ADOBZG CNSTRUCTION ENGINEERING RESEARCH LAB (ARMY) CHAMPAIGN IL F 5 /3

IDENTIFICATION AND QUANTIFICATION OF HYDROCARBON PRODUCTS IN EF--ETC(U)HAY S0 R S VOGEL, W J MIKUCKI. R J FILECCIA

NCLASSIFIED CERL-IR-N-91 NL'II/uuhuuhuhu/I/IIIEEII//IEE/II//EEEII//E/IIIIIIII/I/um

/II//EN/

II 11112 -

,W P P R[OUN l TCARI

construction Uio ttsAm

engineeringresearch My18laboratoryai OiPoltoCotoatMltrIntltos

IDENTIFICATION AND QUANTIFICATION OFHYDROCARBON PRODUCTS IN EFFLUENTS

byR. S. Vogel

Rt. J. FilecciaW. J. MikuckiR. G. Lampo

13 Approved for public release; distribution unlimited. j.~k Ait I.........

The contents of this report are not to be used for advertising, publication, orpromotional purposes. Citation of trade names does not constitute anofficial indorsement or approval of the use of such commercial productsThe findings of this report are not to be construed as an official Departmentof the Army position, unless so designated by other authorized documents.

DESTROY THIS REPORT WHEN ITIS No LONGER NEEDEDDO NOT RETURN IT TO THE ORIGIN4 TOR

UNLM),1 HLU

SECURITY CLASSIFICATION4 OF THIS IMAGE (When Date Fntered)

REPORT DOCMENTATION PAGE BEFORE COMPLETING FR

4 - I. REPORT NUMBER 2.GV CESSION No. 3. RECIPIENT'S CATALOG NUMBER

CERL-IR-N-91 bi4. 5IL_ . TYPE OF REPORT & PERIOD COVERED

'1 DENTIFICATION AND QUANTI FICATION OF HYDROCARBON"PRODUCTS INNTERIM r

7. AUTHOR(&) . ~ --.- ** -8. CONTRACT OR GRANT NUMBER(&)

~ R.S~fogelW. J.Aikucki / q

R. J./Fileccia R. G./;Aampo /

9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT, PROJECT, TASKAREA & WORK UNIT NUMBERS

CONSTRUCTION ENGINEERING RESEARCH LABORATORY -

It. CONTROLLING OFFICE N AME ANO ADDRESSToT

14. MONITORING AGENCY NAME A ADDRESS(i( different fromt Controlling Office) IS. SECURITY CLASS. (of thie report)

Uncl assi fied15.. DECLASSIFICATION/0OWNGRADING

SCHEDULE

16. DISTRIBUTION STATEMENT (of thi. Report)

Approved for public release; distribution unlimited.

17. DISTRIBUTION STATEMENT (of the abstract entered ftn Block 20. ft different from Report)

III. SUPPLEMENTARY NOTES

Copies are obtainable from National Technical Information ServiceSpringfield, VA 22151

IS. KEY WORDS (Continue on revere* side it necessary and identify by block number)

hydrocarbonswaste waterinfrared spectroscopygravimetric analysis

211 AITNAcr (MCithmnae - tvere ebb N naueeopy n~ d Identfiy by block numew)

SThis report documents interim results of a study to (1) investigatethe capability of infrared (IR) spectroscopy for the identification ofspecific hydrocarbon products in effluents from washracks and othervehicle service operations at Army installations, and (2) make a com-parative evaluation of the Environmental Protection Agency's (EPA) gra-vimetric (STORET 00556) and IR (STORET 00§60) methods of quantifying"total hydrocarbons" in wastewater. z

J",~ 17 DTO O O 51oSLT UNCLASSIFIED / s-

Date fntered)

SCCURITY CLASSIFICATION OF THIS PAGE(Whe, Dole Entered)

Block 20 continued.

' Results have shown that IR spectroscopy can provide a basis for theidentification of hydrocarbon products for which reference samples areavailable. In the initial phase of the work, a scheme of sequentialdiscriminant analysis and pattern recognition was used with reference toa file of military specification (MIL SPEC) hydrocarbon products used inservicing wheel- and track-type Amy vehicles. The complexity of IRdata reduction was shown to increase rapidly with the number of hydro-carbon products present in a given sample -- a result of a similarity inthe hydrocarbon components' molecular structure. The use of gas-liquidchromatography (GLC) in conjunction with IR spectroscopy nrovided addi-tional discrimination and simplified data reduction.

A comparative investigation of the gravimetric vs IR methods forthe quantification of "total hydrocarbons" in wastewater showed that theIR method was capable of higher precision than the gravimetric method athydrocarbon concentrations less than 10 mg/k, and could provide informa-tion on chemical composition and consequently the origin of thehydrocarbon product... However, quantification of "total hydrocarbons" bythe infrared method required greater skills and technical background ofthe analyst than the gravimetric method.

p

UNCLASSIFIEDSECURITY CLASSIFICATION OF THIS PAGEW?,en Date Xnter.E),

-" - _.. .. . ._i . ...... . .. ....__.. . . . -p 11

FOREWORD

This report was prepared for the Directorate of Military Programs,Office of the Chief of Engineers (OCE), under Project 4A762720A896,"Environmental Quality for Construction and Operation of Military Facil-ities"; Task Area 02, "Pollution Abatement Systems"; Work Unit 009, "OilPollution Control at Military Installations." The applicable QCR is3.01.OOA. The OCE Technical Monitor is Mr. W. Medding, DAEN-MPO-U.

The report was prepared by the Environmental Division (EN) of theU.S. Army Construction Engineering Research Laboratory (CERL). Dr.E. Novak is Acting Chief of EN.

COL Louis J. Circeo is Commander and Director of CERL, andDr. L. R. Shaffer is Technical Director.

3

4 * ,--

CONTENTS

Page

DD FORM 1473 1FOREWORD 3LIST OF TABLES AND FIGURES 5

1INTRODUCTION............................. ....******....*** 7BackgroundObjectiveApproachScopeMode of Technology Transfer

2 APPROACH................... ..... o ..... .......... 9

3 EXPERIMENTAL PROCEDURES. .. . ............ ..... ... .. .. 13Sample PreparationWeathering EffectsPreliminary Results

4 GAS-LIQUID CHROMATOGRAPHY ...... .. .. .. .. .. .. . ... .. .. .. .. . ... 0 . . .. 27Experimental ProceduresPreliminary Results and Discussion

5 QUANTIFICATION OF HYDROCARBONS IN EFFLUENTS .................... 44GeneralExperimental Procedures and ObservationsFuture Work

REFERENCES 52DISTRIBUTION 59

4

. .S

TABLES

Number Page

1 Hydrocarbon Products To Be Identified in Washrack Effluents 9

2 Characteristic Absorption Peaks of Hydrocarbon Products Usedin Servicing Military Vehicles or Present in Effluents atMilitary Installations 11

3 Optimized Operating Parameters for Beckman IR-20 InfraredSpectrophotometer 14

4 GLC Operating Conditions 28

5 Hydrocarbon Reference Materials for GLC 28

6 Comparative Analytical Results Between Gravimetricand IR Determination of Hydrocarbons in Wastewater 46

FIGURES

1 Relationship Between Performance and Instrumental Variables

in IR Spectrophotometry 14

2 IR Spectrogram of lOW Oil 15

3 IR Spectrogram of 30W Oil 16

4 IR Spectrogram of 90W Oil 17

5 IR Spectrogram of GAA Grease 18

6 IR Spectrogram of Antifreeze 19

7 IR Spectrogram of Brake Fluid 20

8 IR Spectrogram of Hydraulic Fluid 21

9 IR Spectrogram of Solvent Type II 22

10 IR Spectrogram of Kerosene 23

11 GLC Chromatogram of lOW Oil Reference Material 29

12 GLC Chromatogram of 30W Oil Reference Material (Fort Lewis) 30

13 GLC Chromatogram of 90W Oil Reference Material 31

5 _

.4----. . ...

FIGURES (Cont'd)

14 GLC Chromatogram of 30W Oil Reference Material (Fort Carson) 32

15 GLC Chromatogram of Kerosene Reference Material (Fort Lewis) 33

16 GLC Chromatogram of Kerosene Reference Material (Fort Carson) 34

17 GLC Chromatogram of Hydraulic Fluid Reference Material(Fort Lewis) 35

18 GLC Chromatogram of Hydraulic Fluid Reference Material(Fort Carson) 36

19 GLC Chromatogram of Brake Fluid Reference Material(Fort Lewis) 37

20 GLC Chromatogram of Brake Fluid Reference Material(Fort Carson) 38

21 GLC Chromatogram of Diesel Fuel Reference Material 39

22 GLC Chromatogram of Antifreeze Reference Material 40

23 GLC Chromatogram of Solvent Type II Reference Material 41

24 Chromatogram of Field Sample 42

25 IR Cell Holder for Absorption Spectra of Solutions 45

26 Reference Gauge for Determining Sample Volume for TotalHydrocarbon Determination 48

6

4V

IDENTIFICATION AND QUANTIFICATION OFHYDROCARBON PRODUCTS IN EFFLUENTS

1 INTRODUCTION

Background

Operations essential to the maintenance, servicing, and washing ofmilitary wheel- and track-type vehicles generate large volumes of waste-water which contain variable concentrations of hydrocarbon products.Federal regulations require that such hydrocarbon concentrations be mon-itored when they are contained in effluents discharged into waterways;'*military installations could be liable for restoration costs if environ-mental pollution is caused by military-generated outfalls which exceedestablished hydrocarbon concentration limits. 2

Methods for the quantification of "total hydrocarbons" have beendeveloped and approved by the Environmental Protection Agency (EPA);however, these methods cannot be used to identify specific hydrocarbonproducts. 3 Thus, it is difficult to track a given hydrocarben pollutantto its source. Therefore, the U.S. Army Construction Enginee, ing

Research Laboratory (CERL) is attempting to develop a reliable method ofidentifying specific hydrocarbon products in washrack effluent.

