LEABHARLANN CHOLAISTE NA TRIONOIDE, BAILE ATHA CLIATH TRINITY COLLEGE LIBRARY DUBLIN
OUscoil Atha Cliath The University of Dublin

Terms and Conditions of Use of Digitised Theses from Trinity College Library Dublin 

Copyright statement

All material supplied by Trinity College Library is protected by copyright (under the Copyright and 
Related Rights Act, 2000 as amended) and other relevant Intellectual Property Rights. By accessing 
and using a Digitised Thesis from Trinity College Library you acknowledge that all Intellectual Property 
Rights in any Works supplied are the sole and exclusive property of the copyright and/or other I PR 
holder. Specific copyright holders may not be explicitly identified. Use of materials from other sources 
within a thesis should not be construed as a claim over them.

A non-exclusive, non-transferable licence is hereby granted to those using or reproducing, in whole or in 
part, the material for valid purposes, providing the copyright owners are acknowledged using the normal 
conventions. Where specific permission to use material is required, this is identified and such 
permission must be sought from the copyright holder or agency cited.

Liability statement

By using a Digitised Thesis, I accept that Trinity College Dublin bears no legal responsibility for the 
accuracy, legality or comprehensiveness of materials contained within the thesis, and that Trinity 
College Dublin accepts no liability for indirect, consequential, or incidental, damages or losses arising 
from use of the thesis for whatever reason. Information located in a thesis may be subject to specific 
use constraints, details of which may not be explicitly described. It is the responsibility of potential and 
actual users to be aware of such constraints and to abide by them. By making use of material from a 
digitised thesis, you accept these copyright and disclaimer provisions. Where it is brought to the 
attention of Trinity College Library that there may be a breach of copyright or other restraint, it is the 
policy to withdraw or take down access to a thesis while the issue is being resolved.

Access Agreement

By using a Digitised Thesis from Trinity College Library you are bound by the following Terms & 
Conditions. Please read them carefully.

I have read and I understand the following statement: All material supplied via a Digitised Thesis from 
Trinity College Library is protected by copyright and other intellectual property rights, and duplication or 
sale of all or part of any of a thesis is not permitted, except that material may be duplicated by you for 
your research use or for educational purposes in electronic or print form providing the copyright owners 
are acknowledged using the normal conventions. You must obtain permission for any other use. 
Electronic or print copies may not be offered, whether for sale or otherwise to anyone. This copy has 
been supplied on the understanding that it is copyright material and that no quotation from the thesis 
may be published without proper acknowledgement.



BIOREMEDIATION OF UNRESOLVED COMPLEX 

MIXTURES IN MARINE OIL SPILLS:

AN ANALYTICAL PERSPECTIVE

By

Evin Patrick McGovern

A thesis submitted to the University of Dubhn (Trinity College) 

for the degree of Doctor of Philosophy

December 1999



i
\

I agree that the hbrary may lend 

or copy this thesis on request.

V ,

Evin McGovern Trinity College Dublin 1999

II



Acknowledgements
This project was part funded by EU DG XI Environment, Nuclear Safety and Civil 

Protection as project B4-3040/95/928/jnb/C4. The project was carried out at the Marine 

Institute Fisheries Research Centre (FRC) laboratories in Abbotstown.

I wish to express my gratitude to all those at FRC who contributed to the project during its 

course, and especially to Teresa O ’Shea for her excellent technical assistance over the course 

o f the project, to Brendan M cHugh for his technical assistance during phase I o f  the project, 

and to Conor Duffy for his assistance.

As with many multidisciplinary projects, the success o f  this project was only possible due to 

advice and collaboration with other experts. Dr. Barry K iely’s team at Bioresearch Ireland, 

University College Cork carried out the microbiological testing and inocula preparation at 

their laboratories under contract from the Marine Institute. I am very grateful to Barry for his 

expert advice on the microbial aspects o f the project and enjoyed renewing our collaboration 

in this field after many years.

I wish to thank my supervisors in Trinity College, Dr. Jim Wilson and P ro f Brian McMurry 

for their assistance and support throughout the project.

Many thanks to Eugene Nixon for his support for the proposal and his help in carrying out 

the project, particularly with regard to overcoming the administrative difficulties 

encountered. I am grateful to the Marine Institute for their sponsorship o f this project and 

support for my registration for a doctorate.

