length; 50 lt.L
Analytical Column: IonPac@ ICE-AS6
Eluents: El: 1.8 mil4 carbonate/l.7 mM bicarbonate
E2: 0.1 M sodium hydroxide
E3: 0.4 mM heptafluorobutyric acid
E4: dionized water
Eluent flow rate: 1 .O mL/min
Suppressor: Anion micromembrane self-
regenerating suppressor-ICE II
Autosuppression regenerant mode

7. ASSESSMENT OF DATA QUALITY

7.1 CORRESPONDENCE BETWEEN ANALYTICAL METHODS FOR WSO

CONSTITUENTS
5
Correspondence testing of chemical procedures used to analyze the chemical contents of neat oil
and produced water was performed to establish equivalency in data derived through various
analytical protocols within the PERP 98-04 project. Statoil provided characterization data
derived from North Sea samples. Their protocol relied on the analysis of individual semi-volatile
compounds, as defined by USEPA Method SW-846 827OC, and total petroleum hydrocarbon
content, as defined by USEPA Method SW-846 8015B. ORNL supplied data derived from GOM
crude oil and brine. Rather than identifying individual chemical compounds, ORNL
characterized samples based on carbon-range content and general chemical classes. It will be
important to understand the equivalency of the North Sea and GOM data when all the information
is combined into a single mathematical model that will predict organic contamination in produced
water.

ORNL submitted samples of GOM crude and synthetic produced water to Mountain States
Analytical, Inc. (1645 West 2200 South, Salt Lake City, Utah 84119) for standard USEPA
Methods testing. These samples were also analyzed at ORNL, using open-column LC separation
to fractionate WSO into general chemical classes. The resulting fractions were then analyzed by
GC/PID to determine the relative distribution of carbon content in each of these classes.
Comparative results of deep-well crude by various protocols are sumrnarr‘zed in Table 7.1. The
TPH content in the GOM crude, as defined by USEPA Method SW-846 8015B, was 300 g/kg
TPH-DRO C&&; 180 g/kg TPH-GRO Ch-Cra; and 200 g/kg TPH-Oil C&Cas. Approximate
ratios of carbon size by the USEPA method were 37,21, and 42% for C&io, C&C&, and
C&& ranges, respectively. The relative distribution using a similar GC/F’ID procedure at

18
ORNL was found to be 50,42, and 8% in the carbon series. Differences in relative distribution of.
carbon content may reflect error introduced by variances in sampling time and frequency of
sampling. . Off-site data were the result of duplicate analyses from a single grab ,sarnple taken on
arrival of Deep1 crude. ORNL data were derived from the average results of five grab samples
taken over a several month time frame.

The principal semi-volatile compounds in the deep-well oil, as identified by USEPA Method SW-
846 827OC, included acetophenone (0.6 g/kg) and methylnapthalene (0.4 g/kg). These particular
constituents represented less than 2% of the total organic content in either GOM or North Sea
sources of crude oil. Chemical classification
“,_ ..^by open LC protocol helps to accounted for the
remaining 98% of the organic character in either oil orproduced water samples. Methylene
chloride-extractable material (TEM) in deep-well oil was 200-400 g/kg. Approximately 35% of
this material could be exchanged intoahexane “matrix. The relative composition of the hexane
matrix is 35% saturated hydrocarbons, 15% aromatic components, and 50% polar material. The
visibly colored red-brown material was primarily present in the polar fraction.

Table 7.1 Correspondence of analytical methods in the

characterization of deep-well crude ,_ ,
.._(1- I_^_ Ull.%” ..-. ,.n _..,e&*, -, . *__I_ -._ , ^_ _...-_ s “I ” ,” .*__,. ._ ._. jl ,. I.,
Deep-weli or1 analysis (g/kg)
Offsite ORNL
Offsite d u p l i c a t e iORNL duplicate
“. . I _ ..^ .-r__ ., _.; ..,, ,.
USEPA Method SW-846 8015B: , , I
Carbon analysis in total
extractable material:
TPH-DRO Cl,+& 314 310 192 219
TPH-GRO 179 i79
TPH-OIL i92 186
TPH: C&i0 98 120
TPH: ClO-CzO 82 80
TPH: CzO--Czs 12 19

USEPA Method SW-846 8270C

Semivolatile organics (SVO):
Acetophenone 0.60 0.58
Methylnapthalene 0.40 0.43

Open LC analysis (ORNL):

Total extractable material (TEM) 193 220
Total hexane extractable
(HEX)/% TEM 50126 88140
Total saturated (TSAT) / %TEM 70136 84138
Total aromatic (TARO) / %TEM 36/ 19 201 9
Total ~01~ (TPOL) / %TEM 88146 116/ 53
8-c .a.. j U”., . j,. ‘ *%“.. I,,l.ds.r, *en
Note: Total TEM = TSAl”, +TARO
. ,_TW? * ,iTPOI+
e>.<--.**i* ,_ 1, .’* ,- 3,: *e. ._ ... _ 1 1 .” ^ _. ^. ~
. I,‘
The total concentration of W.S? in produced water derived from oil/brine simulant contacts was
20-30 ppm, using either USEPA Method SW-846 8015B or ORNL .Gc/FID procedures.
Identified semi-volatile cqmpounds were 1-methylnaphthalene (10 ppb), 2-methylnaphthalene (9
ppb), naphthalene (I4’ppb);and phenol (30 ppb). Again, compounds identified by USEPA
Method SW-846 (8270C) accounted for very little of the total WSO content. Chemical

1 9
fractionation of the total extractable materials suggested that 80-90% of WSO is present as polar
compounds; the next largest fraction is that of aromatic materials. The colored material was
primarily present in the lightly yellow aromatic fraction.

In combining data from the North Sea and GOM samples, it appears that data generated by
USEPA Method SW-846 8015B or ORNL GC/FID procedures correlate well to indicate the total
organic content and relative carbon ranges from these two drilling sites. The toxicity of the
WSO can be estimated using data generated by USEPA Method SW-846 8270C from North Sea
samples and the chemical nature of the material can be estimated using the open LC method
derived from GOM samples.

7.2 SURROGATE RECOVERY AND BLANK LEVELS FOR OPEN LC

FRACTIONATION AND GC/FID ANALYSIS

The analysis of produced water generated by contacting shipped oil samples with simulant GOM
brine probably reflects a negative bias in total WSO content. The oil had been contacted with
injection water during its collection onsite; therefore, a portion of the TPH had already been
stripped from the sample of crude prior to testing in this project. An estimate of the analytical
bias might be achieved by analyzing WSO in actual produced water from this particular site.

The gravimetric analysis of evaporate residue provided a gross estimate of TPH content. This
analysis was usually performed on either crude oil dissolved directly in methylene chloride or
WSO present in the concentrated methylene chloride extracts (TEM fractions) of produced water.
The lower limit of detection for gravimetric analysis was dictated by the 0.2-mg weight limit of
an electronic balance. This measurement corresponded to a gravimetric limit of detection of 0.2
g/L in 650 g/L WSO found in crude. The WSO content in produced water was significantly
lower (20-30 ppm). Since the TEM fraction is derived from the extraction of one liter of
produced water, the balance weight limit corresponded to a gravimetric limit of detection of 12
ppm in aqueous samples.

GC/FID provided a much more sensitive estimate for WSO present in oil or produced water.
Quantitation was based on instrument response relative to that of a series of certified n-alkane
standards. Identification of mass within carbon size ranges is dependent on retention times of the
various alkane standard components (Fig. 7.1). Both C,+& and C1s-C~ ranges had similar FID
response factors; FlD response over the CZ0-C2s range was slightly lower due to column bleed at
the elevated GC temperatures in this portion of the chromatogram. The relatively constant FID
response factor over all carbon ranges indicated that GC/FID analysis provided a relatively
accurate estimate of true TPH mass in the LC fractions. The material present in the TPOL
fraction is made up of compounds containing nitrogen, oxygen and sulfur atoms. FID response to
non-hydrocarbon components is more variable. Therefore, GC/FlD analysis of this fraction in
particular provides an estimate of organic mass.