Objective

The overall objective of this study is to (1) develop methods thatcan be used to identify specific hydrocarbon products in wastewater fromvehicular service and washrack areas at Army installations; and (2)evaluate the effectiveness of infrared (IR) spectroscopy vs gravimetricprocedures for quantifying "total hydrocarbons" in wastewater.

The objective of this report is to document initial investigationsinto (1) and (2) above.

Approach

This phase of the study included (1) an extensive literaturesearch, (2) IR spectroscopy and gas-liquid chromatography (GLC) ofreference materials and an actual field sample from an Army installa-tion, (3) redesign and fabrication of experimental equipment, and (4)revisions to testing.

* Superscript numbers refer to reference list beginning on page 52.

7

The hydrocarbon products to be included for identification in thisstudy were limited to those products used to service and maintain mili-tary vehicles -- e.g., lubricating oils and greases, hydraulic fluids,coolants, and solvents. Also included for the purpose of identifyingcontributing hydrocarbon pollutant sources other than vehicular serviceareas were plant/animal-derived hydrocarbon types that are present inwastewaters from mess halls or other domestic activity areas.

Scope

The ecological scope of this study is limited to effluent streamsor accumulations which potentially could find paths through surface andground water into aquifers. The study does not include atmospheric com-ponents.

Mode of Technology Transfer

Research findings on the identification and quantification ofhydrocarbon products in wastewater will be compiled and published as aTechnical Bulletin. This research supports on-going work on thedevelopment of source control and treatment systems for wastewaterdischarges from Army tactical equipment maintenance operations. Facil-ity design criteria developed and results of field evaluations will bepublished both as a laboratory technical report and as an EngineerTechnical Letter. True technology transfer will not occur until thesubmitted guidance is incorporated into Design Guide DG 1110-3-80, TOEVehicle Maintenance Complexes.

8

2 APPROACH

The value of IR spectroscopy for characterizing organic specieswith regard to molecular composition and structure is well known. How-ever, the task of translating spectral infonnation on molecular struc-ture so proprietary hydrocarbon products in a mixture of such hydrocar-bon products can be identified becomes increasingly complex as thenumber of constituents in the mixture increases. To limit the possiblecombinations of analyte variables to reasonable numbers, CERL's analyti-cal approach was designed to be specific for and limited to theidentification of hydrocarbon products used at Army vehicular servicesareas (Table 1).

Extensive research and development efforts by other agencies andindependent groups have been directed toward methods of analyticallycharacterizing petroleum products. A number of techniques have beenused, including infrared spectroscopy,' - " absorption and fluorescencespectrophotometry,1 7 - 2 2 GLC, 1 6'2 3 -

41 and trace-elemental analysis.'

, 1 6 ,;,43

Reviews of recent applications of these methods are given by Bentz;Brown, et al.; Kawahara; Adlard; and Gruenfeld. 7 '3 2 ,

'9'4,4

Table 1

Hydrocarbon Products To Be Identified in Washrack Effluents

Specific GavityDescription (25 /25

low Oil 0.880

30W Oil 0.904

90W Oil 0.899

GAA Grease

Antifreeze 1.125

Brake Fluid 1.0149

Hydraulic Fluid --

Solvent Type II 0.800

Kerosene 0.808

9

i ,~~ , -AW '

However, the general focus of these studies was identification of acrude oil or petroleum derivative as the single component in an acciden-tal or clandestine marine spill. Former work in this area has notaddressed itself to identifying multicomponent mixtures of hydrocarbonproducts.

Three methods of IR data reduction were evaluated for use in thisstudy:

1. The multivariate statistical approach effectively used by Matt-son , et 5al., to fingerprint a given crude oil specimen to determine itssource. 0However, with equipment initially available at CERL, thismethod was not readily adaptable to mixtures of hydrocarbon products inwhich the relative concentrations of the constituents were variable.

2. Linear discriminant function analysis used by Kawahara and Yangto discriminate between given specimens of a single petroleum product toidentify the geographical origin of the niaterial.51 This method wasalso found not to be directly applicable to mixtures of proprietaryhydrocarbon products.

3. Work by Vogel in identifying and quantifying trace constituentsin a complex matrix by means of a limited number of appropriately chosenspectral band combinations unique to each given constituent. 2 Thismethod seemed adaptable to the CERL study, and seemed likely to providesufficient information to allow discrimination among the hydrocarbonslisted in Table 1.

To test how this approach might work in practice, an "evaluation-ary"I experimentation procedure was used. First, a data file of IRreference spectra from samples of nine military specification (MIL SPEC)materials obtained from the Logistics Center, Fort Lewis, WA (Table 1)and reference spectra from an animal- and a vegetable-derived oil suchas those used in food preparation were recorded under controlled condi-tions. Table 2 summnarizes a peak-by-peak log of these reference spec-tra. This log was searched for absorption bands at wavenumbers uniqueto each given hydrocarbon material. As was expected, the similarity inmolecular structure of a majority of the products was reflected in boththe universal occurrence of the CH2 (2930/cm) band used for quantifying"total hydrocarbons," and the similarity of the majority of the otherprincipal bands. Consequently, minor peaks were required to provide thenecessary spectral information for differentiating the hydrocarbon pro-ducts.

A major portion of the initial experimental work focused on optim-izing the operating parameters of the available IR spectrophotometer.Such optimum parameters were essential to insure that the absorptionspectra of the reference materials would be accurately recorded.

Spectral data from an unknown sample were correlated with thatof standard reference materials, following a scheme of sequential

10

Iz

4-

°- "I 2U

2Un An g

>0

IV z

--' 1) M

-- t IN LIn Z.. n2

4- E

4)LU

- 4.

z z jm w A 4A3111

C') 0

a 4-) 24-) 0L (00 u

0 1:!' CL Ini 0 m3

a)o

(fl~4 4.0L

1-.n _ ' D In InQ

4.) 0 K m

0.c a% OD ell 5

2!~)

discriminant analysis and visual pattern recognition. Operations con-sisted of several stages of class characterization with reference tomajor peaks (group frequencies), followed by correlation of minor peaksby pattern recognition of intensity ratios in the unknown samples to thereference materials. By reducing the number of possible candidates inTable 1 through preliminary characterization, the probability of identi-fying or matching multiple components in the unknown sample with theknown reference materials was enhanced. This multistage discriminantanalysis approach seemed to lend itself to data reduction through eitheran appropriate computer program or a simple keysort card technique.

12

4 -L__ __

'7 EXPERIMENTAL PROCEDURES

The Beckman Model IR-20 Infrared Spectrophotometer, used during theinitial phase of this study, is an automatic-recording, filter-grating,double-beam instrument. Spectral information is presented as an analogplot of percent transmission vs wavenumber in the spectral range between4000 to 240/cm on a formatted strip chart. Sample and reference cellsused for hydrocarbon characterization were of the sealed demountabletype, with Luer fittings and potassium bromide (KBr) windows spaced at a0.05-nun path length.

Since most of the spectral information for hydrocarbon characteri-zation would be obtained from minor peaks of greater than 75 percenttransmission (while quantification of hydrocarbons would involvetransmission measurement of peak and baseline values between 15 and 90percent transmission), instrumental variables of the IR spectrophoto-meter needed to be critically balanced to record both major and minorpeaks with minimum distortion and with high photometric accuracy andprecision. To obtain this required performance, the four major instru-mental variables (slit program, gain, response time, and scan rate)required systematic optimization. The interrelationship of these fourvariables on instrumental performance is shown in Figure 1. Optimiza-tion of these four variables involved a suitable balance between resolu-tion, noise level, photometric accuracy, and time-related efficiency.Since an accurately represented spectrum is needed for intensity-ratiomeasurements (to identify hydrocarbons) and for peak height measurements(to quantify hydrocarbon concentrations), systematic efforts to estab-lish instrumental parameters to maximize the accuracy of the recordedspectra were essential.

The procedure for optimizing the values of the instrumental parame-ters was an experimental one. Initial, realistically chosen values weresuccessively refined to obtain the following performance criteria:

1. Spectral resolution of 4 to 7 wavenumbers/cm

2. Photometric distortion of 2 percent relative

3. Peak-to-peak maximum noise level of 0.5 percent transmission.

These experimentally established operating parameters, summarized inTable 3, were used as the basis for all succeeding spectral recording.

With the instrumental settings so established, absorption spectrawere recorded for the nine MIL SPEC materials listed in Table 1; spectraof samples of vegetable-derived oil (safflower oil) and animal oil(lard) were also recorded. These absorption spectra are shown in Fig-ures 2 through 10.