Other people to whom thanks are due include those who supplied oil: Statoil Norway (Thure 

Ingebriksted), Shell Ireland (PJ Cummins), Chevron UK (Seamus O ’Connor), Texaco UK. 

Thanks to those who gave advice and assistance directly or indirectly, especially at the outset 

o f  the project -  Dr. Per Wrang, Dr. Jose Biscaya, Dr. Steve Grigson, Dr. John O ’Brien, Dr. 

Alan Dobson and anybody I may have overlooked.

Finally, I am indebted to my wife, Ailve, for her patience over the entire project, but most 

particularly during the lengthy and antisocial period o f  thesis preparation.



For my wife Ailve, 

and in memory of my parents.

iv



BIOREMEDIATION OF UNRESOLVED COMPLEX 

MIXTURES IN MARINE OIL SPILLS:

AN ANALYTICAL PERSPECTIVE

Table of Contents

Declaration i

Permission to Copy ii

Acknowledgem ents iii

Dedication iv

Contents v

List o f Tables xiii

List o f Figures xvi

Sum mary xxi

Abbreviations and Terms xxiii

Chapter 1. Background

1.1 Background I

1.2 Petroleum hydrocarbons and unresolved complex mixtures (UCM) 2

1.3 Petroleum hydrocarbons in the marine environment

1.3.1 Inputs and sources 6

1.3.2 Physical and chemical fate o f  oil spills 6

1.3.3 Toxicity o f  petroleum hydrocarbons and impacts o f  marine oil spills 8

1.3.4 Clean-up o f marine oil spills 9

1.4 Bioremediation o f oil spills

1.4.1 Application and case studies 1 1



1.4.2 Microbial biodegradation of petroleum hydrocarbons 12

1.4.3 Metabolic pathways 14

1.4.4 Factors affecting and optimisation of hydrocarbon biodegradation 17

1.4.4.1 Nutrients 17

1.4.4.2 Oxygen 18

1.4.4.3 Seeding 18

1.4.4.4 Physical state o f  the oil and bioavailability 20

1.4.4.5 Other factors -  pH, temperature, competitive substrates 20

1.5 Analysis of petroleum hydrocarbons 22

1.5.1 Fractionation techniques 22

1.5.1.1 Column and thin layer chromatography 22

1.5.1.2 Molecular sieve and urea and thiourea adduction 22

1.5.2 Spectroscopic techniques

1.5.2.1 Infrared spectroscopy 23

1.5.2.2 Fluorescence/luminescence spectroscopy 24

1.5.2.3NMR 24

1.5.3 Chromatography

1.5.3.1 Gas chromatography 25

1.5.3.2 HPLC  25

1.5.4 Mass spectrometry (MS)

1.5.4.1 Gas-chromatography-mass spectrometry (GC-MS) 26

1.5.4.2 Direct Sample introduction mass spectrometry 28

1.5.4.3 Chemical ionisation mass spectrometry (CI-MS) 29

1.5.4.4 Chromic acid oxidation 29

1.5.4.5 Stereochemistry and nomenclature o f  hydrocarbons 29

1.6 Objectives o f the project 32

1.7 Outline of the work programme 33

VI



Chapter 2. Analytical Development and Characterisation of 

Oils (PHASE I A)

2.1 Purpose 35

2.2 Selection & acquisition of oils for phase I of project 35

2.2.1 Forties crude oil 35

2.2.2 Gullfaks crude oil 36

2.2.3 Alba crude oil 37

2.2.4 Lubricating oils 37

2.3 Analytical Methods 38

2.3.1 Gravimetric 38

2.3.2 Urea and thiourea adduction 39

2.3.3 Gas Chromatography (GC) 41

2.3.4 Spectroscopic 44

Fluorescence 44

Infrared 50

2.3.5 High performance liquid chromatography (HPLC) 51

2.3.6 Mass spectrometry 5 3

2.3.7 Other methods evaluated 59

2.4 Conclusions 60

Chapter 3. Inocula Development and Evaluation 

(PHASE IB)