The successive solvent extractions and multiple concentration steps in either the USEPA protocol
or the LC fraction protocol contributed to losses of WSO material. Although negative bias is
inherent to the WSO analysis, discharge permitting is based on the same USEPA method.
Therefore, data derived from such procedures will still reflect a value for WSO content that can
be used for modeling and subsequent development of engineering practices that will reduce the
generation of WSO. The extraction recovery of WSO were typically estimated by the
introduction of surrogate recovery standards to the produced water prior to manipulating the
sample. Deuterated forms of the spike standards were selected to provide the option for future

20
sample analysis by GC-MS if the latter was considered necessary. These compounds are listed in
Table 7.2, together with the recovery found for each LC fraction. Tribromophenol was
substituted for phenol midway through the experimental campaign. Phenol identification was
difficult due to the proximity of its retention peak to the peaks of the majority of WSO
compounds found in the TEM and HEX fractions.

Table 7.2. Surrogate s@ke recovery compotinds for LC fractionation prdtocol

Recovery Constituent -’ Spike ‘_ ‘*Recovery in each LC fraction (%)
standard surrogate level (pg) TEM HEX TSAT TARO’ T P O L
n-Hexanoic acid Organic-acids 4,635 I”otJ __ 5‘5” -.‘. 0 0 l-10
n-Dodecane-c&h Saturated TPH -- 115 100 60 5-30 0 0
Naphthalene-ds Aromatic TPH 100 100 55 0 20-60 5-40
N-phenol& Polar TPH 70 -_- --- 0 0 O-40
tribromophenol polar TPH 196 100 75 0 0 50-90
,I . . . , _- *“.- e a.. . 1^..~ v.4 )I, / . * _ ,* j _, , __ I _ 8 ,.-i.. ., __ > _ir, ‘j
Positive bias can be introduced into the analytical procedure through the contamination present in
solvents and labware. Therefore, a procedural blank containing all surrogate recovery standards
was run to determine the WSO blank level for each LC fraction. Figure 7.2 presents a
chromatogram of the blank levels and recovery standards associated with the TEM fraction
derived from the methylene chloride extraction of 1 L simulant brine spiked with recovery
standards. The solvent front completely eluted after 2 min. After solvent and recovery surrogate
data was stripped from the chromatogram, regression analysis of the resulting peak area for this
fraction yielded a calculated blank level of -0.2 pg TPH. This level compares with a 0.2~pg TPH
detection limit for the GUFID. Complete processing through the LC protocol yielded the blank
levels shown in Table 7.3. Comparable WSO blank levels were calculated for 1 L of produced
water. A positive blank was associated with the TPOL fraction:. A low level of apparent WSO
resulted from methanol extraction of the organic matter from fired sand. This necessitated an
additional wash of fired sand before a packed LC column was loaded with sample.

Table 7;3 Calculated blank levels

associated with each LC fraction
Equivalent WSO in . ’ ’
LC fraction TPH ( ) produced water (ppm)
,$ ,..- y I( ,-6~82’~.“w~’ . I ,;.. ., _.,, ..j
. lj_ -I-EM
I_. _
‘HEX -0.3 - -0.d2
TSAT O;O 010
TAR0 0.08 0.08 .- .
TPOL 0.2 0.02

7.3 BIASES ASSOCIATED WITH Lc’ FRACTIONATION PROTOCOL

A series of chromatograms were prepared to illustrate the typical quantities of TPH found in
individual LC fractions, as well as the relative losses encountered(“1, with the “.,lfractionation
.._ r__,_ .” .,,,. __, protocol.
These chromatograms followed the processing of produced water generated from contacting 750
mL of deep-well crude with 3000 mL of GOM brine simulant at pH 7,65,000 TDS, 25OC, and
ambient pressure (denoted as experiment “PERF-20” in appendix data tables). As in
chromatograms of diluted oils, the WSO in produced water contained.a,multitude of organic
components. Figure 7.3 presents the WSO contained in the concentrated methylene chloride
extract of produced water (i.e., the TEM fraction). Because of the complexity of the composition,

2 1
it is evident that a general classification, rather than specific component identification, should be
used to characterize WSO in produced water economically. The majority of WSO elute at
retention times comparable to C&i,, and CI&U, n-alkanes. Because of the quantity of material
present in TEM fractions, manual baseline construction was performed to determine the organic
mass in these fractions.

Most of the TEM fraction concentrate was converted to a hexane (HEX) matrix prior to its
addition to the LC column. This procedural step accomplished two objectives. The first was to
acquire TPH data equivalent to data obtained from extractions with hexane, a typical solvent used
to determine WSO in field testing. Second, the LC protocol required an initial nonpolar matrix.
Organic fractionation was then accomplished by serially increasing the polarity of the eluent.
Figure 7.4 presents the chromatogram of WSO in the hexane matrix. Concentration of hexane
requires a heating temperature of 85OC, which is 20°C higher than that needed for methylene
chloride evaporation. Therefore, some of the more volatile WSO constituents in the C&Z10 range
were less prevalent in the chromatogram of HEX extracts, yielding a carbon range distribution
observed using USEPA methods based on hexane extractions. Approximately 50% of the
surrogate recovery standards are lost in the solvent exchange. Beginning with experiment PERF-
15, solvent substitution was modified to reduce organic losses as a result of elevated
temperatures. The volume of hexane was reduced by evaporation at 35OC under a flow of nitrogen
gas. Again, the quantity of material present in TEM fraction was such that the chromatograms
were typically not baseline resolved. This may have contributed to a slight negative bias in the
calculation of percent TPH transferred to the HEX fraction.

Pentane was used to elute saturated organic compounds from the LC column. Typically, the
TSAT chromatogram (Fig.7.5) contained only the n-decane surrogate standard at a recovery of
approximately 25%. Aromatic compounds were present in the produced water as indicated in the
number of peaks found in the chromatogram of the TAR0 fraction. Naphthalene was usually
recovered at 40% of initial spike levels. The TAR0 chromatogram (Fig. 7.6) contained column
contaminants in the CS-C$ carbon range equivalent to about 0.1 ppm WSO in produced water.
This contaminant level was subtracted from all TAR0 fraction data.

The LC protocol developed by Mills et al. was used to calculate the quantity of polar constituents
in produced water by subtracting the combined WSO mass found in TAR0 and TSAT fractions
from that found in the TEM fraction. This type of calculation was made for each contact
experiment performed in the study reported here. These values are denoted as TPOL(subtract) in
the appendix data tables. This form of calculation may have overestimated the contribution of
polar material, considering that only 40% of TEM material was actually transferred to the LC
column in the form of the HEX extract. If it is assumed that losses in the solvent exchange from
methylene chloride to hexane were primarily due to the evaporation of low molecular weight
organic acids, then the TPOL(subtract) data might adequately compensate for this loss and thus
would not significantly bias the calculation of total WSO found in this fraction.

The polar material in produced water was also estimated by adding a final elution step to the LC
protocol defined by Mills et al. After the TAR0 fraction had been collected, a polar solvent was
employed to elute the polar WSO from the column. Initially, methylene chloride was used as the
eluent; however, methanol was found to be more effective in removing polar constituents and was
used in experiment PERF-11 and all experiments thereafter. Even with a methanol solvent, the
presence of colored material remaining on the fired sand at the head of the column from the LC
separation of diluted crude indicated that some polar constituents were irreversibly adsorbed to
the column. Nonetheless, the chromatogram (Fig. 7.7) of the eluted TPOL fraction, denoted as

22
TPOL(gc) in the appendix data tables, implies that most of the WSO compounds were present as
polar constituents in the C10-C20 carbon range. The hexanoic acid and tribromophenol recovery
standards in this fraction were present at 5 and 70%, respectively.

7.4 COMPARISON OF LC FRACTIONATION PROTOCOL WITH IWRARED

SPECTROMETRIC A N A L Y S I S , _

A secondary comparison of methods is presented below. The TPH content of a portion of

produced-water samples was determined by the GC/PID analysis of TEM fractions and,
secondarily, by the IR analysis of their tetrachloroethylene extracts. Results based on IR data
tended to underestimated the TPH content; the salinity present in the produced water depressed
lR absorbance readings in the samples (see Table 7.4).