13

(') Spectral Resolution EwNs

(2) Noise a rI/2

(3) Servo Energy a w g

where E is servo loop energy

w is monochromator slit width

N is noise

s is wavenumber scan rate

r is response time

g is gain setting

Figure 1. Relationship between performance and instrumental variablesin IR spectrophotometry.

Table 3

Optimized Operating Parametersfor Beckman IR-20 Infrared Spectrophotometer

Mode: Double beam

Spectral Range: 4000 to 250 cm-I

Absorption Cells: KBr plates spaced at 0.05 mm

Slit Program: I

Gain: 6

Period: 1-second time constant

Scan Time: 240 minutes (17.2 cm 1/min)

14

£:-

-- - I-- -

-I

N -

N I 0

I .U

iN. I

N - - -o - 1 - - - - - - - -o

* I

.4-3

03303d

16

22 ao~;ernswvwL Lk32?~3d 2 ~

U.---- ----- I

-Iif

a- - - - - - - - ~0

0

-- 02S

S9-o

- - 0* 0

- - --- - - 0

o~ 0 0 0000...0......0.0 4-o a a 9- 5 5 ~ -~

S

* ** U 'U- -~ ~

- hi- UU

4JUa,0.(n

0 - -E

b~4

S

a)-~ - - - - - - 1..

- - =- - - - - - -- - 0~

.9-Li~

S

NOISIINSNVMI iN2~M1d

80

'9- ii

17

.1

p____________________________ ___ ~

33

2 -0 9 0 O

N0I --S-V-l - -ot-

..... ...-

19.NOII1* 4;v kN3 H.

k

0

I ans~, I IMI

r s ~a2~ 2 9

* - - -+J

NJ I I I

- -- - -

21L

!IM

z a

N - -- - -ca_

I - - - - - - -.

0' -- - - -~~w~

-004

22-

-4

IN31U14

oi

23-

Although the recording of IR spectra is straightforward, carefulattention to established operating procedures dnd techniques of cellmanipulation was necessary to insure valid and reproducible results. Toavoid cross-contamination between samples, cells were flushed with fivesuccessive 1O0-vi portions of Freon 113. Cells were vacuum dried aftereach spectral scan. Since Luer fittings are potential reservoirs fortraces of a previously contained sample, meticulous attention was paidto both external and internal surfaces of the Luer fittings and to thecell windows when cells were rinsed between samples.

It was determined that the conventional method of using a syringeto inject samples into absorption cells entailed an unacceptable risk ofcross-contamination (in addition to contaminating exterior surfaces whenhighly volatile solvents such as Freon were used). Therefore, animproved, though unorthodox, manipulation technique was chosen. Cellswere loaded using a 100-0.Z Eppendorf pipette assembly; the pipette'splastic tip provided a liquid-tight connection with the capillary ori-fice at the base of the Luer fittings. The calibrated 100-lik volume didoverfill the 0.05-pin pathlength cell, and the disposable pipette tipinsured against sample cross-contamination. The KBr cell windowsremained matched within experimental error over, the duration ofspectral-data gathering.

Sample Preparation

Field hydrocarbon samples were typically skim samples consisting ofan oily phase containing solid debris and occluded water overlying anaqueous phase. Since even trace amounts of water cause strong spectralinterference and etch alkali-halide cell windows, complete separation ofthe organic phase from the aqueous phase and debris was necessary. Thiswas accomplished on a physical basis, without the addition of solvent,by centrifugation of 2- to 5-mZ portions of a sample in tapered 20-wrtubes in a clinical-type centrifuge. A portion of the supernatantorganic layer was then carefully drawn off with a Pasteur pipette. Thecycle was repeated until a clear organic phase was obtained. Gravityfilterin~g of the separated organic phase through anhydrous sodium sul-fate reduced the remaining traces of water to levels that could not bedetected in the recorded spectra.

The conventional method of extracting the organic solvent from theoriginal sample, followed by evaporation of the solvent, was not usedbecause of (1) the inevitable addition of interfering solvent peaks ifevaporation was performed at low temperatures, and (2) the loss of rela-tively volatile fractions of the organic phase -if the solvent evapora-tion was conducted at higher temperatures. Residual solvent peaks werepartially cancelled by using differential spectroscopy with pure solventin the reference beam in a variable pathlength cell, except when theconcentration of residual solvent in the sample extract required a cellpathlength less than the mechanical limit of the variable pathlengthcell.

24

ma etr u-ae ,I avgti the presence ofeach other are quantified by evluatiAny ;& L oocrption measurements

diglycerides, triglycerides, and other f2. 'I 1 Ctcrs which consti-tute a significant portion of iThe ve,-,I. tw- atr i& -derived oils andwhich are essentially absent i~ 's airE representedat an absorption band at 1745 curl- ~ 'le uroleu and1vegetable/animal-based oils absorb a, tic ,e: jer of ?903 cm- ; thismeasure is used in the official EPA rof "totai hydrocar-bons. 111 ) Petroleum-b'asedi oil s .an 'j T i-n, vegetaole!animal-based oils in cin 111Known 3a2 zu1,-)* cKm mea Sur ement sare made by simply contacting an (o17 (- e-J in CC] with sil-ica gel. Vegetable/aninial-base6 .wc '--veiY adsorb~d;* thepetroleum-based oils remain in s,)flti;

Weathering Effects

When oils or other hydrocart'urs )6r : rend inrtc an aqueouseffluent and are exposed to the atmos:,f 2iws 10IS or films overly-ing an aqueous phase, changes in physi_ f-n'rnmnical characteristicsare caused by the effects of sol,,- irre'oLcn, 1einperature, agitation,adsorption on solids, bacterial ciction, jn. ceeia interaction withother chemical compounds in the effluent 'r- so-cal led "weathering"must be taken into account in arac~ s .ba sedo on comparison ofstandard reference materials with the -a, 1 .xore-ts in a recoveredsample that has undergone environmerLJ ' l The changes thatoccur in weathered hydrocarbons are r& - Kn tne differences betweenthe analytical response of a giver 0 ,' ,';trolleum product andthe same product present in a sample c ~~c r~all etfluent stream.

On the assumption that a st.- ! : 'wu is approached in theweathering process, techniques to e;-, -'r( ,renLal weathering havebeen developed. The U.S. Coast Guard, N-IG-riient of the Interior,the EPA, and others concerned with the n rih of crude oil andpetroleum products in marine and terresi ] erv-'cnents have devotedconsiderable research and development Pffort to stueilas of the composi-tional changes caused by weathering, and~ ito ".he development of tech-niques for simulating weathering of reterenca samples to compensate forthe environmental weathering effect on tne .wknoiwn sample. '5 ,1 1 6 3

The reliability of analytical f in(,erpri ri 1,-n methiods improves whensynthetically weathered standards are usc ) identify recovered crudeoil or petroleum products. Methods deve'.oped aind adopted by the EnergyResearch and Development Agency"* nd ,he JAne e-I(- Society for Testingand Materials (ASTM) for the synthetic wceather'ng of crude oil involve a

* adsorbed =bound on particle surface-_**Superseded in 1977 by the Department w~ tnrivy.

25

distillation (under reduced pressure) of the suspect and recovered oilsamples, and collection of the 275 C (atm) fraction; this fraction isthen analyzed. 17,22 This distillation technique rejects the light frac-tions of petroleum components having n-paraffin carbon numbers of 15 orless. This loss of the low carbon-number fraction does not hinder theidentification of crude oils containing a homologous n-paraffin seriesextending to CA0 or more. However, such synthetic weathering methodsmay be unsuited to the synthetic weathering of recovered samples thatmay contain diesel fuel, solvent, and other light hydrocarbon productssuch as might be encountered in effluent streams at Army installations.More applicable synthetic weathering techniques must be found soenvironmental weathering can be simulated without extensive loss oflow-boiling hydrocarbons. Ultraviolet irradiation and aeration of stan-dard reference materials at temperatures below 400C offer potentiallyuseful alternatives.

4' s54

Prel iminary Results

Two effluent stream samples were analyzed by IR spectroscopythrough spectral correlation of absorption peaks with those of the knownreference materials listed in Table 2, using discriminant analysis andpattern recognition of analog data. In a field sample (a contaminatedoutfall sample from Flora Road outfall, Fort Lewis, WA), two consti-tuents -- motor lubricating oil and diesel fuel or kerosene -- wereidentified by IR spectroscopy and confirmed by GLC as the principalhydrocarbon products present. Other minor peaks only slightly abovethreshold intensity were observed, but were not sufficiently defined tobe analytically useful, since the Beckman IR-20 Spectrophotometer didnot have the capability for scale expansion in the analog presentationof absorption values. Facilities for scale expansion and analog-to-digital data conversion would be required for accurate representation ofminor peaks and for peak-rationing techniques.64

26

'9

i Ll GAS-LIQUID CHROMATOGRAPHY

Over the past 10 years, GLC has gained a prominent place in analyt-

ical methods which fingerprint crude oil samples recovered from oilspills.60 This has been due, in part, to the unique ability of GLC todifferentiate between hydrocarbon species having similar structures.Petroleum crudes have a predominant component of n-paraffin hydrocarbonsin an homologous series of carbon numbers to C40 or more. The chromato-gram of a typical crude petroleum sample is characterized by a periodicsequence of approximately equally spaced peaks representing unit incre-ments in carbon numbers of the n-alkanes contained in the given sample.The relative concentration of each n-alkane is a function of the areaunder the corresponding peak; information about peak height distributionof the respective n-alkanes allows recovered crude oil samples to bematched to known reference samples of the suspect petroleum.6"

GLC was used during this study to complement IR spectroscopic tech-niques. The GLC carbon-number criterion made it possible to identifyboth the low-boiling n-paraffin fractions (carbon numbers less than 11)which characterize kerosene and diesel fuel and the higher-boilingn-paraffin fractions which constitute motor lubricating oils. However,weathering of n-paraffin series hydrocarbons results in partial loss oflower-boiling components; this loss is reflected in an attenuation orabsence of the lower carbon-number peaks. As in the case in IR measure-ments, weathering effects must be taken into account when chromato-graphic data on field samples are interpreted.