3.1 Background 61

3.2 Enrichment o f microbial consortia from marine sediments 62

3.3 Evaluation o f enriched cultures 64

3.3.1 Determination of suitable propagation medium for crude inoculum 64

3.3.2 Comparison of enrichments on UCM as sole carbon source 65

v i i



3.3.3 Efficiency o f isolated strains vs. crude enrichment broths 71

3.4 Preparation o f  inocula for microcosm trials -  phase II 78

3.5 Conclusions 78

Chapter 4. Design of Bioremediation Microcosm Trials 

(PHASE II)

4.1 Background 80

4.2 Experimental design o f bioremediation trials 80

4.2.1 Design 80

4.2.2 Selection o f oils for trials 81

4.2.3 Nomenclature 82

4.3 Microcosm unit design 83

4.3.1 Spiked sediment 83

4.3.2 Microcosm unit design 83

4.3.3 Seawater/nutrients 85

4.3.4 Inoculation o f T1 and T2 tests 85

4.3.5 Location 85

4.3.6 Temperature 8 5

4.4 Monitoring o f microcosms and oil biodegradation

4.4.1 Assessment and testing o f the units 86

4.4.2 Respirometry trials to assess UCM biodegradation potential o f  87

microcosm microbial populations

4.4.3 Sampling for biodegradation monitoring 87

4.4.4 Chemical monitoring o f  oil biodegradation 87

4.5 Quality assurance 87

v i i i



Chapter 5. Bioremediation Trial Results - Microcosm Design

5.1 Contamination o f sediment 90

5.2 Seawater cycle 90

5.3 Temperature 91

5.4 Nutrients 92

5.5 Sediment characteristics 95

5.6 Water parameters 97

5.7 Oxygen 99

5.8 M icrobiology 100

5.9 M icrocosm and field situation: a comparison 102

Chapter 6. Bioremediation Trial Results II -  Chemical 

Assessment of Biodegradation

6.1 Visual 104

6.2 Gravimetric assessment

6.2.1 Extractable m atter 104

6.2.2 Fractionation 108

6.3 Chromatography

6.3.1 Gas chromatography 110

6.3.2 High performance liquid chromatography 121

IX



6.4 Spectroscopy

6.4.1 Fluorescence spectroscopy

6.4.2 Infra red and nuclear magnetic resonance spectroscopy

123

124

6.5 Mass spectrometry

6.5.1 Gas chromatography-mass spectrometry -  Total ion chromatogram 124 

(TIC) analysis

6.5.2 Gas chromatography-mass spectrometry -  (Individual components -  134 

selected ion analysis)

Chapter 7. Oil Spill Identification and Fingerprinting 

Techniques

7.1 Background 146

7.2 Pristane:phytane ratio 148

7.3 Total fluorescence spectroscopy 151

7.4 PAH patterns and ratios 154

7.5 Biomarker patterns 160

7.6 Biomarker indices 163

Chapter 8. General Discussion and Conclusions

8.1 Microcosm design 165

8.2 Bioaugmentation 166

8.3 Biostimulation 168

X



8.4 Changes to oil composition during bioremediation trials 171

8.5 Analytical aspects and monitoring of bioremediation 172

8.6 Oil spill fingerprinting 176

8.7 Overall findings and recommendations 178

8.8 Further research 182

Chapter 9. References

Appendices
A. Analytical methodology

A.I. Extraction o f hydrocarbon from contaminated sediment (microcosm 203

I samples) and gravimetric analysis 203

I A.2. Analytical regime for microcosm experiments

A.3. Oil dissolution method & gravimetric analysis (phase I) 204

A.4. Fractionation by column chromatography 205

I A.5. Urea adduction 206
(■

i A.6. Thiourea adduction 207

A.7. Gas chromatography (flame ionisation detection) [GC-FID] 208

A.8. HPLC 209

A.9 Fluorescence spectroscopy 210

I A. 10. Infrared spectroscopy 212

I A. 11. Mass spectrometry 213

} 215

B. Selected mass chromatograms for Alba oil

C. Minimal media to support growth on UCM 219

D. Analytical data for respirometry studies and microcosm trials 225

D. 1 GC-MS data from analysis of oils from respirometry studies 227

D.2 Gravimetric analysis results Forties and Alba microcosms samples 228

D.3 Gas chromatography (FID) results for microcosm trials -  peak/UCM 232

areas 234

xi



D.4 GC-MS TIC total ion abundance data for analysis o f  microcosm 

samples

D.5 GC-MS SIR data for analysis o f  microcosm



List of Tables

Table 1.1. The chemistry o f crude oil

Table 2.1: Gravimetic recovery o f  oil after hexane extraction/dissolution

Table 2.2: Fractionation o f  oils expressed as a percentage o f  hexane extracted oil