Table 7.4. Correspondence of TPH results

from LC fractionation and IR spectroscopy
- TPH (mg/L)
.._,,. _.. ..I . “. ___“... -. .iI _, ;_ , _
.. ,,,I Experiment LC fractionation_-
Tl!?M fraction _ IR spectroscopy
s I”e 1 _. -_ ._*. “.^ __. .x “. ,.., ~, 1,, ”
PERP-6 12 ‘I-
PERP-7 37
PERP-8 25 6
PERP-9 30 13
PER&10 22 13
PER&l 1 2 6 25.
PERP-12 11
PERP-13 21 16
PERK14 10 49.8
PERP-15 12 12.3
PERP-16 14 10.5
PERP-17 18 17.6
PER&18 2 1 ’ 23.1
PERP-19 10 8.7
. PERP-20 .21 ld:6 >’
._I .., ,.-_. < 4 ., _ , *.*, ,. ._;

8. RESULTS

8.1 CHEMICAL CHARACTERIZATION OF DEEP-WELL CRUDE

The LC column fractionation procedure was used‘to characterize deep-well crude according to
the quantity of organic matter in each chemical class, as well as carbon size distribution. Results
presented in Table 8.1 summarize the data for three to five deep crude analyses, in which lg of
mixed oil was diluted 25-fold with methylene chloride. Data were calculated in un&of grams
organic mass per liter of oil but can be‘converted to a weight basis using a density value of 0.8432
g/L at 25OC. The TEM fraction was made up of 45% each of C&~O and C&C& material. The
&-Czs data indicate that 35% of the organic mass was transferred to the HEX fraction; the major
loss of material was fromthe C&r0 carbon range. Of the organic material transferred to the
HEX fraction roughly 75% of the mass was contained in the saturated hydrocarbon class. The
remaining mass was equally divided between aromatic and polar constituents. The standard
deviation in the data reflects a combination of error introduced in sampling the mixed oil and in
preparing the LC fractions.

Table 8.1. Chemical fraction data for GOM deep-well crude samples
LC fraction &--c28 (g/L) c6-cl0 (g/L) CIO-c20 (g/L) c20-c28 (g/L)
TEM 2Olk 26 114521 90. * 14 9.7 + 6.1
HEX 70. f 19 5.3 f 3.5 54* 19 10. Zk 4.7
TSAT 74 * 9.1 8.3 z!z 4.0 61+11 3.5 I!z 0.7
TAR0 21+ 14 4.0 I?I 3.4 15 r!~ 9.6 4.0 f 3.0
TPOL(gc) 16+ 3 . 6 14 AZ 4.2 0.57 + 0.06 0.67 z!z 0.57

8.2 WSO AS A FUNCTION OF PERCENT WATER CUT

Methylene chloride-extractable material in the equilibrated produced water was typically present
at 20-30 ppm. Its chemical character differed significantly from that of the deep crude. The
values listed in Table 8.2 reflect the average of three oil/brine contacts in which the experimental
baseline conditions were set at 80% GOM synthetic brine/oil, equilibration for 4 days at 25OC, pH
7 brine, and 65,000 TDS. The brine volumes added to the WAF vessel varied from 1 to 3 L,
while the oil volume varied from 0.25 to 0.75 L. TEM organic matter was distributed between
the low and midrange carbon masses, similar to the oil data. However, the WSO eluted from the
LC column are primarily polar compounds with a minor contribution by aromatic constituents.
The slight yellow coloration of produced water appears to be associated with the aromatic
fraction. As would be expected, saturated compounds were not noticeably extracted into the
produced water. -

Table 8.2. Chemical fraction data for GOM synthetic produced water*
LC Fraction Cb-C2s (mg/L) c6-cl0 (mg/L) ClO~cU, bg/L) c20-c28 @g/L)
TEM 21k4 7 z!z 0.4 13 +3 0.8 zk 1.3
HEX llf5 0.4 f 0.4 lOf4 0.01 + 0.01
TSAT 0.02 & 0.02 0.02 z!z 0.01 0.00 f 0.00 0.00 zk 0.00
TAR0 0.2 * 0.2 0.07 z!z 0.08 0.04 zk 0.02 0.00 zlz 0.00
TPOL(gc) 2.4 + 1.7 0.00 AI 0.00 2.4 f 1.7 0.00 f 0.00
“80% GOM synthetic brine/oil, 4 days at 25 *C, pH 7 brine, 65,000 TDS WAF vessel

Figures 8.1 and 8.2 summarize the data trends for various water/oil ratios. There was a slight
negative trend in WSO content in all chemical fractions as the water cut increased. This effect
was a combination of decreasing solubility of the Cl&&, mass and increasing solubility of the
C,+Z1o carbonaceous material.

8.3 WSO DISTRIBUTION COEFFICIENTS AS A FUNCTION OF PERCENT WATER

CUT

The analytical results obtained in the characterization of crude oil were used to determine values
of the approximate distribution coefficients (&s) of WSO in GOM brine as a function of percent
water cut and brine volume. The l&s were calculated for brine simulant at baseline conditions of
65,000 TDS, pH 7 brine, ambient pressure, and 25*C. Equilibrium conditions were assumed after

24
a 4day contact time. Calculations based on the following equation were performed for each
chemical class:

Ka = (initial oil concentration) x (oil

.- volume) - (WSO in brine) X (brine volume!,..
_. (WSO in brine) k (oil volume)’ - ‘^

The data presented in Fig. 8.3 imply that the value for Log I& of each chemical fraction remained
constant with water cut. Average values for Log Kd were 4.0,3.8,5.0, S.l,and 6.3 for TEM,
HEX; TPOL, TAR0 and TSAT fractions, respectively. The value of Log I(d for each carbon
range was also independent of water cut, as borne out in Fig. 8.4 as well as in the data
summarized in Tables 8.3 and 8.4. The,only deviation from this trend was observed with slightly
higher values for the data set obtained for a 73% water, cut. These,d.ata,were generated using the
small pressure vessel. Although all data acquired with the pressure vessel were consistent,in the
subset, WSO values in the produced water were lower than those obtained with the WAF vessel.
,
8.4 WSO AS A FUNCTION OF SALINITY

The variation of TDS (as salinity) did not significantly alter the WSO content of the produced
water. As seen in Fig. 8.5, the total WSO remained at lo-15 ppm over the chloride concentration
range of 40,000 to 115,000 ppm. Neither the chemical character nor the carbon content (Fig. 8.6)
varied with TDS.
‘.
8.5 WSO AS A FUNCTION OF pH

The pH of the GOM brine simulant. was-buffered _ ._7-7.5 by the presence of an oil layer. In
i -.._.._ to
order for the oil/water contact to be performed at a pH outside this range, NaOH or HCl had to be
added continuously to maintain the desired pH level. The use of a 0.25 M borate buffer was not
sufficient to maintain the pH of the aqueous phase at more alkaline values. The required pH
adjustment was reduced.with each contact day, such that no further adjustment was required on
the fourth day. Thus, equilibrium was assumed to have been reached at this point. As expected,
the pH of the brine simulant affected the concentrations of COs2- and HCOJ-. Acidic solutions
reacted with HCOs- anion. Basic solutions increased the initial carbonate concentration
sufficiently to precipitate alkaline earth metals, thereby reducing the equilibrium concentration of
both COi2- and HC03-.

The WSO content almost doubled for pH values above 7 (Fig. 8.7). This was primarily due to the
enhanced solubility of polar organic compounds that ionize at higher pH levels. An enhanced
solubility of the Cro-Cm carbon-range material and a decline in C&r0 mass for pH values
greater than 7 were also observed
. (Fig. 8.8).

8.6 WSO AS A FUNCTiOti ~ti’~titiP~tii’URE : .,

The chemical character of the WSO did not. appear to be dependent on brine temperature (Fig.
8.9). However, the percentage mass in each carbon range was altered; the mass of WSO in the
C&Jr0 carbon range decreased while that in the C ro-C20 range increased (Fig. 8.10). Thus, the
net quantity of organics in the produced water increased by only 5 ppm for a 60°C elevation in
solution temperature.

25
8.7 WSO AS A FUNCTION OF PRESSURE

The distribution of organic compounds in GOM brine was studied as a function of pressure over
the range of 1 to 58 bar. The results are presented in Fig. 8.11 and 8.12. It was found that WSO
content in the TEM fraction increased by less than ten percent over the pressure range. The total
WSO content for this subset of experiments were less than expected, based on pervious data
obtained with the WAF vessel. Experiments performed with the WAF vessel were typically
conducted with 300 mL oil in contact with 1 L of brine. The total WSO levels were generally 16-
20 ppm for WAF experiments. The WSO of the 200 mL brine drawn from the pressure vessel
consistently averaged only 4-6 ppm. At this point, the reason for this variance is unknown.
Pressure data presented in these graphs also exhibit a fair amount of scatter, attributed to both the
low level of WSO in the pressurized samples and tie difficulty in separating the phases after
contact in the opaque pressure vessel.