Experimental Procedures

GLC information was collected using a Hewlett-Packard 5754 Gas-Liquid Chromatograph equipped with temperature programming and a flameionization detector. Data output was through a Hewlett-Packard 3380Arecording integrator which provided (1) an analog plot of the chromato-gram (with retention times at each recorded peak), and (2) a table ofthe respective peak areas as percent of the total (for quantifying indi-vidual molecular species).

Optimum GLC operating conditions were established by experimentingwith various liquid phases and values of instrumental parameters withrespect to the sample materials to be analyzed (Table 4). Althoughother liquid phases might have given equal performance, Dexsil 300 waschosen for its thermal stability and its capability to provide reason-able resolution of peaks in a real-life sample.

After these analytical conditions were established, chromatogramswere obtained from the 13 reference materials listed in Table 5. Therecorded chromatograms appear in Figures 11 through 23.

27

- a - ---

Table 4

GLC Operating Conditions

Instrument Type: Hewlett-Packard 5754

Detector: Flame ionization detector

Injection Port Temperature: 2500C

Detector Temperature: 3000C

Column: 365-m x 3.2-ram stainless steelcolumn packed with 6 percentDexsil 300 on 80/100-meshChromosorb WHP

Column Oven Temperature: Proggammed from 70 to 3250Cat 4 C~minute. Heldat 325 C

Carrier Gas and Flow Rate: Helium at 30 cc/minute

Electrometer Range: 105

Chart Speed: 0.5 cm/minute

Sample Volume Injected: 1 to 5 it as required

Table 5

Hydrocarbon Reference Materials for GLC

Reference Sample Source

lOW Oil Fort Lewis30W Oil Fort Lewis90W Oil Fort Lewis30W Oil Fort CarsonKerosene Fort LewisKerosene Fort CarsonHydraulic Fluid Fort LewisHydraulic Fluid Fort CarsonBrake Fluid Fort LewisBrake Fluid Fert CarsonDiesel Fuel Fort CarsonAntifreeze Fort LewisSolvent Type II Fort Lewis

28

04'i Ji,

ITI

14

4244

451

4T.85

54.37

58.35L.30

63.00

Figure 11. GLC chromatogram of lOW oil reference material.

29

12 20

is to

1? 72

4931

4S 21

4"

9 1

S2 63-

54. $9

62 7?63 74

65.01

Figure 12. GLC chromatogram of 30W oil reference material (Fort Lewis).

30

T- NJ

tzart

19AI20.29

sastoZtss

52.76

54.4.36303

544

43.51

54332

555.5

$724

.4

64.41

Figure 13. GLC chromatogram of 90W oil reference material.

31

29 3

32. 31

32. I?

41. 46

42 17

438

41 16of 3

44; 94'

474 '54

46 a49 2?

/i.

2 16

2 2

7 46 7

62 7?

63 82

STOP

Figure 14. GLC chromatogram of 30W oil reference material(Fort Carson).

32

6 20 3.636 66

0 90966

(Forti Lews)

33 A

-. . . --- -___ ____ ___ ____ ____ ___ .-2- 25

-1 NJ

9 '

,"J 1212. 75

1)22 54 2.42

25j. 4q

29. 66

32-43

3.4

3514

037.74

0*26

4734

Figure 16. GLC chromatog ram Of kerosene reference material(Fort Carson).

344

-L2 43

3-. 46

6.I 24

1 65

(Forti Lews)

35 _______

4--3

14

as 7t

22 23

S £024 42

24. 40

&21. 28

soo

4. 22

SO 96

If. L| 1;

64 41

23 35

46 42

Figure 18. GLC chromatogram of hydraulic fluid reference material(Fort Carson).

36

2 63

4 72

17 S2

22. 40

p. 2

-7. 20

igloo

Figure 19. GLC chromatogram of brake fluid reference material(Fort Lewis).

37

4 S

22 3?

4. 44

37. 1

at. 7

Figure 20. GLC chromatogram of brake fluid reference material(Fort Carson).

38

-___

tit,

4f 52

12 1?

214 31

16 13

89

9215

23 2424

26 46

29. 3.46'2

2 07

3::2

D37o34

Mal• 39*06S40*794140r42-20"

44.1

_4*06

STOP

Figure 21. GLC chromatogram of diesel fuel reference material.

39

Id 2

12 40

47. 23

I TOP

Figure 22. GLC chromatogram of antifreeze reference material.

40

ELI

STOP

Figure 23. GLC chromatogram of solvent type 11 reference material.

41

19 41

21 92

-11

42

A -~- - ~ . ~ '~ - -4g

Preliminary Results and Discussion

At this phase of the study, visual pattern-recognition was used tocategorize component classes. The presence of n-alkane homologousseries was apparent in the kerosenes and lubricating oils in which thewell-defined series of peaks are superimposed on an "unresolvedenvelope."6' The lower carbon-number distributions in the kerosenes(with respect to the lubricating oils) were correlated to the respectiveboiling point ranges of the kerosene and motor-oil products. The dis-similarity between Fort Carson and Fort Lewis hydraulic fluid chromato-grams denote different chemical compositions, although, according to MILSPECS, the physical properties of these substances are the same. Thebrake fluid samples from the Fort Carson and Fort Lewis stores producedsimilar chromatograms, indicating that they had the same chemical forma-tion. The "tailing" of the last three eluted peaks indicates highlypolar compounds such as glycols or higher alcohols. The absence of theperiodic-peak series of the n-alkanes in the chromatogram of solventtype II indicates it is not composed of n-paraffin hydrocarbons.

The field sample was used to test the analytical interdependence ofGLC in combination with IR spectroscopy. The GLC operating conditionsused in the preliminary tests on this sample were as given in Table 4,except that temperature programming was between 100 and 3250C at a ramprate of 100C per minute.

The outfall sample chromatogram (Figure 24) had a periodic seriesof peaks, indicating an homologous series which corresponded to then-paraffin series in kerosene. Subsequent analysis by GLC/mass spec-troscopy confirmed these peaks to be an homologous series of n-alkanesranging from C14 to C24. The attenuation or absence of peaks for carbonnumbers less than 12 in the field sample chromatogram (as compared topeaks of carbon numbers less than 8 in the reference sample of kerosene)indicated the expected losses of these lower-boiling n-alkanes throughevaporation in the weathering process.

Since a sophisticated data-reduction system which would generatespecific compound information was not available, GLC chromatogram pat-terns were visually screened to establish what classes of organic com-pounds may or may not be present. This step simplified the interpreta-tion of IR spectra wen the two techniques were used together, since apreliminary GLC screening reduced the population of probable componentsin a mixture of hydrocarbon products.

43

__ __ --

4

5 QUANTIFICATION OF HYDROCARBONS IN EFFLUENTS

General

The determination of oils and other hydrocarbons in effluents atmilitary installations can provide information essential to the monitor-ing and control of environmental pollution by these substances.

CERL used the official EPA gravimetric method (STORET 00556) todetermine "total hydrocarbons" in wastewater. This method is based onthe weight of the residue on evaporation of a Freon extract of a givenvolume of a wastewater sample. Analytical results are expressed as mil-ligrams of oil per liter. The stated analytical range is 5 to1000 mg/t.

An alternate EPA method (STORET 00560) uses IR spectroscopy and isbased on the linear relationship of optical absorbance by the CH2 bandat 2930/cm, with concentration of the hydrocarbon.65 The measuredabsorbance of a Freon extract of the sample is converted to concentra-tion of "total hydrocarbons" in the original sample by means of ananalytical curve prepared from absorption measurements of prepared stan-dards of known concentrations of a standard oil in Freon. The analyti-cal range is 0.1 to 40 mg/Z with conventional absorption cells. Theupper limit can be extended several orders of magnitude by volumetricdilution of the Freon extract.

The EPA method using IR spectroscopy can provide information oncomposition as well as concentrations of the hydrocarbons, and thereforemight be more appropriate in some cases than the nonselective gra-vimetric method.