Table 2.3: Percentage urea adducted and thiourea adducted material from oils

Table 2.4: GC parameters. UCM description in terms o f  «-alkane retention

Table 2.5: Ratios o f  fluorescence intensity at Ex?^=270nm, EmA,=340nm and 

ExA,=230nm, Em>,= 365nm

Table 2.6 : IR response o f oils relative to Alba crude oil

Table 2.7: Aromatic component distribution for oil samples relative to Alba 

reference oil

Table 2.8: Total ion abundance ratios for repeated alba oil analysis by GC-MS

Table 2.9: Composition o f the lubricating oils calculated from mass spectra (based 

on ASTM method D 2786-9I)

Table 3.1: Setup o f respirometry trials for evaluating efficiency o f isolated strains 

vs. crude enrichment broths

Table 4.1: Bioremediation trials experimental design

Table 4.2: Hydrocarbon monitoring programme for bioremediation microcosm 

trials

Table 5.1: Microcosm trial air temperatures -  descriptive statistics



Table 5.2: Particle size distribution o f sediment 95

Table 5.3: M oisture content o f  drained microcosm sediments at t=350 as % volume 

o f pore space

Table 5.4: Seawater parameters in microcosms

Table 6.1: Gravimetric results for t=0 reference sediments

Table 6.2: Gravimetric results -  percentage reduction o f  oil concentrations in 

microcosms at t=473 days

Table 6.3: Estimated rate constants (k), 95% confidence intervals, and coefficients 

o f  determination (R^) for non-linear solutions o f  gravimetric biodegradation using 

the model [A]t =  [A]oe’'“

Table 6.4: Biodegradation rate constants o f Forties oil components with 95% 

confidence intervals and coefficients o f determination (R^) determined from GC 

results using the model [A], = [A]oe '“

Table 6.5: Biodegradation rate constants o f  Alba oil components with 95% 

confidence intervals and coefficients o f determination (R^) determined from GC 

results using the model [A]t = [A]oe''“

Table 6.6: Composition o f saturate fraction o f Alba oil in microcosm sediments at 

t=473 days as calculated by mass spectrometry fragment peak analysis (ASTM 

2786-91).

Table 6.7: Percentage reduction in sediment concentrations o f C30-hopane (P6), 

20R-ethylcholestane (triaromatic sterane TAS4) and P6/TAS4 ratio at t=473 days 

wrt. t=0 sediments.

Table 7.1: Pristane:phytane ratios and summary statistics for Alba microcosms as 

measured by GC-FID (peak area) during microcosm trials.

97

98

104

105

107

1 1 2

117

133

137

151



Table 7.2: Mean values and standard deviations for biomarker indices for nutrient 164

supplemented and poison control samples from Alba microcosms and for t=0

samples.

Table A. 1: Ions (m/z) recorded in GC-MS SIR mode 218

Table B .l: Biomarker peak assignations 224

XV



List of Figures

Fig. 1.1; Molecular structures of example crude oil compounds. 5

Fig. 1.2: Simple schematic of the primary weathering effects on spilt oil 7

Fig. 1.3: Pathway for pristane oxidation by Brevibacterium etythrogenes 15

(adapted from Singer & Finnerty 1984 in Petroleum Microbiology ed. Atlas)

Fig. 1.4: Pathway for the bacterial oxidation of naphthalene 16

(adapted from Cemiglia 1984 in Petroleum Microbiology ed. Atlas)

Fig. 1.5: Stereochemistry and nomenclature of petroleum hydrocarbons. 31

Fig. 1.6: Overview of the project 34

Fig. 2.1: Gas chromatogram of weathered Forties crude oil 36

Fig. 2.2: Gas chromatogram of Gullfaks reservoir biodegraded crude oil 36

Fig. 2.3: Gas chromatogram of Alba reservoir biodegraded crude oil 37

Fig. 2.4: Gas chromatograms of lubricating oils 37

Fig. 2.5: GC traces of urea adduct and non-adduct fractions 41

Fig. 2.6: Gas chromatograms for dried oils 42

Fig. 2.7: UCM area ‘segments’ between «-alkanes as percentage o f the total area. 44