The slight increase observed in total WSO as a function of pressure appears to be associated with
the polar fraction. The influence of pressure on the measured fractions of saturated hydrocarbons
(TSAT) and aromatic hydrocarbons (TARO) were investigated, but no trends were observed.
The enhanced quantity of WSO observed in the polar fraction was the result of additional
dissolved C&Z10 material (Fig. 8.12). A slight decrease in quantities in the C2&2s range was
also evident.

8.8 COMPOSITION OF Cl-C6 ORGANIC ACIDS IN PRODUCED WATER

Three organic acids were typically found in the produced-water samples: formic, acetic, and
propionic acids. No higher-carbon acids were found. Acetic acid, which made up the bulk of the
acid mass, was present at levels similar to that of TPH (lo-30 ppm). Propionic and formic acids
were present at approximately 60 and 20%, respectively, of the acetic acid concentration.

It is important to understand the extent to which low-molecular-weight acids might elevate the
apparent TPH concentration found in methylene chloride and HEX extracts. The NPDES limit
set for the oceanic discharge of GGM produced water is 42 ppm TPH. If the acid and TPH
concentrations in produced water were roughly equivalent, the TPH analysis might be artificially
overstated by the presence of high levels of organic acids. To estimate the proportion of acids
extracted by methylene chloride and hexane, 1 L of GGM brine simulant was spiked with 1000
pg each of formic, acetic, propionic,‘and n-butyric acids. Both a TEM fraction and a HEX
fraction were generated from the spiked brine using the standard extraction protocol for produced
water. The fractions were analyzed by GC/FID and IC. Only propionic and n-butyric acids were
sufficiently resolved from the solvent fronts to quantitate these particular acids by GC/FID.
These acids fell within GC retention times of CT and Cs n-alkanes. Less than 5% total acid was
recovered in the TEM fraction, corresponding to an equivalent TPH content of 51 ppm in a
produced water sample. Less than 1% organic acids were recovered in the HEX fraction,
corresponding to a value of 23 ppm TPH in produced water. The organic acid concentration
observed in PERF oil/water contacts indicates that no more than 30 ppm organic acids were
typically present. Estimating that no more than 5% of the acids are recovered in the TEM
fraction, TPH would typically be overestimated by less than 2 ppm.

The TEM and HEX fractions were analyzed by IC by first diluting the samples by a factor of 20
with acetonitrile. Relative ratios for formic, acetic, propionic, and n-butyric acids were
1: 100:500:2500, respectively. The total acid concentration found in the TEM fraction was
approximately four times that determined by GC/FID, due to the ability to the ability of the

26
method to resolve the formic and acetic acid peaks from the solvent front. Only propionic and
butyric acids were observed in the HEX fraction at a concentration ratio of 1: 10, respectively.
Eight percent of the total acid mass found in TEM was transferred into the HEX fraction. Losses
of formic and acetic acids from this fraction are attributed to volatilization in the solvent
exchange necessary to produce the HEX fraction. Assuming that the typical TPH concentration
in HEX fractions derived from.the processing of produced water samples was approximately 12
ppm and that 30 ppm of organic acids was present in the water, the TPH concentration would be
elevated by no more than 0.5 ppm in the HEX fraction.

Several experimental parameters were found to affect the concentrations

&- e ., _ of.) .-organic
*, acids found in
produced water. The oil/water ratio did not consistently affect the levels of organic acids in
produced water (Fig. 8.13). However, the acid levels decreased as the brine simulant became
more alkaline, a trend opposite to that observed for hydrocarbon content in the TEM fraction
(Fig. 8.14). The organic acid concentration was near zero at 42,000 ppm chloride; acetic and
propionic acids were consistently present at about 10 ppm at the higher salinity levels (Fig. 8.15).
Concentrations, of both acetic acid and TPH were elevated at temperatures greater than 50°C
(Fig. 8.16). Elevating pressure did not significantly affect organic acid content in produced
water.

In addition to organic acids, the chromatograms of produced water included elution peaks for the
primary anion, chloride, as well as a minor phosphate constituent (Fig. 8.17). Phosphate was
present at slightly greater than 1200 ppm at room temperature; however it was lost as the
temperature increased. Only 300 ppm of phosphate was contained in the produced water of a
75OC oil/water contact experiment. Phosphate &as most prevalent at near-neutral pH levels and
was rapidly lost when the pH was increased or decreased. Phosphate concentration showed no
consistent trends with either the salinity or the water cut.

8.9 PRESENCE OF INORGANIC CON$TITUENTS

.._. ,. IN CRUDE AND PRODUCED
WATER

Alkaline and alkaline-earth metals were the primary metal components in the deep-well crude
according to results of ICP-AES analyses; their total concentration was about 7 g/kg. The
primary transition metals are nickel (25 mg/kg); silicon (248 mg/kg); iron (22 mg/kg); vanadium
(67 mg/kg) and zinc (6 mg/kg). Only selenium (3.5 mg/kg) approached the level to be considered
a toxicity characteristic of the oil.

Results of inorganic chemical analysis of produced water samples indicated that only cadmium
and antimony were occasionally extracted into the water phase at the level of 0.1 ppm. The
concentrations of these metals did not appear to be affected by variations in pH or water/oil cut.
The alkaline and alkaline-ear@ mefal,concentratjons in produced water were equivalent to those
found in the starting brine simulant.
.I .’

27
 Table 8.3. WSO distribution coefficients within chemical classes

Simulant Oil Distribution coefficients, &

Water cut volume volume TEM HEX TSAT TAR0 TPOL (GC)
Experiment (“72) (mL) (mL)
PERF-9 20 1000 4000 6,719 5,095 1,457,002 88,523 NAa
PERF-7 50 1000 1000 5,471 3,208 1,746,309 46,785 2,953
PERF-10 50 1100 1100 8,257 7,339 1,267,370 144,889 2,560
PERF-6 67 1000 500 17,291 6,671 1,188,170 627,449 NA
PERF-8 80 1000 250 8,132 4,522 2,156,409 139,427 NA
PERF-16 80 2400 600 12,097 6,606 4,794,989 128,948 4,468
PERFPO 80 3000 750 9,698 12,009 no extraction no extraction 14,358
PERF-25 73 200 73 52,006 20,220 4648566 128,756 3,927
aNA = not analyzed.

Table 8.4. WSO distribution coefficients within carbon ranges for TEM fractions
Experiment Water cut Simulant Oil Kd in TEM Fractions
(“Yo) vdlume volume c6410 Go-C20 c20-c26
(mL) (mL)
PERF19 20 1000 4000 12,195 4,375 No extraction
PERF-7 50 1000 1000 9,525 3,637 No extraction
PERF-10 50 1100 1100 18,654 4,937 No extraction
PERF-6 67 1000 500 21,306 14,346 No extraction
PERF-8 80 1000 250 16,025 5,876 No extraction
PERF-16 80 2400 600 16,792 9,153 No extraction.
PERF-20 80 3000 750 14,941 6,873 No extraction
PERF-25 73 200 73 51.388 54.557 No extraction
 >

9 . CONCLUSIONs ._

Quantitative characterization data.were.,c_ql!ecfed,for the WSO

.- A, ‘. / .Feajxwe”, , rs>,m >Icontent
”.<, =a ,., . b>, of
. a produced water as a
function of several experimental parameters-The WSO content was defined on the basisof an 1
open LC column procedure in which the total organic material was fractionated into methylene
chloride-soluble, hexane-soluble, saturated hydrocarbon, aromatic hydrocarbon, and polar
hydrocarbon classes. The proportions of gas-, diesel- and oil-TPH within each fraction were then
determined by GCYFID analysis. Data from this protocol were supported by the results using
USEPA SW-846 Method 8270 C for semi-volatiheorganic analytes and infrared spectroscopy for
TPH content. Analyses were also”pe~orme;;i”‘a”number ,of in-organic constituents .