Before CERL could evaluate the overall benefits and limitations ofthe IR and gravimetric methods with regard to precision, accuracy, pro-cedural characteristics, and applicability to routine use, the BeckmanIR-20 Infrared Spectrophotometer had to be retrofit with quartz-windowcells of 1-, 10-, and 50-mm pathlengths for absorption measurements ofsolutions. The necessary cell holders of the dimension stabilityrequired for this purpose were not commercially available, and so proto-types were made at CERL (Figure 25). These prototypes were designed toaccommodate the range of cells needed for both the initial phase of thisstudy, and for planned investigations into low concentrations of hydro-carbons that would require relatively long pathlength cells.

Next, a "hands on" stepwise execution of the EPA IR procedure wasconducted; critical attention was paid to procedural details of prepar-ing and calibrating standards, development of the analytical curves,sample preparation, extraction techniques, absorption measurementmethods, data reduction, and calculations.

44

- 4 - - -- ---- ____ ___ - - ---- ---- ----

4J

E 0

00

(n 4-.)

5n u

4..

0E o E

(0 4-)E+1 "

dc E N+N v 0

01))

EE

5-

ON45

The EPA IR method was then applied to the determination of hydro-carbons in a representative group of wastewater samples previouslyanalyzed by the gravimetric method. Comparative results of the initialgroup of samples are listed in Table 6.

After a sufficient number of analytical results are accumulated(during the second phase of this study), these data will be analyzed andcorrelation statistics will be computed to assess bias between the IRand gravimetric methods. Accuracy and precision of the IR and gra-vimetric methods will be assessed through the use of synthesized waste-water samples containing known concentrations of the MIL SPEC hydrocar-bon products (Table 1).

Experimental Procedures and Observations

The procedural details of the EPA IR method follow the solventextraction and IR techniques used in the petroleum industry for monitor-ing refinery wastes and effluents.1 6 Acidification of the sample to pH 2

Table 6

Comparative Analytical Results BetweenGravimetric and IR Determination of

Hydrocarbons in Wastewater

Concentration (mg/Z)

Sample Number Gravimetric Method IR Method

FB-4a 10.2 11.5

FB-4b 2.0 8.2

FB-5 11.0 9.5

FB-6 9.7 11.9

FB-8 5.6 6.1

FB-1O 1.0 1.9

FB-15 21.5 23.5

FB-18 32.7 32.0

*Standardized with reference to 30W MIL SPEC motor oil.

46

- _ _ _ _ _ _ _ _ _ _ _ _ _ _ -

or less with sulfuric acid at the time of sampling arrests biochemicalchanges and increases hydrocarbon extraction efficiency.4 9 The use ofFreon 113 as the extractant to replace the previously used carbon tetra-chloride was a result of increased awareness during recent years oftoxic substances, rather than for chemical reasons. The slight incon-venience of using the more volatile Freon 113 instead of carbon tetra-chloride is compensated by the 100-fold lower toxicity of Freon relativeto carbon tetrachloride.

The cell holder design shown in Figure 26 provided accurate andreproducible alignment of cylindrical cells of 1 to 10 cm pathlengthwith the optical axis of the Beckman IR-20 spectrometer. Preciselymachined v-ways were used instead of conventional clamping devices tomaintain accurate and reproducible positioning of the cells. This per-mitted convenient access for cell-filling operations and minimizedmechanical stresses on the cells during their placement and removal.

In the selection of hydrocarbon compounds for preparation of thereference "'oil" standard, the EPA JR procedure categorized hydrocarbonsinto two groups:

1. Known oils. These are oils in which the composition of thehydrocarbon material in the wastewater is known and a sample of thematerial is available for preparation of calibration standards.

2. Unknown oils. These are oils for which a sample of thematerial i., not available for the preparation of standards.

In unknown oils, a hydrocarbon mixture consisting of 37.5 percentn-hexadecane, 37.5 percent isooctane, and 25 percent benzene (volume/volume basis) is given as the calibration material. However, the0.769 specific gravity of that mixture is considerably less than the0.900 average specific gravity of hydrocarbon products listed in Table 1that might be present in washrack effluents, and consequently, wouldintroduce a systematic error of approximately +18 percent in the analyt-ical results. Therefore, it was decided to prepare calibration stan-dards from a material of approximately 0.900 specific gravity. A MILSPEC 30W motor oil with a 0.904 specific gravity was selected as thestandard material; calibration standards were prepared over the rangebetween 0.5 and 40 mg/2,.

Oil calibration standards were prepared according to conventionalpractices for serial dilutions of the MIL SPEC 30W motor oil with Freon113. To minimize evaporation losses, the prepared standard series werestored in volumetric flasks sealed with latex caps over ground glassstoppers. Absorption spectra of the standa-rds in the region of the2930/cm band were recorded using the integral strip-chart recorder ofthe Beckman IR-20.

Data reduction required the conversion of the percent transmissionreadout of the IR-20 to absorbance. This was done graphically by means

47

55 CAP SCREW PRSFI

6 -3 2 kl /4 d. HA RD ALU M INUM

TYP 4-40 --

SCREW B *~

3/,3x 2/4 LUG I----- 4FLAT FOR

SPRNG I DIAL CALIPER

LOADED BALL - 4 A-O RFL

ARM LENGTHS AS-- </-A-O RFL

INDICATED -_

0.125" SS RO-1

SLIDING FIT

5TAPER

Figure 26. Reference gauge for determining sample volumefor total hydrocarbon determination.

48I

of a linear-function plot of percent transmission vs absorbance (%T vsA) on semilog paper, with absorbance on the linear axis and percenttransmission on the logarithmic axis. The plotted intercepts weretransmission = 100 vs absorbance = 0 (T 100 vs A = 0), and transmis-sion = 10 vs absorbance = 1 (T = 10 vs A = 1). Peak-height absorbancevalues were obtained from the strip-chart recording by subtracting thebase-line absorbance value from the total absorbance value at peakcenter. The analytical curve of absorbance vs peak height was plottedin the conventional manner. In plotting the analytical curve, the useof log-log instead of the conventionally used linear-coordinate graphpaper provided constant relative precision over the calibrated rangewhen graphical data reduction methods were used.

Wastewater samples for oil determination were received in l-quart"Mason jar"-type containers. Sample volume was established by attachinga tape reference mark on the jar at the liquid level, transferring thecontents to a separatory funnel, rinsing with portions of Freon, refil-ling to the tape mark with water, then pouring the contents into a1000-mt graduated cylinder and estimating the volume by visual interpo-lation between the 10-mk graduations on the cylinder. The uncertaintyin establishing the liquid level in the sample ,jar and graduatedcylinder through parallax and meniscus errors in routine operations wasfound to be approximately +2 mm -- equivalent to +20 mR of samplevolume. The resultant error contribution to the analytical results wasapproximately ± 6 percent (for a sample volume of 300 mk).

A CERL-designed device was used to directly determine samplevolumes through a liquid level vs volume calibration. This device pro-vided both a measurement of sample volume four times more precise thanthe prescribed method and a significant reduction of operator time. Todetermine the sample volume, the device is placed on the Mason jar rim,then the depth from the rim to the liquid surface is measured with avernier caliper integral with the device, and the sample volume is readdirectly from a previously prepared plot of sample volume vs measureddepth from rim to liquid level. The relative error in the use of thismethod was approximately 1.5 percent for typical sample volumes.

The gravimetric method requires evaporation of the 100 mi of theFreon solvent from the extracted hydrocarbon at a temperature of 70 Cand drying at 800C. This results in partial losses of analyte fractionssuch as gasoline and other products having significant vapor pressure atthose temperatures. Since the IR method does not require evaporation ofthe solvent from the hydrocarbon extract, losses of low-boiling hydro-carbon fractions are significantly lower than by the gravimetric method.

The determinative step of the gravimetric method requires oil resi-dues to be weighed in 125-mk distilling flasks. The uncertainty ofreproducing the tare weight of a vessel of this mass and volume afterthe necessary handling, evaporation, and drying operations is of theorder of 10 mg, or approximately 200 percent or more of the reportedvalue at oil concentrations below 5 mg/t. While the relative precision

49

-' --. . . .__ __ _,,__ _... ..____ __ ___ __ __

of the gravimetric method varies inversely with oil concentration, therelative precision of the IR method is essentially constant over thecalibrated range.

CERL's comparison of the gravimetric and IR methods included a con-sideration of the degree of expertise required to perform the analyticaloperations. The sampling and extraction operations of the two methodswere identical -- assuming the EPA-prescribed separatory funnel tech-nique was used. The weighing operations of the gravimetric method hadto be carefully done to minimize tare changes in the flask containingthe oil residue. However, since the determinative steps of the gra-vimetric procedure involve only an analytical balance, the procedure canbe successfully performed by analysts who have minimum technical back-ground. The infrared method, however, demands considerably more skilland technical background of the analyst. Technical judgment in select-ing and preparing standards, geterating analytical curves, establishingoperating conditions and maintaining performance levels of the IR spec-trophotometer, interpreting spectra, and recognizing spectral aberra-tions when they occur require that these operations be performed orsupervised by a person competent in spectrochemical methods.

Future Work

The second phase of the study will generate enough comparativeresults from gravimetric and IR determinations to permit a statisticalcomparison. Phase 2 will also evaluate the effect of differences inabsorptivity of specific hydrocarbons on observed concentrations by theIR method compared to the gravimetric method. Stability of spectralabsorptivities of the various materials in Freon must also be examined,to assure validity of prepared standards stored over a given time inter-val. An analysis of variance of the procedural steps of the IR methodwill provide information on the relative attention required in sampling,extraction, solution preparation, standardization, and photometry foroptimizing cost-effective operations.