Measured by GC-FID for Forties crude oil

Fig. 2.8: Total fluorescence spectra for 6 oils 48

Fig. 2.9: Total fluorescence spectra of 6 PAH compounds. 49



Fig. 2.10: Mass spectra for crude oils and lubricating oils (Combined spectra over 

entire UCM and combined baseline subtracted)

Fig. 2.11: Ion abundance ratios from E1+ mass spectra o f  dried oils

Fig. 3.1: Weathered Forties crude oil GC-MS (TIC) trace (t=0) and same oil after 

treatment with «-alkane degraders for 24 days

Fig. 3.2: Respirometry trial - sample and control

Ion abundance ratios from mass spectra o f  UCM after 2 weeks respirometry for 

sample subjected to enriched consortia and for an abiotic control

Fig. 3.3: Evaluation o f rapid propagation media on subsequent capacity o f 

enriched consortia to biodegrade UCM

Fig. 3.4: Respirometry on UCM by enriched and non-enriched microbial 

consortia

Fig. 3.5: Respirometry trials for enrichment flasks - Ion abundance ratios from 

UCM mass spectra after 6 weeks

Fig. 3.6: Reduction in UCM component groups during respirometry tests for 

consortia efficacy on UCM. GC-M S component group:C30-hopane ratios 

expressed as a percentage o f component group:C30-hopane ratios for abiotic 

control sample

Fig. 3.7: Respirometry -  comparison o f pooled selection o f micro-organisms and 

crude enriched and unenriched consortia

Fig. 3.8: GC chromatograms o f extracts from respirometry trial.

Fig. 3.9: GC-MS ion abundance ratios for total oils from respirometry trials

Fig. 3.10: Biodegradation o f  aromatic components in respirometry test flasks. 

Degradation expressed normalised to the refractory internal m arker C30-hopane.

55

56

63

65

67

68

70

70

73

74

75

76

x v i i



Fig. 3.11: Biomarker fragmentograms for a) steranes b) triaromatic steranes c) 77

pentacyclic triterpanes -  for biodegraded UCM and abiotic control samples

Fig. 4.1: GC chromatograms o f the initial oil extract from sediment (t=0 82

samples).

Fig.4.2; Bioremediation trials: Microcosm unit design 84

Fig. 4.3: Photograph o f the experimental setup, showing microcosm units 86

Fig. 5.1: Approximate microcosm seawater cycle periods 91

Fig. 5.2: Air temperature fluctuations over the course o f the experiment 92

Fig. 5.3: Particle size distribution o f the microcosm sediment 95

Fig. 5.4: Respirometry trials on microcosm sediment washings (sampled at t=280 101

days)

Fig. 6.1: Gravimetric reduction o f hydrocarbon in microcosms (normalised to 108

Alba reference samples)

Fig. 6.2: Gravimetric results expressed as % aliphatic, aromatic and polar 109

components for Alba

Fig. 6.3: Forties C18:phytane ratios vs. time for microcosms 111

Fig. 6.4: UCM biodegradation for Forties UCM ‘segm ents’ and isoprenoids at 114

t=280 days and 473 days

Fig. 6.5: A comparison o f the biodegradation o f total n-alkanes, total UCM and 115

total hydrocarbon for FCl and FT3

Fig. 6.6: UCM biodegradation for Alba UCM segments and isoprenoids at t=473 118

days

x v i i i



Fig. 6.7: The biodegradation o f  different oil com ponents (Forties o il) for T3 

treatment

Fig. 6.8: P lots representing biodegradation o f  o il com ponents in Forties 

m icrocosm s

1 F ig. 6.9: P lots representing rem oval o f  oil com ponents in A lba m icrocosm s

I  Fig. 6 .10: A rom atic biodegradation for a) A lba and b) Forties m icrocosm s at
i

j t= 280  days as determ ined using HPLC
1

1
I F ig. 6.11: F luorescence spectroscopy: ratios o f  em ission  intensities for

ex .^ = 230n m , em.X, =  340nm  and ex .^ = 270n m , em .A,=360nm  for m icrocosm s at 

I t= 407  days

Fig. 6 .12: Ion abundance ratio trends for A lba m icrocosm s 127-

Fig. 6 .13: Ion abundance ratio trends for Forties m icrocosm s 129-

Fig. 6 .14: Saturate and m onoarom atic com p osition  in A lba m icrocosm  sedim en ts  

at t=473 days, expressed  as percentage o f  total petroleum  extracts.