The G& crude was obtainedfrom a single deep-well site. The TPH contents in this sample
were 50,42, and 8% for the C&Z+ _ C1+&, .- . .- and C,-Czs ranges, respectively. The principal
SVOCs in the deep-well oil were acetophenone (0.6 .&g) andmethylnaphthalene
(0.4 g/kg), which represent less than 2% of the total organic content. Chemical classification by
the open LC protocol was used to account for the remaining 98% of the organic character of the
oil. Methylene chloride-extractable material (TEM) in the deep-well oil was present at
200-400 g/kg. Approximately 35% of this material could be, exchanged-into a hexane matrix.
The relative composition of the hexane matrix was,35% saturated . ,.. %(“^ ., hydrocarbons, 15% aromatic
components, and 50% polar material. The visibly colored red-brown material was primarily
present in the polar fraction. Alkaline and alkaline-earth
....Id.. _..-. Ii.metals
“. __A,./ ^Il\.ewere
”sV,‘l(_~.LIII.the
.>~ primary ICP metal
components in the deep-well crude; their total concentration was.about,.lJ g/kg. Several RCRA
metals were detected-slightly above the ICP detection limit; however, they were not present
above the USEPA-defined toxicity limits.

The sample of GOM crude described above was contacted with a GGM ,_.... brine.simulant
, I., ,i ..,* /+in; a WAF
vessel to generate produced water samples. The total concentration of WSG observed rn the
oil/brine simulant contacts ranged from 20-30 ppm. Four SVOCs were identified:
1-methylnaphthalene (10 ppb), 2-methylnaphthalene (9 ppb), naphthalene (14 ppb), and phenol
(30 ppb). The distribution within tte” chemical.fractions differed~significantly from that of the
deep crude. Chemical fraction~~~~~pf~~~~~~~l’~~~~~~~als ,” -*“--*--“_I_ ixr--dsia-*.. x” ._ suggested that 80 to 90% of
the WSO was present as polar compounds; the next largest fraction was that of aromatic
materials. The visibly colored material was primarily present in the light-yellow aromatic
fraction. TEM organic matter was distributed between the”low ,and midrange carbon masses,
similar to oil data.
I “, ~_ ” **_ . .,- _ .,
Variation in the water/oil ratio indicated that there was a slight
. -_ -.-L.*~,.“~^LI,l^ __.... I”. negative trend in the WSO content
in all chemical fractions as the w,ater,.cut increased, This, effect,was a combination of, the
decreasing solubility of the Ci&Z, mass and the increasing’solubility of the C&K, carbon
material. Analytical results obtained in the characterization of crude oil were used in combination
with data from water cut experiments to determine approximate distribution coefficients (&
.
values) of WSO in GOM brine as a function of percent water cut and brine volume. Distribution
data imply that that brine was saturated with WSO for water cuts of lessthan. 50%. ,

The variation of TDS with salinity did not significantly alter the WSO content in produced water;
however, the pH of the brine did affect TPH content, Normally, the pH of the GOM brine
simulant was buffered to 7-7.5 by the presence of an oil layer. When the buffering capacity of
oil was exhausted with base, the WSO content almost doubled for pH values above 7. This was
primarily due to the enhanced solubility of polar organic compounds that deprotonate at higher

29
pH values. There was an enhanced solubility.of the Cr,-#& carbon range material and a decline
in C&ra mass for pH values greater than 7.

The chemical character of WSO did not appear to be significantly dependent upon solution
temperature, although the percentage mass in each carbon range was altered. The mass of WSO
in the C&&-, carbon range decreased while that in the Cra-Cm range increased. Thus, the net
quantity of organics in the produced water increased by 5 ppm for a 60°C elevation in solution
temperature.

Over the pressure range of l-60 bar, total WSO content increased by less than 10%. The
additional quantity of WSO was primarily associated with the polar fraction. The.most notable
difference in the content of produced water derived from high pressure contacts was the increased
amounts of C&Cra range material, particularly in the polar fraction and to a lesser extent in the
aromatic fraction. There was considerable scatter in pressure data due to the low volumes of oil
and water used in this series of experiments. The low quantities of saturated hydrocarbon
fractions made it difficult to discern any trends for this chemical fraction.

Three organic acids were typically found in the produced-water samples: formic, acetic, and
propionic. Acetic acid, which was present at 10-30 ppm (similar to that of TPH), made up the
bulk of the acid mass. However, acid levels decreased as the brine simulant became more
alkaline, a trend opposite from that observed for the hydrocarbon content in the TEM fraction.
Both salinity and temperature slightly elevated the organic acid concentration, particularly that of
acetic acid. Less than 5% total acid in produced water was recovered in the TEM fraction; less
than 1% organic acids was recovered in the HEX fraction. For 30 ppm of total acid present in
produced water, it was estimated that the TPH content would be overestimated by less than 2 ppm
if methylene chloride were used to extract WSO from produced water. The TPH content would
be overestimated by 0.5 ppm if hexane were used for solvent extraction. Assuming a constant
distribution coefficient, overestimation would then increase in a linear fashion relative to total
acid concentration in the produced water.

Of the five experimental parameters tested (percent water cut, pH, salinity, temperature, and
pressure), the factor that most controlled the total WSO in produced water was that of aqueous
phase pH. Beyond a value of pH 7 significant quantities of Cra-Cz~ range material became
markedly soluble at they deprotonated in basic GOM brine. Both the absolute and relative
volumes of GOM brine and crude additionally affected total WSO. Produced water appeared to
reach a saturation level of WSO at a 50% water/oil ratio. Pressure slightly enhanced WSO by
increasing the relative quantity of C&ta range material. Temperature primarily altered the
relative ratio of the carbon ranges within the WSO without significantly elevating the total WSO
in the GOM brine. Salinity had the least affect on the chemical character or the carbon size of
WSO in produced water.

10. REFERENCES

M. A. Mills, T. J. McDonald, J. S. Bonner, M. A. Simon, and 8. L. Autenrieth, “Method for

Quantifying the Fate of Petroleum in the Environment,” Chemosphere 39,2563-82 (1999).

M. A. Franson, (ed.), Standard Methods for the Examination of Water and Wastewater, 18th ed.,
American Public Health Association, Washington, D.C., 1992, pp. 2-23 to 2-28.

30
,
Figures for Text of Report
Fig. 4.1. Schematic’diagram of oil/water contact pressure vessel.

34
.
Fig. 4.2. Oil/brine contact pressure vessel.
I’ I

35
 5,000

0,000

5.000

0 8 16 24 32

Time (min)

Fig. 7.i. GC/FID chromatogram of n-alkane standards.

(~-FL injection of 15 ppm of each alkane)

-Hexanoi- _ - Naphtha,ene- ~ ~-‘ -

i II
I
_ acid 8.5 mi _ _ /
_ _3
_ _ _ -n-l& m _ _ _ _ _
I 11.3 i;t\

I
I.5 n-tin _ _ _ _ _ _ _ _ _ _ _ _
Tribromophenol

/
min _ _ _ _ _ _

0 10 20
Time (min)

Fig. 7.2. GC/FID chromatogram of surrogate recovery standards.

(2-a injection of methylene chloride (TEM) extract)

36
 I n-Dodecane 11.35 min

15
Time (min)

Fig. 7.3. GC/FID chromatogram of TEM fraction.

(0.5~pL injection of 1.2 mL of TEM-20 fraction)

Hexanoic acid 8.5 min

._ _ ____------ --
._______

. .
____________________________

_ _ _ _. I _-_______.
G __________________
it n -Dodecane 11.35 min

, I I Tr>bromophenol 17.1 m

Time (min)

Fig. 7.4. GC/FID chromatogram of HEX fraction.

(OS-pL injection of 1.0 mL of HEX-20 Fraction)

37
f 10,000 __ _ ._----_ --__---_--____-_______

8 y-Docecane
_ _ -__--__--------_ 11.35 min
---_--__---__
II I
~,

20 3,o
Time (min)

Fig. 7.5. GUFID chromatogram of TSAT fraction.

(3-PL injection of 1.1 mL of TSAT-20 fraction)

10,000

0
0 IO 20 30
Time (min)

Fig. 7.6. GUFID chromatogram of TAR0 fraction.

(3-PL injection of 0.95 mL of TARO-20 fraction)

38
 __________________ -__ -_ ------.
Tribromophenol 17J min

I I 1

0 10 20 30
Time (min)

Fig. 7.7. GC/FID chromatogram of TPOL fraction.

(1-pL injection of 1.0 mL of TPOL-20 fraction)
 _ _ _ _ --+ _ ___---_--__ --------------

n HEX
A TSAT
x TAR0
. ~Ww) 1

20 40 60 80 100
% Water Cut

Fig. 8.1. WSO chemical character versus percent water cut.

80

ACIO-C20
[e C20-C28

0 20 40 60 80 100
% Water Cut

Fig. 8.2. WSO carbon content versus percent water cut.