50

--. A. . ,mTTOW

6 CONCLUSION

As an initial step in developing methods for identifying hydrocar-bon products in Army vehicular service and washrack effluents, IR spec-troscopy and GLC methods were evaluated and tested.

Results have shown that IR spectroscopy may provide a basis for theidentification of hydrocarbon products for which reference samples areavailable. In the initial phase of the work, a scheme of sequentialdiscriminant analysis and pattern recognition was used with reference toa file of MIL SPEC hydrocarbon products used in servicing wheel- andtrack-type Army vehicles. The complexity of IR data reduction was shownto increase rapidly with the number of hydrocarbon products present in agiven sample -- a result of a similarity in molecular structure of thehydrocarbon components. The use of GLC in conjunction with IR spectros-copy provided additional discrimination and simplified data reduction.

A comparative investigation of the gravimetric vs IR methods forthe quantification of "total hydrocarbons" in wastewater showed that theIR method was capable of higher precision than the gravimetric method athydrocarbon concentrations less than 10 mg/k, and could provide informa-tion on chemical composition and, consequently, the origin of the hydro-carbon product. However, quantification of "total hydrocarbons" by theinfrared method required greater skills and technical background of theanalyst than the gravimetric method.

51

4- -- -u ___ __ __ ___ __-_ ___ _ _ .

F7

REFERENCES

1. The Clean Water Act of 1977, Public Law 95-217 (91 Stat 1566).

2. Methods for Chemical Analysis of Water and Wastes, EPA-600/4-79-020, PB297686 (Environmental Protection Agency [EPA], March1979).

3. Bentz, A. P., "Oil Spill Identification," Analytical Chemistry,Vol 48, No. 6 (1976), p 454A.

4. Bentz, A. P., Oil Spill Identification Bibliography, ADA 029126(Coast Guard Research and Development Center, March 1976).

5. Adlard, E. R., "A Review of the Methods for the Identification ofPersistent Hydrocarbon Pollutants on Seas and Beaches," Jour-nal of the Institute of Petroleum, Vol 58, No. 560 (MarcE1972), pp 63-74.

6. Ahmadjian, M., C. D. Baer, P. F. Lynch, and C. W. Brown, "InfraredSpectra of Petroleum Weathered Naturally and Under SimulatedConditions," Environmental Science and Technology, Vol 10,No. 8 (August 1976), pp 777-781.

7. Bentz, A. P., "Oil Spill Identification," Analytical Chemistry,Vol 48, No. 6 (May 1976), pp 455A-470A.

8. Brown, C. W., P. F. Lynch, and M. Ahnadjian, "Applications ofInfrared Spectroscopy in Petroleum Analysis and Oil SpillIdentification," Applied Spectroscopy Review, Vol 9, No. 2(1975), pp 223-248.

9. Brown, C. W., P. F. Lynch, and M. Ahmadjian, "Monitoring Narragan-sett Bay Oil Spills by Infrared Spectroscopy," EnvironmentalScience and Technology, Vol 8 (July 1974), pp 669-670.

10. Brown, C. W., "Identification and Differentiation of Heavy ResidualOil and Asphalt Pollutants in Surface Water by ComparativeRatios of Infrared Absorbances," Environmental Science andTechnology, Vol 3, No. 2 (February 1969), pp 150-153.

11. Kawahara, F. K., and D. G. Ballinger, "Characterization of OilSlicks on Surface Waters," Industrial Engineering Chemistry,Product Research and Development, Vol 9, No. 4 (1970),pp 553-558.

52

REFERENCES (Cont'd)

12. Kawahara, F. K., J. F. Santner, and E. C. Julian, "Characterizationof Heavy Residual Fuel Oils and Asphalts by Infrared Spectro-photometry Using Statistical Discriminant Function Analysis,"Analytical Chemistry, Vol 46, No. 2 (February 1974),pp 266-273.

13. Lynch, P. F., and C. W. Brown, "Identifying Source of Petroleum byInfrared Spectroscopy," Environmental Science and Technology,Vol 7, No. 13 (December 1973), pp 1123-1127.

14. Mattson, J. S., "'Fingerprinting' of Oil by Infrared Spectrometry,"Analytical Chemistry, Vol 43, No. 13 (November 1971),pp 1872-1873.

15. Spencer, M. J., Oil Identification Using Infrare Spectrometry,UM-RSMAS #7502, DOT-CG-81-75-1364 (Department of Transporta-tion, United States Coast Guard, June 1975).

16. Wilson, C. A., E. P. Ferrero, and H. J. Coleman, Crude Oil SpillsResearch: An Investigation and Evaluation of Analytical Tech-n , PB 242128, Bureau of Mines Report of Investigation802471975).

17. Chisholm, B. R., H. G. Eldering, L. P. Giering, and A. W. Hornig,Total Luminescence Contour Spectra of Six Topped Crude Oils,BERC RI-76/16 (1976).

18. Freegarde, M., C. G. Hatchard, and C. A. Parker, "Oil Spill at Sea:Its Identification, Determination, and Ultimate Fate," Labora-tory Practice, Vol 20, No. 1 (1971), pp 35-40.

19. John, P., and I. Soutar, "Identification of Crude Oils by Synchro-nous Excitation Spectrofluorimetry," Analytical Chemistry,Vol 48, No. 3 (March 1976), pp 520-524.

20. Parker, C. A., and W. J. Barnes, "Spectrophotofluorimetry of Lubri-cating Oils: Determination of Oil Mist in Air," Analyst,Vol 85 (1960), pp 3-8.

21. Thruston, A. D., and R. W. Knight, "Characterization of Crude andResidual-Type Oils by Fluorescence Spectroscopy," Environmen-tal Science and Technology, Vol 5, No. 1 (January 1971T,pp 64-69.

22. ASTM Method D3650-78, "New Standard Method of Test for Comparisonof Waterborne Petroleum Oils by Fluorescence Analysis, "Bookof ASTM Standards, Part 31 (American Society for Testing andMaterials [ASTMj).

53

. ANEW96 6 0 0im

REFERENCES (Cont'd)

23. Adlard, E. R., L. F. Creaser, and P. H. D. Matthews, "Identifica-tion of Hydrocarbon Pollutants on Seas and Be(aches by GasChromatography," Analytical Chemistry, Vol 44, No. 1 (January1972), pp 64-73.

24. Garza, M. E., and J. Muth, "Characterization of Crude, Semirefinedand Refined Oils by Gas-Liquid Chromatography," EnvironmentalScience and Technology, Vol 8, No. 3 (March 1974), pp 249-255.

25. Petrovic, K., and D. Vitorovic, "Recognition and Qualitative Char-acterization of Commercial Petroleum Fuels ana Synthetic Fuelsby Gas Chromatographic Fingerprinting Technique," Journal ofChromatography, Vol 119 (1976), pp 413-422.

26. Ramsdale, S. J., and R. E. Wilkinson, "Identification of PetroleumSources of Beach Pollution by Gas-Liquid Chromatography,"Journal of the Institute of Petroleum, Vol 54, No. 539(November 1968), pp 326-332.

27. Rasmussen, D. V., "Characterization of Oil Spills by CapillaryColumn Gas Chromatography," Analytical Chemistry, Vol 48,No. 11 (September 1976), pp 1562-1566.

28. Zafiriou, 0. C., "Improved Method for Characterizing EnvironmentalHydrocarbons by Gas Chromatography," Analytical Chemistry,Vol 45, No. 6 (May 1973), pp 952-956.

29. Adlard, E. R., L. F. Creaser, and P. H. D. Matthews, "Identifica-tion of Hydrocarbon Pollutants on Seas and Beaches by GasChromatography," Analytical Chemistry, Vol 44, No. 1 (January1972), pp 64-73.

30. Cole, R. D., "Recognition of Crude Oils by Capillary Gas Chromatog-raphy," Nature, Vol 233 (October 1971), pp 546-548.

31. "Analytical Methods for the Identification of the Source of Pollu-tion by Oil of the Seas, Rivers, and Beaches," Journal of theInstitute of Petroleum, Vol 56, No. 548 (March 1970),pp 107-117.

32. Kawahara, F. K., "Characterization and Identification of SpilledFuel Oils by Gas Chromatography and Infrared Spectropho-tometry," Journal of Chromatographic Science, Vol 10 (October1972), pp 629-636.

33. Kawahara, F. K., "Trace Organic Components as Fingerprints in GasChromatographic Identification of Spilled Asphalts," Environ-mental Science and Technology, Vol 10, No. 8 (August T976),pp 761-765.

54

-A

REFERENCES (Cont'd)

34. Kreider, R. E., "Identification of Oil Leaks and Spills," Proceed-ins of Joint Conference on Prevention and Control of OilSpills (American Petroleum Institute, June 15-17, 1971).

35. Petrovic, K., and D. Vitorovic, "Recognition and Qualitative Char-acterization of Commercial Petroleum Fuels and Synthetic Fuelsby Gas Chromatographic Fingerprinting Technique," Journal ofChromatography, Vol 119 (1976), pp 413-422.