I

Fig. 6 .15: C 30-hopane (P6) rem oval from m icrocosm s

Fig. 6 .16  a-d): R em oval o f  biomarker param eters from m icrocosm  sedim ent 

relative to P6 (C 30-hopane) and norm alised to reference oil

F ig. 6.17: R em oval o f  triaromatic sterane (T A S 4 -20R  eth ylch olestan e) from  

m icrocosm s, norm alised to A lba reference oil.

i

Fig. 6 .18  a-f): Biodegradation o f  C 1-, C 2-, and C 3- phenanthrenes and 142-

d ibenzothiophenes norm alised to T A S 4 and reference oil for m ean o f  duplicate  

A lba m icrocosm s

Fig. 6.19: B iodegradation o f  individual param eters for treated m icrocosm s at 

t=473 days com pared w ith  t=0 sam ples.

118

119

120

122

123

128

130

132

139

140

141

143

144

XI X



Fig. 6.20: Biodegradation o f substituted PAH in Forties oil for treated 

microcosms at t=473 days.

145

Fig. 7.1: Pristane:phytane ratios for Alba microcosms during the course o f the 150

monitoring trials

Fig. 7.2: Total fluorescence spectra o f Alba oil samples from microcosms at t=407 153

days

Fig. 7.3: Alteration o f GC-MS substituted-PAH patterns based on selected peak 158

areas for biodegraded oil, (mean AT1-T3), and non-biodegraded oil, (AC2 and 

At=0), at t=203 days

Fig. 7.4: Effect o f severe biodegradation on C2D/C2P and C3D/C3P PAH source 159

indicator ratios

Fig. 7.5: Alteration o f biomarker GC-MS patterns based on selected peak areas 161

for biodegraded oil (mean AT1-T3), and non-biodegraded oil (AC2 and At=0) at 

t=203 days and t=473 days

Fig. 7.6: Biomarker patterns at t=473 days for A iiT2 and AiiC2 162

X X



Summary

Unresolved complex mixture (UCM) is the fraction o f some petroleums that cannot be 

resolved by conventional gas chromatography and is characteristic o f  biodegraded petroleum 

oils and some refined products such as lubricating oils. This study aimed to investigate the 

effects o f environmental biodegradation, and more specifically shoreline bioremediation, on 

the chemistry o f UCM. A number o f  wet chemical and instrumental analytical 

methodologies, including gravimetry, urea adduction, spectroscopy, chromatography and 

mass spectrometry, were evaluated for their suitability to monitor UCM compositional 

changes during the course o f biodegradation trials. Six UCM rich oils were also 

characterised using these methods, three crude oils and three lubricating oils. In parallel, an 

investigation o f microbial degradation o f UCM in solution was carried out in the laboratory. 

Using respirometry trials, undefined consortia o f  natural marine micro-organisms enriched 

on UCM demonstrated an enhanced ability to metabolise UCM. The pooled crude enriched 

consortium also performed considerably better at metabolising UCM when compared with a 

cocktail derived from the apparently dominant individual colonies (not identified) isolated 

from the crude consortium using standard techniques. Initial investigations o f  UCM 

biodegradation using gas chromatography (GC) and gas chromatography-mass spectrometry 

(GC-MS) were carried out at this stage. An inoculum based on the undefined consortium was 

prepared for testing in microcosm trials.

Microcosms were designed that provided a controlled environment with a basic simulation 

o f shoreline conditions, including a simple tidal cycle, to assess the effect o f bioremediation 

treatments for the decontamination o f UCM contaminated sediments. Treatments included 

nutrient supplementation (T3) only, inoculation with the enriched consortium (T l), and 

augmentation with rapidly grown micro-organisms isolated from the sediment used as a base 

for contamination (T2). Both T l and T2 were also nutrient supplemented. A seawater control 