40
 100,000 m
X

75,000 _..-----.--.-.---.-.---~-----

m
y” 5oJ)O~ -. - _ - . - _ _ _ _ _ _ ic- _ _ _ _ . _ _ _ - _ _ _ -

25,000 -. _ _. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ -x~. _‘_ _

l

I I
0 I I
8 n

0 20 40 60 80 100
% Water Cut

-Fig. 8.3. WSO I<rl k&suS percent watkr cut.

100,000 m
X
)I: m
75,000 __ _ _ _ _ . . _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _
m
y” 50,000 -. _ . _ _ _ x- _ _ _ _ _ _ _ _ _ _ _ - _ _ - _ _

25,00() _. _ _ _ m- _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _
l

; @
. f2
0 I I

0 1000 2000 3000 4000

Simulant Volume (ml)

Fig. 8.4. WSO & versus GOM simulant volume.

 40,000 80,000 120,000
Final [Cl’]

Fig. 8.5. Chemical fraction as a function of total dissolved solids (as Cl-).

80

60
I=C6-Cl0 1
40 lAC10-C20/
1 e C20-C28j

I
n

20 ________--_-_ __-------------_

0
40,000 80,000 120,000
Final [Cl’]

Fig. 8.6. Carbon content in TEM fractions as a function of total dissolved solids (Cl-).

42
 *TEM
n HEX
A TSAT
x TAR0
x TPOL-sub
l TPOL-gc

6 8 IO
Final pH

Fig. 8.7. Chemical character as a function of GOM brine pH.

-._________ -_--------- A _ _ _ _ _ _ _ _
A
.C6-Cl0
A cl0 -C20
e C20 - C28

0
4 6 8 IO
Final pH

Fig. 8.8. Carbon con&t in TEiM fractions versus GOM brine PH..
 ___-_~----_----__---_________ -_ __ pK---
I* * n HEX
A TSAT
x TAR0
m TPOL-sub
1 l TPOL-gc

0
20 40 60 80 100
Temperature (“C)

Fig. 8.9. Chemical character as a function of GOM brine temperature.

1 0 0 1
A
80 _____ -.-__------_
A
_ _---_._ ____

60
40 .m---------------------------

n
20
n
__._____-_--___---___________
t
e n I’8

0 --. = I - I I

20 60
Temperature (“C)

Fig. 8.10. Carbon content in TEM fractions as GOM brine temperature.

44
 8

n HEX

A TSAT

x TAR0

m FOL-
s u b
. TPOL-

0 10 20 30 40 50 60 70
Pressure (bar)

r Fig. 8.11 Chemical Character as a function of GOM brine pressure.

100 1
80 _ _ . _ . _ .A _.__________________ ___ 1

60 _ __- -_.__________ n
_ __ _._ --._-__-

40
20
1
n

A
A __.______________--_
n

_ --------.-
A
1 a C20-C28

0
0 10 20 30 40 50 60
Pressure (bar)

Fig. 8.12 Carbon content in TPOL as a function of GOM brine pressure.

40

+ l . Formic
-.---_------- - --------__-- e?.-.-. n Acetic
A A Propionic
n i
A
-.____---__- l
-------___--
n l TEM
--A---‘-
n

n
0
I T I -

25 50 75 100
Water Cut (%)

Fig. 8.13. Influence of percent water cut on organic acid content.

60

1
n

l Formic
n Acetic
A Propionic
l TEM
n
4 l 2
0 A
p t n

6
PH

Fig. 8.14. Influence of brine pH on organic acid content.

46
 l Formic
n Acetic
A Propionic
l TEM

40,000 80,000 120,000

Salinity (ppm Cl’)
I

Fig. 8.15. Influence of s&i&y on organic acid .content.

40 1 I
n

g30
.___- _ ______.______..-__ _ --- _ --
1 I l Formic
n Acetic
A Propionic
l TEM

d 20 40 60 80
Temperature (“C)

Fig. 8.16. Influence of temperature on organic acid content.

-47
 I I I I

0 10 20 30 40
Time (min)

Fig. 8.17. Ion chromatogram of organic acids in produced water.

0

I APPENDIX
 APPENDIX

The tables in this appendix list data acquired for the oil/water contact experiments, analyses of oil
samples, certification of GOM brine simulant, and protocol standards. These experiments are
denoted by “PERP-“, “Deep- “, “GOM-“, and “STD-“, respectively, in each of the tables. The
data for PERP designations are further delineated according to the experimental parameter that
was being studied. Baseline experimental conditions were defined as 80% water cut, 65,000 ppm
salinity, pH 7 brine, 25’C, and ambient pressure. Experiment numbers for the oil/water contacts
are listed in table A. 1. At least two of these experiments were performed to study the variation in
WSO content for a given parameter performed outside the baseline values.

Table A.1 PERF experiments associated with

each oil/water contact parameter
Parameter Oil/water contact
Baseline PERP-8; PERP- 16; PERP-20
Water cut PERP-6; PERF-7; PER&9; PERF-10
PH PERF-11; PERP-12; PERP-13
Salinity PERP-14; PERF-15
Temperature PERP-17; PERF-18
Pressure PEW-2 1; PERP-22; PERF-23; PERP-24;
PERF-25; PERF-26; PERF-27; PERF-28

The details pertaining as to the starting conditions of each experiment are summarized in Table
A.2. The equilibrium concentrations of major anion species and the pH levels at the completion
of an oil/water contact are listed in Table A.3. Corresponding cation data are presented in Table
A.4. Summation of WSO content for oil/water contacts, as-received oil, and recovery standards
are contained in Tables A.5 through A.7. The WSO results based on gravimetric and IR analyses
are located in Table A.5; those based on the open LC protocol are presented in Tables A.6 and
A.7.

52
 ‘.

L .