36. Pym, J. G., J. E. Ray, G. W. Smith, and E. V. Whitehead, "PetroleumTriperpane Fingerprinting of Crude Oils," Analytical Chemis-try, Vol 47, No. 9 (August 1975), pp 1617-1622.

37. Ramsdale, S. J., and R. E. Wilkinson, "Identification of PetroleumSources of Beach Pollution by Gas-Liquid Chromatography,"Journal of the Institute of Petroleum, Vol 54, No. 539(November 1968), pp 326-332.

38. Rasmussen, D. V., "Characterization of Oil Spills by CapillaryColumn Gas Chromatography," Analytical Chemistry, Vol 48,No. 11 (September 1976), pp 1562-1566.

39. Zafiriou, 0. C., "Improved Method for Characterizing EnvironmentalHydrocarbons by Gas Chromatography," Analytical Chemistry,Vol 45, No. 6 (May 1973), pp 952-956.

40. ASTM Method D-3328-78, "Method of Test for Comparison of WaterbornePetroleum Oils by Gas Chromatography, Book of ASTM Standards,Part 31 (ASTM).

41. Saner, W. A., G. E. Fitzgerald, and J. P. Welsh, "Liquid Chromato-graphic Identification of Oils by Separation of the MethanolExtractable Fraction," Analytical Chemistry, Vol 48, No. 1747(October 1976).

42. Brunnock, J. V., D. F. Duckworth, and G. G. Stephens, "Analysis ofBeach Pollutants," Journal of the Institute of Petroleum,Vol 54, No. 539 (November 1968), pp 310-325.

43. Duewer, D. L., B. R. Kowalski, and T. F. Schatzki, "Source Identif-ication of Oil Spills by Pattern Recognition Analysis ofNatural Elemental Composition," Analy-tical Chemistry, Vol 47,No. 9 (August 1975), pp 1574-1583.

44. Adlard, E. R., "A Review of the Methods for the Identification ofPersistent Hydrocarbon Pollutants on Seas and Beaches," Jour-nal of the Institute of Petroleum, Vol 58, No. 560 (MarcW-1972), pp 63-74.

55

REFERENCES (Cont'd)

45. Gruenfeld, M., Proceedings of the Joint Conference on Preventionand Control of Oil Spills (American Petroleum Institute,1973), p 179.

46. Gruenfeld, M., "Extraction of Dispersed Oils From Water for Qudnti-tative Analysis by Infrared Spectrophotometry," EnvironmentalScience and Technology (July 1973).

47. Gruenfeld, M., "Quantitative Analysis of Petroleum Oil Pollutantsby Infrared Spectrophotometry," Water quality Parameters, ASTMSpecial Technical Publication (STP 573.290-308 (ASTM, 1975).

48. Brown, R. A., et al., Measurement and Interpretation of Non-Volatile Hydrocarbons in the Ocean Part I: The Measurementin the Atlantic Mediterranean, Gulf of Mexico) and PersianGulf, Contract C-I-35049, COM 74-11634 (Department of Com-merce, Maritime Administration (July 1974).

49. Oil and Grease, Total Recoverable (Infrared), EPA STORET No. 00560.

50. Mattson, J. W., et al., "Multivariate Statistical Approach to theFingerprinting of Oils by Infrared Spectrometry," AnalyticalChemistry, Vol 49, No. 2 (1977), pp 297-302.

51. Kawahara, F. K., and Y. Y. Yang, "Systems Chemical Analysis ofPetroleum Pollutants," Analytical Chemistry, Vol 48, No. 4(1976), pp 651-655.

52. Vogel, R. S., A Spectrochemical Method for the Determination ofIndi',idual Rare Earth Elements in Uranium Ore Concentrates andHigh-Purity Uranium Compounds, MCW 1498 (1966).

53. Kahn, L., B. F. Dudenbostel, D. N. Speis, and G. Karras, "Determi-nation of Petroleum-Derived Oils and Vegetable/Animal Oils inthe Presence of Each Other," American Laboratory (March 1977).

54. Gruenfeld, M., and R. Frederick, "The Ultrasonic Dispersion, SourceIdentification, and Quantitative Analysis of Petroleum Oils inWater," Rapports et Procesverbaux des Reunions, Vol 171 (Con-seil International Pour L'Exploratlon de la Mer, 1977),pp 33-38.

55. Brown, C. W., P. F. Lynch, and M. Ahmadjian, Identification of OilSlicks by Infrared Spectroscopy, Final Report, U.S. CoastGuard Contract DOT-CT-81-74-1099/ADA040975 (August 1976).

56. Frankenfeld, J. W., Weathering of Oil at Sea, Final Report, U.S.Coast Guard Contract DOT-CG-23035-A/AD787789 (September 1973).

56

A

(REFERENCES (Cont 'd)

57. Flanigan, G. A., R. D. Copperthite, and S. Buchanan, LaboratoryWeathering of Petroleum, 1978 Pittsburgh Conference,Paper 556.

58. Anderson, C. P., T. J. Killeen, J. B. Taft, and A. P. Bentz, Artif-icial Oil Weathering Techniques, paper submitted to 1979Pittsburgh Conference.

59. Zurcher, F., and M. Thuer, "Rapid Weathering Processes of Fuel Oilin Natural Waters: Analyses and Interpretations," Environmen-tal Science and Technology, Vol 12, No. 7 (july 1971)9

60. Mecks, D. C., J. J. Bommarito, R. S. Schwantje, and A. T. Edgerton,Transfer, Installation and Flight Testing of the Modified Air-borne Oil Surveillance System (AOSS II) in an HC-130B Air-craft, Final Report, U.S. Coast Guard CG-D-60-77/ADA052434T-gust 1977).

61. Grizzle, P. L., and H. J. Coleman, "GLC Analysis of N-Paraffin Dis-tribution in Crude Oils and Topped Crude Oils for Oil Identif-ication," BERC/RI-77/12 (1977).

62. ASTM Method D-3415-75T, "Practices for Identification of WaterborneOils," Annual Book of ASTM Standards, Part 31 (ASTM).

63. ASTM Method D 3326-78, "Preparation of Samples for Identificationof Waterborne Oils" (ASTM).

64. Bentz, A. P., Oil Spill Identification and Remote Sensing, U.S.Coast Guard Research and Development Center, presented atDivision of Petroleum Chemistry, American Chemical SocietyMeeting (September 1978).

65. Manual on Disposal of Refinery Wastes, Vol IV, Method 733-58 (Amer-ican Petroleum Institute, 1958).

57

CERL DISTRIBUTION

Chief of Engineers inst. for Water Res.. ATTN: Library HSCATTN: Tech Monitor HQ JSAHSC, ATTN: HSLO-FATTN: DAEN-RD Army Instl. and Major Activities (CONUS) ATTN: Facilities EngineerATTN: DAEN-MP DARCOM - Dir.. Inst., & Svcs. Fitzsimorn, Amy Medical CenterATTN: DAEN-ZC ATTN: Facilities Engineer Walter Reed Army Medical CenterATTN: DAEN-CW ARRADCOMATIN: DAEN-RM Aberdeen Proving Ground USACCATTN: DAEN-CCP Army Matis. and Mechanics Res Ctr. ATTN: Facilities EngineerATTN: DAEN-ASI-L (2) Corpus Christi Army Depot Fort Huachuca

Harry Diamond Laboratories Fort RitchieUS Army Engineer Districts Dugway Proving Ground

ANN: Library Jefferson Proving Ground 1TMCAlaska Fort Monmouth HQ, ATTN: MTMC-SAAlbatin Letterkenny Army epot ATTN: Facilities EngineerAlbuquerque Natick Research ana D-v. Ctr. Oakland Army BaseBaltimore New Cumberland Army Depot Bayonne MOTBuffalo Pueblo Army Depot Sunny Point MOTCharleston Red River Army DepotChicago Redstone Arsenal US Military AcademyDetroit Rock Island Arsenal ATIN: Facilities EngireerFar East Savannah Army DepotFort Worth Sharpe Army Depot USAES, Fort Relvoir, VACalveston Seneca Army Depot ATTN: FE Mgmt. Br.Huntington Tobyhanna Army Depot ATTN: Const. Mgmt. Br.Jacksonville Tooele Army Depot ATTN: Engr. LibraryJapan Watervliet ArsenalJidda Yuma Proving Ground Chief Inst. Div., I&SA, Rock Island. ILKansas City White Sands Missile RangeLittle Rock USA ARRCOM, ATTN: D'r., Instl & SvcLos Angeles FORSCOM TARCOM, Fac. Dio.Louisville FORSCOM Engineer, ATTN: AFEN-FE TECO I. ATTN: DRSTE-LG-FMemphis ATT: Facilities Engineers TSARCOM, ATTN: STSAS-FMobile Fort Buchanan NARAD COM, ATT : DRDNA-FNashville Fort Bragg AMMRC, ATTN: DROiR-WENew Orleans Fort CampbellNew York Fort Carson HQ, XVIII Airburne Corps andNorfolk Fort Devens Ft. BraggOmaha Fort Drum ATTN AFZA-FE-EEPhiladelphia Fort HoodPittsburgh Fort Indlantown Gap HQ. 7th Army Training CommandPortland Fort Irwin ATTM: AETTG-DEH (5)Riyadh Fort Sam HoustonRock Island Fort Lewis HQ USAREUR and 7th ArmySacramento Fort McCoy ODCS/EngineerSan Francisco Fort McPherson ATTN AEAEN-EH (4)Savannah Fort George G. MeadeSeattle Fort Ord V CorpsSt. Louis Fort Polk ATTN: AETVDEH (5)St. Paul Fort RichardsonTulsa Fort Riley VII CorpsVicksburg Presidn of San Francisco ATT: AETSDRH (5)Walla Walla Fort SneridanWilmington Fort Stewart 21st Support Command