(C l) and poison control (C2) were also employed. Two sets o f  microcosms using different 

contaminating oils were set up. The two oils used were environmentally weathered Forties 

crude oil (-90%  UCM) and Alba crude oil, a reservoir biodegraded crude. A comprehensive 

monitoring programme was undertaken. The microcosms proved a very effective tool for 

assessing biodegradation. Results indicated that, despite the enhanced capacity o f the T l 

inoculum for UCM biodegradation exhibited in respirometry trials, bioaugmentation had no 

appreciable effect on biodegradation o f UCM or any UCM component monitored, compared 

with the nutrient only treatments. Nutrient treatment, biostimulation, had a discernible effect,



the extent o f  which differed for various parameters monitored. A simple model [A], = [A]o.e‘ 

was used to compare biodegradation rates over the course o f  the microcosm trials (473 

days) for GC and gravimetric results. A modest enhancement for total oil and UCM 

biodegradation was observed for nutrient supplemented microcosms, but a marked 

acceleration o f  PAH biodegradation was evident in these microcosms. Compositional 

changes were also more advanced in nutrient treated microcosms than seawater control units. 

Results suggested that biodegradation proceeded along intuitive lines, with lower molecular 

weight UCM material being more degradable. PAH biodegradation rates increased with 

■ decreasing ring number and alkyl substitution. In the aliphatic fraction, acyclic components

were more biodegradable than cyclic and recalcitrance increased with increasing condensed 

i  ring number. The saturate fraction was more biodegradable than the aromatic fraction with
r
I the resin fraction most recalcitrant. Washout was a notable feature for Alba microcosms.

This study enabled the comparison o f a range analytical tools and techniques for monitoring 

; UCM biodegradation, such as the n-octadecane:phytane ratio which was observed to be only

useful for monitoring the early stages o f biodegradation. A methodology developed for oil 

spill fingerprinting purposes was adapted for long term GC-MS monitoring o f 

biodegradation and proved very effective, overcoming some o f  the inherent difficulties in 

using GC-MS for this purpose. The biomarker compound 17 a , 21 P-C30-hopane has been 

used as a conserved internal marker in recent years for assessing biodegradation o f 

hydrocarbons. For Alba nutrient supplemented microcosms, this was seen to be substantially 

biodegraded and would thus lead to an underestimation o f biodegradation and 

bioremediation. C28 (20R) triaromatic sterane was chosen as a preferred alternative as it was 

abundant, resolvable by GC-MS and considerably more resistant to biodegradation than 

C30-hopane. The use o f relative ion abundance ratios was also developed as a tool for 

gaining an insight into bulk compositional change in UCM and this proved very useful. 

Finally, an investigation into the effect o f severe biodegradation on a range o f oil spill source 

identification parameters and patterns was carried out, and the results are presented in the 

thesis.

[

X X l l



Abbreviations and Terms

A M A P A rctic M onitoring and A ssessm en t Program m e
amu A tom ic m ass unit
A ST M A m erican  S ocie ty  for T esting and M aterials
A TR A ttenuated total reflectance
C Carbon
C 30-hopane 1 7 a (H ), 2 1 a (H )h o p a n e (C 30)
CFC C hlorofluorcarbon
C nD  (n = l,2 ,3 ) Substituted d ibenzoth iophenes (m eth yl-, d im eth yl- and trim ethyl- for 

n= 1,2 3 resp ective ly )
CnP (n = l,2 ,3 ) Substituted phenanthrenes (m ethyl-, d im eth yl- and trim ethyl- for n=  

1,2 3 resp ectively)
C x-b en zene A lk y lb en zen e. (alkyl substitution o f  x carbons)
C x-cycloh exan e A lk y lcyc loh exan e . (alkyl substitution o f  x carbons)
C l C onfid en ce interval
C l (M S ) C hem ical ion isation
DCM D ichlorom ethane

D D T D ich lorodiphenyltrichloroethane { l,l ,l - tr ic h lo r o -2 ,2 -b is  (4- 
ch lorophenyl) ethane}