Table A.2 Starting conditions for PERF experiments

Water Oil Total
Experiment Simulant GOM Temp. Pressure Salinity volume volume contact Comments
batch pH (“C) (in. Hg) (mg/L) (mL) (mL) time (days)
GOM-1 4L-Prep 7.5 --- --- 62975 --- -em ---
PERF-3blka GOM-1 7.5 25.0 24.06 62975 1000 0 --- + 312.5 ppm Na+ plus 8640 ppm SO,*-
PERF-4oil --- 7.5 25.0 --a 0 2 ---
GOM-2 4L-Prep 6.9 --- --- 60793 --- --- --- Start: no iron added to GOM simulant
PERFS GOM-2 6.9 25.0 29.06 60793 1000 500 4.83 Start: addition of recovery standards
PERF-7 GOM-2 6.9 25.0 29.06 60793 1000 1000 5.76 Start: gravimetric analysis for TEM and HEX
fractions only
PERF-8 GOM-2 6.9 25.0 29.06 60793 1000 250 4.00
GOM-3 4L-Prep 6.9 --- --- 65308 --- -mm ---
PERF-9 GOM-3 6.9 25.0 29.06 65308 1000 4000 4.00
PERF-10 GOM-3 6.9 25.0 29.06 65308 1100 1100 5.75 GOM pH adjustment to 8.7 with 50% NaOH
prior to oil contact
GOM-4 4L-Prep 6.8 --- --^ 62143 --- --- ---
PERF-11 GOM-4 9.03 24.5 29.06 65308 1200 300 2.96 pH continuously adjusted with 50% NaOH
PERF-12 GOM-4 4.73 24.5 29.1 65308 1200 300 5.94 pH continuously adjusted with
concentrated HCI
PERF-13 GOM-4 9.5 24.5 29.1 65308 1200 300 5.86 pH continuously adjusted with 50% NaOH
GOM-5 4L-Prep 7.4 m-e --- 40638 --- --- --- 50% NaCl content in simulant
PERF-14 GOM-5 7.4 24.5 29.14 40638 1200 300 4.89 50% NaCl content in simulant
PERF-Stdl b GOM-5 7.4 24.5 29.14 40638 1000 0 --- Start standards recovery with hexanoic acid
GOM-6 2L-Prep 6.82 --- --- --- --- --- ---
PERF-15 GOM-6 6.82 25.0 29.31 114500 1200 300 5.33 150% NaCl content
PERF-Deep1 --- 25.0 29.31 -ms --- 0.995 s-m 0.8391 g oil in 17.66 mL CH&12
GOM-7 4L-Prep --a --- --- 62143 --- --- -me Simulant used for Std2 and Std3
GOM-8 GL-Prep 7.1 --- 72900 --- 2 L GOM-7 + 4 L GOM-8
PERFStd2 GOM-5 --- 24.5 29.14 40638 1000 0 --- Standards recovery with hexanoic acid
“Process blanks
bRecovery of surrogate standards
 Table A.2. (continued)
Water Oil Total
Experiment Simulant GOM Temp. Pressure Salinity volume volume contact Comments
batch PH (“Cl e-m-) Fw (mL) time (days)
---
GOM-5 --- 24.5 40638 1000 0 Standards recovery with hexanoic acid and
tribromophenol
PERF-16 GOM-8 7.1 25.0 29.41 72900 2400 600 6.02 Baseline with double water & oil
PERF-17 GOM-8 7.42 47.2 29.40 72900 1200 300 5.83 Temperature (47.2’C) in water bath
PERF-18 GOM-8 7.42 74.5 29.19 72900 1200 300 3.88 Temperature (74OC) in water bath
PERF-19 GOM-9 5.98 23.5 29.01 60908 1200 300 4.10 pH continuously adjusted with cont. HCI
PERFPO GOM-9 7.02 21 .o 29.07 61911 3000 750 4.92 Baseline with triple water & oil
GOM-9 5L-Prep 7.1 --- 65820 --- --- --- 4 L prepared on 11/08 plus 1 L on 11 /I 4/00
GOM-10 2L-Prep 6.7 72900 --- --- ---
GOM-11 2L-Prep 6.8 --- --- 67860 -me _-- ---
PERF-25 GOM-11 50.0 29.09 67860 200 72.8 4.00 No Hoke vessel data taken
: PERF-26 GOM-11 50.0 200.00 67860 200 60.6 4.00 Constant 200 psi
PERF-27 GOM-11 50.0 500.00 67860 200 57.1 4.90 Initial pressure 600 psi; He. added once for
K 500+6OOpsi; final 500 psi
PERF-28 GOM-11 50.0 850.00 67860 200 46.7 4.04 Initial pressure 1 OOO.psi; He added for 600-
>9OOpsi; final 850 psi
PERF-Deep2 25.0 29.31 0.991 --- 0.8377 g oil in 18.025 mL CH2C12
PERF-Deep3 --- 25.0 29.31 0.796 --- 0.6727 g mixed oil in 17.12 mL CH2C12
PERF-Deep4 --- --- 25.0 29.31 1.689 --- 1.4273 g mixed oil in 30.605 mL CH2C12
PERF-Deep5 --- --- 25.0 29.31 --- --- 0.995 --- 0.8407 g mixed oil in 19.76 mL CH2C12
“Process blanks
bRecovery of surrogate standards
 Table A.3 Ec&brium concentrations of
inorganic constituent
Experiment Salinity ‘HCO< CO,‘-
Wm CO ( P P ) (pp ) pH
PERFS 62,487 137 0” 6.97
PERF-7 64,734 163 0 7.02
PERF-8 63,289 164 0 7.24
PERF-9 63,788 101 0 6.99
PERF-10 67,856 77 0 6.25
PERF-11 59,232 139 6 9.03
PERF-12 72,793 27 0 4.73
PERF-13 ” 71,603 32 0 8.09
PERF-14 40,140 123 0 7.6
PERF-15 114,480 89 0 7.14
PERF-16 65,791 124 0 7.09
PERF-17 66,606 74 0 7.14
PERF-18 66,333 54 0 7.26
PERF-19 66,333 21 0 5.98
PERFPO 61,911 128 0 7.02
PERF-25 65,721 NA NA NA
PERF-26 71,772 NA NA NA
PERF-27 79,647 NA NA NA
PEfiF-28 *- 67,050‘ NA NA N/i
.- ,**.. 1;. ,v -. _xI Pj* am.\ .-a.- k.. “6. ,.< . -,
NA = Not atialyied.

Table A.4. Equilibrium concentrations of ICP tietals

Experiment RCRA metals by ICP (ppm) Non-RCRA metals by ICP (ppm)
Se Cd. ” ‘” ‘-.-I’^ ^ S b I Ba __ @ _ , Mg Nci ‘. Sr
PERFB Ua 0.079 f 0.005 co.75 f 0.14 42.3 3200 850 33,400 140
PERF-7 U 0.080 ct 0.006 co.75 f 0.29 44 3300 860 33,800 140
PERFS ’ U 0.080 zt 0.009 co.75 f 0.20 43 3180 840 32,900 138
PERF-9 U 0.080 ct 0.010 co.75 f Q.,l F, 45. 3?4? 860 33,500 144
PERF-10 U 0.080 zt 0.010 GO.75 f 0.15 40 3300 880 34:2&I 148
PERF-11 0.9 z!I 0.4 0.085 k‘O.015 co.75 f 0.15 44 3040 850 33,300 146
PERF-12 l.OkO.6 0.086 zt 0.022 co.75 zt 0.27 41 3150 840 33,300 146
PERF-13 0.7 zk 0.3 0.076 f 0.016 cl .5 & 0.85 37.5 3080 840 32,270 138
PERF-14 U 0.077* 0.015 U 3290 860 17,250 145
PERF-15 0.7 T!z 0.7 0.074’ f 0.020 ‘<1.5*0.7? ii: 3170 823, 49,418 137
PERF-Deep1 3.5 Ik 1.0 U U 26 924 138 7,000 250
PERF-16 U 0.073 f 0.007 cl .5 *to .53 53 3260 853 32,800 140
PERF-17 U 0.075 f 0.019 cl.5 *to.45 44 3280‘__ 864 33,230 142
PERF-18 U 0.066 f 0.004 51.5 f 0.92 41 3320 877 33,600 144
PERF-19 U 0.073 f 0.001 cl .5 zt 0.76 43 3240 834 32,430 138
.PERF-20 U 0.071 f 0.005 cl .5 TJZ 0.32 53 3255 844 32,930 140
“U = undetected above ICP detection limit for tFs elkmeriLO~

55
 Table A.5 Equilibrium concentrations of panic matter by gravimetric md IR analyses
Experiment WSO (ppm), gravimetric Oil (g/kg), gravimetric= VSO (pw-0, IR
TEM fraction HEX fraction TEM fraction HEX fraction I produced water
PERFB 27.0 27.0 0.064 0.048
PERF-7 35 26.3 0.042 0.025
PERF-8 18.0 47.8 0.085 0.176
PERF-9 19.2 33.3 0.0057 0.0074
PERF-10 14.4 7.71 0.0171 0.0071
PERF-I 1 28.9 27.97 0.137 0.078
PERF-12 28.9 28.0 0.068 0.028
PERF-13 14.4 9.00 0.094 0.057
PERF-14 18.6 10.8 0.088 0.034 49.8
PERF-15 16 7.14 0.076 0.024 12.2
PERF-Deep1 mm- --- 673 547
PERF-16 23.4 26 0.11 0.09 10.5
PERF-17 35.2 33.0 0.17 0.13 17.6
PERF-18 46.25 42.9 0.22 0.17 23.1
PERF-19 16.1 10.9 0.08 0.04 8.7
PERF-20 16.8 14.4 0.08 0 . 0 6 10.5
PERF-25 32.2 7.4 0.11 0.02 NA
PERF-26 83.0 37.7 0.33 0.12 NA
PERF-27 39.2 26.1 0.16 0.09 NA
PERF-28 29.3 35.9 0.15 0.18 NA
PERF-Deep2 --- 561 518 NA
PERF-Deep3 434 ‘--- NA
PERF-Deep4 731 537 NA
PERF-Deep5 ___ 636 --- NA

a For produced water samples, an equivalent oil value was calculated by multiplying the quantity of WSO
in the water by the volume of water and then dividing this number by the volume of oil used in the contact.
 Table A.6 Eauilibrium concentrations of WSO in open LC fractions TEM, HE CX, and TSAT
Experiment WSO (ppm) in TEM fraction WSO (ppm) in HEX fraction WSO (mg/L) in TSAT fraction
Name Total c6-410 Go-C20 C20428 Total G-Go Go420 C20428 Total CSCIO Go-C20 C20-&!8