Fort Wainwright ATTN: AEREh (5)US Army Engineer Divisions Vancouver Bks.ATTN: Library US Army BerlinEurope TRADOC ATTN: AEBA-EN (2)Huntsville HQ TRADOC, ATTN: ATEN-FELower Mississippi Valley ATTN: Facilities Engineer US Army Southern European Task ForceMiddle East Fort Belvoir ATTN: AESE-ENG (5)Middle East (Rear) Fort BenningMissouri River Fort Bliss US Army Installation Support Activity,New England Carlisle Barracks EuropeNorth Atlantic Fort Chaffee ATTN: AEUL-RPNorth Central Fort DixNorth Pacific Fort Eustis Bth USA, KoreaOhio River Fort Gordon AITN: EAFEPacific Ocean Fort Hamilton Cdr, Fac Engr Act (8)South Atlantic Fort Benjamin Harrison AFE, Yongsan AreaSouth Pacific Fort Jackson AFE. 20 Inf DivSouthwestern Fort Knox AFE. Area II Sot Det

Fort Leavenworth AFE, Cp HumphreysWaterways Experiment Station Fort Lee AFE, PusanATTN: Library Fort McClellan AFE, Taegu# Fort Monroe

Cold Regions Research Engineering Lab Fort Rucker DLA ATTN: DLA-WIATTN: Library Fort Sill

Fort Leonard Wood USA Japan (USARJ)US Government Printing Office Ch, FE Div, AJEN-FEReceiving Section/Depository Copies (2) INSCOM - Ch, Instl. Div. Fac Engr (Honshu)

ATTN: Facilities Engineer Fac Engr (Okinawa)Defense Technical Information Center Vint Hill Farms StationATTN: DDA (12) Arlington Hall Station ROK/US Combined Forces Command

AIIM: EUSA-HHC-CFC/EngrEngineering Societies Library

WESTCOM

New York, NY ATTN: Facilities EngineerFort Shafter

FESA, ATTN: LibraryMDW

ETL, ATTN: Library ATTN: Facilities EngineerCameron Station

Engr. Studies Center, ATTN: Library Fort Lesley J. McNairFort Myer

4------------ , ___ -_ ,.._,___ .______ ,_, __""

EE 13-. li..trlbution

Piratinny Arsenal US Army Engineer District US Army F gineer DivisionATTN: SMIJPA-VP3 Philadelphia Missouri River

AITN: Chief, NAPEN-E ATTN: Chief, MRDD-TDirector of Facilities Engineering Norfolk SouthwesternMiami, FL 34004 ATTN: Chief, NAOEN-O ATTN: Chief, SWDED-TH

Huntington South Pacific

DARCO4 STIT-EUR ATTN: Chief. ORHED-H ATTN. Chief, SPDED-TG

APO Ne. York 09710 Wilmington ATTN: Laboratory

ATTN: Chief, SAWEN-PM PdCific OceanWelt Point, NY 10996 ATTN: Chief, SAWIEN-E ATTN: rhief, Engr liv

ATTN. Dept of Mechanics Charleston ATN: Lnief, PODED-MPATN: Library ATTN: Chief, Engr Div ATTN: Chief, PODED-P

Savannah North PacificHQA (SCR0-4E) ATTN: Chief, SASAS-L ATTN: Cnief, Engr Div

JacksonvilleChief of Engineers ATTN: Env Res Br HQ FORSCIM

AMT DAEN-MPO-8 Mobile ATTN: ArFN-CDATTN DAN-MPO-U ATTN: Chief. SAEN-C Ft McPher-on, NA 30330Aorr 0AEN-MPR VicksburgATTN DAEN-rPl-A ATTN: Chief. Engr Div AF/ROXTATTN DAEN-RDL Louisville WASH Of ?0310ATTN DAEN-ZC ATTN: Chief, Engr Div

St Paul M. J. KerbyNational Defense Headquarters ATTN: Chief. ED-H HQ AICOM/DEMUSDirector General of Construction Chicago Peterson AFB, CO R0914Ottawa. Ontario KIAOK2 ATN: Chief, NCCCO-R

Canada ATN: Chief, NCCED-H John WallATTN: Chief, NCCPB-ER 2Ph4 ABG/DEEC

Division of Building Research St Louis Tinker AFR. O 73145National Research Council ATTN: Chief, ED-BMontreal Road ATTN: Chief, ED-D LT David C. HallOttawa, Ontario KlAOR6 Kansas City 2852 APG/DECanada ATTN: Chief, Engr Div McClellan AF, CA 95652

OmahaAirports and Construction Services Dir ATTN: Chief. Engr Div AFESC/PRTtechnical Information Reference Centre Little Rock Tyndall AFB, FL 32403

KAO[, Transport Canada Building ATTN: Chief, Engr DivPlice de Aille Tulsa AF/PREU

Ottawa, Ontario KIAON8 ATTN: Chief, Engr Div Bolling AF, DC ?0332Carada Fort Worth

ATTN: SWFED-D Little Rock AF8British Liaison Officer (5) ATIN: SWFED-MA ATTN: 314/DEE (Mr. Gillham)').S. Army Mobility Equipment GalvestonResearch and Development Center ATTN: Chief, SWGAS-L US Naval Oceanographic Office

ft Belvoir, VA 22060 ATTN: Chief. SWGCO-M Library

Los Angeles Bay St Louis, NS 39522Aberbeen Proving Ground, ND 21005 ATTN: Chief. SPLEB-EATTN: AMXHE/J. D. Weisz San Francisco Naval Training Cquipment Center

ATTN: Chief, Engr Div ATTN- Technical LibraryFt Belvoir. VA 22060 Sacramento Orlando, FL 32R13ATIN. Learning Resources Center ATTN: Chief, SPKED-DATTN: ATSE-TB-TL (2) Far East Naval facilities Engr CommandATTN: Kingman Bldg. Library ATTN: Chief, Engr Div ATTN: Code 04ATTN: MAJ Shurb (4) Seattle Alexandria, VA 22332

ATTN: Chief. NPSEN-FMFt Leavenworth, KS 66027 ATTN: Chief, EN-DB-SE LT Neil B. Hall, CEC, USN (Code 101)ATZLCA-SA/F. WOlcott ATTN: Chief. NPSEN-PL-WC A84-6366

ATN: Chief, NPSEN-PL-ER OS Navy Public Works CenterFt Monroe, VA 23651 Walla Walla Box 6, FPO San Francisco 96651ATTN: ATEN-AD (3) ATTN: Chief, Engr DivATTN: ATEN-FE-E Alaska Port Hueneme, CA 93043ATTN: TN-FE-U ATTN: NPASA-R ATTN: Library (Code LOBA)

ATTN: Morrell LibraryFt Lee, VA 23801 US Army Engineer DivisionATTN: DRXMC-D (2) New England Washington, DC

ATTN: Chief, NEDED-E ATTN: Building Research Advisory BoardUS Army Foreign Science B Tech Center North Atlantic ATTN: Transportation Research BoardATTN: Charlottesville, VA 22901 ATTN: Chief, NADEN-T ATTN: Library of Congress (2)ATTN: far East Office Middle East (Rear) ATTN: Dept of Transportation Library

ATTN: MEBEO-TSth US Anmy South Atlantic W. N. Lofross, P.E.ATTN: AFKM-LG-E ATTN: Chief, SADEN-TE Dept of Transportation

Huntsville Tallahassee, FL 323045th 0S Army ATTN: Chief, HNOED-CSATTN: AFKC-FN ATTN: Chief, HNDED-ME James T. Burns

ATTN: Chief, MNDEB-SR Base Civil Engineering SqdnOS Army Engineer District Lower Mississippi Valley Patrick AFB, FL 32925New York ATTN: Chief, PD-R

ATTN: Chief, NANEN-E Ohio RiverATTN: Chief. Design Br ATTN: Chief, Engr Div

Pittsburgh North CentralATTN; Chief, Engr Div ATTN: Chief, Engr Planning Br

' .. . ..lip

Vogel, Raymond SIdentification and quantification of hydrocarbon products in effluents/

by R. S. Vogel ... t al. -- Champaign, IL : Construction Engineering ResearchLaboratory ; Springfield, VA : available from NTIS, 1980.

57 p. ; 27 cm (Interim report ; N-91)

1. Sewage -- analysis. 2. Hydrocarbons -- analysis. 3. Organic waterpollutants -- analysis. 1. Fileccia, Robert J. II. Nikucki, Walter J.111. Lampo. Richard G. IV. Title. V. Series: U.S. Army ConstructionEngineering Research Laboratory. Interim report M-91.