PERFS 11.6 5.35 6.29 0 10.5 2.56 7.98 0.0 0.062 0.062 0.0 0.0
PERF-7 36.8 12.0 24.81 0.030 21.9 6.52 15.35 0.0 0.042 0.04 0.0 0.0
PERF-8 24.8 7.1 15.35 2.293 15.5 0.86 14.68 0.0 0.034 0.03 0.0 0.0
PERF-9 30.0 9.3 20.63 0.000 13.8 3.29 10.51 0.0 0.051 0.05 0.0 0.0
PERF-10 24.4 6.1 18.28 0.000 9.6 2.13 7.45 0.0 0.058 0.06 0.0 0.0
PERF-11 28.6 7.4 21.21 0.004 10.5 0.70 9.76 0.0 Lost Lost Lost Lost
PERF-12 11.9 6.0 5.87 0.014 5.8 0.24 5.57 0.0 0.039 0.04 0.0 0.0
PERF-13 23.5 4.3 19.19 0;038 10.0 0.50 9.46 0.0 0.010 0.00 0.0 0.0
PERF-14 12.4 4.5 7.86 0.064 5.4 0.48 4.86 0.01 0.066 .0.03 0.02 0.008
PERF-15 14.5 5.1 9.34 0.007 6.6 1.21 5.39 0.00 0.253 0.05 0.21 0.000
PERF-Deep1 251,028 139,915 109,871 1241 73,478 9,514 58,658 5,306 66,590 11,991 50,192 4,408
PERF-Std3 0.0 0.0 0.00 0.000 0.0 0.00 0.00 0.00 0.000 0.00 0.00 0.000
PERF-16 16.6 6.8 9.86 0.000 10.6 0.34 10.28 0.02 0.015 0.02 0.00 0.000
PERF-17 21.9 5.2 16.63 0.022 12.6 0.49 12.08 0.01 0.216 ‘0.08 0.14 0.000
2 PERF-18 24.8 2.2 22.57 0.046 15.7 0.44 15.27 0.03 0.000 !O.OO 0.00 0.000
PERF-19 11.7 4.9 5.89 0.920 5.2 -0.18 5.35 0.01 0.005 .o.oo ‘0.00 0.000
PERFQO 20.8 7.6 13.12 0.006 5.9 io.04 5.89 0.00 0.000 :o.oo 0.00 0.000
PERF-25 3.9 2.2 1.7 0.000 3.48 0.59 2.89 0.00 0.016 !o.oo 0.01 0.00
PERF-26 46.4 8.4 37.9 ,0.107 24.36 1.13 23.22 0.0125 0.103 :0.05 ‘0.04 0.01
PERF-27 6.1 3.4 2.7 :o.ooo 2.69 0.42 2.27 0.00 0.015 ,O.Ol 0.00 0.00
PERF-28 4.7 2.6 2.1 0.006 6.01 1.54 4.47 0.0000 0.071 0.02 ‘0.01 0.04
PERF-Deep2 220,140 120,782 80,496 18,863 88,178 5,331 70,619 12,228 83,880 8,751 71,202 3,927
PERF-Deep3 142,268 79,016 52,579 10,673 NAa NA NA NA NA NA NA NA
PERF-Deep4 192,785 98,182 82,188 12,415 50,321 2,528 33,407 14,386 69,514 3,828 63,158 2,529
PERF-Deep5 201 ,116 97,367 88,838 14,912 NA NA NA NA NA NA NA NA
a NA = not analyzed.
 Table A.7 Equilibrium concentrations of U ;O in open LC fractions TARO, TPOL(subtract), and TPOL(GC)
Experiment WSO (ppm) in TAR0 fraction WSO (ppm) in TPOL fraction WSO (ppm) in TPOL fraction (GC)
(subtract)
Total C&,0 Go-C20 C20428 T o t a l &j-C,0 C&&J C20--C28 Total C&O cl0420 c20-c28
PERF6 0.03 0.03 0.00 0.00 11.6 5.30 6.30 0 NA NA NA NA
PERF-7 0.46 0.40 0.05 0.00 36.3 11.52 24.75 0.030 5.4 1.67 3.73 0.016
PERF8 0.15 0.12 0.03 0.00 24.6 6.96 15.32 2.293 NA NA NA NA
PERF-9 0.24 0.21 0.03 0.00 29.7 9.09 20.60 0.000 NA NA NA NA
PERF-10 0.15 0.14 0.01 0.00 24.2 5.91 18.27 0.000 6.2 1.46 4.78 0.002
PERF-11. 0.28 0.27 0.01 0.00 28.4 7.17 21.20 0.004 Lost Lost Lost Lost
PERF-12 0.07 0.05 0.02 0.00 11.7 5.71 5.85 0.014 3.1 1.18 1.92 0.000
PERF-13 0.11 0.09 0.01 0.00 23.4 4.18 19.17 0.038 27.9 3.90 11.78 12.3
PERF-14 0.03 0.01 0.02 0.00 12.3 4.46 7.82 0.057 4.6 1.35 3.03 0.25
PERF-15 0.05 0.04 0.01 0.00 14.2 5.05 9.12 0.007 6.1 2.70 3.37 0.003
PERF-Deep1 8,438 1,098 5,726 1,614 175,999 126,826 53,954 -4,781 14,847 13,555 445 847
PERF-Std3 0.00 0.00 0.00 0.00 0.0 0.00 0.00 0.000 0.0 0.02 0.00 0.000
PERF-16 0.17 0.09 0.05 0.00 16.5 6.68 9.81 0.000 3.6 0.00 3.57 0.004
PERF-17 0.08 0.03 0.05 0.00 21.6 5.11 16.44 0.022 6.7 0.00 6.41 0.314
PERF-18 0.33 0.03 0.23 0.07 24.5 2.16 22.34 -0.028 12.6 0.00 12.46 0.145
PERF-19 0.04 0.01 0.03 0.00 11.7 4.87 5.86 0.920 5.3 0.02 5.29 0.005
PERF-20 -0.01 -0.02 0.02 0.00 20.8 7.65 13.11 0.006 1.1 0.00 1.11 0.005
PERF-25 0.370 0.20 0.17 0.00 5.0 2.16 1.92 0.90 4.07 2.65 1.42 0.004
PERF-26 0.093 0.00 0.05 0.05 46.2 8.32 37.8 0.05 3.43 2.59 0.84 0.004
PERF-27 0.024 0.00 0.01 0.01 6.05 3.41 2.64 0.00 6.25 5.44 0.78 0.027
PERF-28 0.213 0.12 0.07 0.01661 5.09 3.34 1.78 0.00 2.08 0.09 1.44 0.5529
PERF-Deep2 19,700 3,963 13,235 2,502 116,560 106,700 -61,326 2,708 19,645 18,972 673 0
PERF-Deep3 NAa NA NA NA NA NA NA NA NA NA NA NA
PERF-Deep4 35,628 6,619 24,758 4,251 87,643 91,826 -14,377 -4,499 13,297 11,429 648 1,220
PERF-Deep5 NA NA NA NA NA NA NA NA NA NA NA NA
a NA = not analyzed.
 ORNL/TM-2001/78

INTERNAL DISTRIBUTION

l-5. D. T. Bostick
6. D. W. DePaoli
7. R. T. Jubin
8. R. R. Judkins
9. K. T. Klasson
10. H. Luo
11. J. McFarlane
12. S. M. Robinson
13. T. W. Schmidt
14. W. V. Steele
15. C. Tsouris
16. ORNL Laboratory Records-OSTI
17. ORNL Laboratory Records-RC
18. Central’Research Library

EXTERliAL DISTRIWTION

19. D. Alleman, DOE-PTO, P.O. Box 3628, Tulsa, OK 74101

20. Veronica Blackwell, ChevronTexaco, 100 Chevron Way, Richmond, CA 94802-0627
21. Doug Bourg, Phillips Petroleum Company, 6330 West Loop South, Bellaire, Texas 77401
22. Nancy Comstock, DOE-NPTO, P.0. Box 3628, Tulsa, OK 74101
23. Michael Dubrovsky, ChevronTexaco ERTC, 100 Chevron Way, Richmond, CA 94802
24. Art Hartstein, DOE Fossil.Energy, 19901 Germantown Road, Germantown, MD 20874
1290
Sung-I Johnson, Analytical and Environmental Services, 256 RF Phillips Research Center,
Bartlesville, OK 74004
26. Zara I. Khatib, SIEP, SEPTAR-AWW, 2288 GD, Rijswijk, Netherlands
27. William F. Lawson, DOE-NPTO, P.O. Box 3628, Tulsa, OK 74101
28. Nicholas J. Spiridakis, ChevronTexaco ERTC, 100 Chevron Way, Richmond, CA 94802
29. Dexter Sutterfield, DOE-NPTO, P.O. Box 3628, Tulsa, OK 74101

.
.,