DO D isso lved  oxygen
El Electron ion isation
Ex. A, E xcitation  w avelength
Em. X E m ission  w avelength
EU European U nion
FIA Fluorescent indicator adsorption
FID Flam e ion isation  detection
FT Fourier transform
GC G as chrom atography
G C -M S M ass spectrom etry
G EM G enetica lly  engineered  m icro-organism
G E SA M P Joint Group o f  Experts on the S cien tific  A sp ects o f  M arine Pollution
H H ydrogen
H 2O 2 H ydrogen peroxide
HCl H ydrochloric acid
HPLC H igh perform ance liquid chrom atography
IM O International M aritim e O rganisation
IR Infra-red
MI FRC M arine Institute F isheries R esearch Centre
M S M ass spectrom etry
M T B E M ethyl-/er?-butyl ether
M W M olecular w eigh t
m /z M assxh arge  ratio
N N itrogen
N 2 M olecular nitrogen
N E T A C N ational Environm ental T ech n ology  A pp lications Corporation
N F B C N ational Food  B io tech n o logy  Centre
N M R N uclear m agnetic resonance
nm N anom etre
N S O N itrogen , sulfur, oxygen
N R C N ational R esearch C ouncil (U S A )
O 2 M olecular oxygen

x x i i i



p Phosphorus
PAC Amino-cyano (HPLC column)
PAH Polynuclear aromatic hydrocarbon
PCA Plate count agar
PCB Polychlorinated biphenyl
PCE Perchloroethylene
Pr Pristane
Phy Phytane
PTFE Polytetrafluorethylene
PSD Particle size distribution
Psu Practical salinity unit
Rf Retention factor
RI Refractive Index
RIC Reconstructed ion chromatograph
RSD Relative standard deviation
SARA Saturates, aromatics, resins, asphaltenes
sd Standard deviation
SEC Size exclusion chromatography
SEEC Sea Empress Evaluation Committee
SIR Selected ion recording
TAS4 (20R) Ethylcholestane
TDB Tridecylbenzene
TIC Total ion chromatogram
TLC Thin layer chromatography
TPH Total petroleum hydrocarbon
Tm 17a(H)-22,29,30-trisnorhopane (C27)
Ts 18a(H)-22,29,30-trisnorhopane (C27)
UCC University College Cork
UCM Unresolved complex mixture
USEPA United States Environmental Protection Agency
UV Ultrviolet
UVF Ultraviolet/visible fluorescence
v/v Volume to volume
WWDW Wet weight/ dry weight
w/w Weight to weight

XXIV



1. BACKGROUND

1.1 BACKGROUND

Shoreline bioremediation o f oil spills involves intervention by man to accelerate biodegradation 

o f  the spilt oil, through optimisation o f conditions that may otherwise be rate limiting. Often this 

simply involves the addition o f nitrogen and phosphorus, as natural background environmental 

concentrations o f  these elements may limit biodegradation rates. However, in some cases it may 

entail more complex manipulations. The potential o f  bioremediation in enhancing biodegradation 

o f oil spills has been demonstrated in the laboratory and the field.

Crude oil and most petroleum products are very complex mixtures o f organic chemicals. In 

general, preferential biodegradation o f simpler substrates in petroleum, in particular «-alkanes, 

occurs before biodegradation o f more complex molecular structures. This results in a pronounced 

‘unresolved complex mixture’ (UCM), which is a mixture o f  cyclic and branched hydrocarbons 

that is unresolvable by capillary gas chromatography analysis with conventional detection 

techniques. Should clean up goals be defined in terms o f  concentration or even percentage 

removal o f  contamination, removal o f the UCM may be critical to achieving desired clean-up 

objectives. UCM is especially pronounced in weathered crude oils, some fuel oils and also in 

reservoir biodegraded crude oils. Also, the bulk o f petroleum based lubricating oils consists o f 

aliphatic UCM.

This project aimed to ascertain whether there is a role to play for bioremediation in clean-up o f 

shores contaminated with petroleum containing significant UCM. This was to be achieved by 

investigating a number o f bioremediation strategies and evaluating the effect on UCM removal. 

Broadly, the strategies investigated were biostimulation (supplementation with nutrients) and 

bioaugmentation (addition o f seed microbial cultures). Although there has been little independent 

evidence to support the effectiveness o f bioaugmentation, it was considered useful to investigate 

the role o f  addition o f specialist cultures with a demonstrated capacity for UCM degradation. A 

further objective was to gain an insight into the alteration o f  UCM during bioremediation, 

ascertain what component groups were most resistant to biodegradation, and to study 

biodegradation o f  specific compounds in the UCM. The study was also intended to give an

1



indication o f the limitation o f petroleum hydrocarbon bioremediation as a technology in terms o f 

substrate recalcitrance, i.e. which UCM substrates offer greatest resistance to biodegradation.

There are many analytical methods for determining the biodegradation o f oil and selection o f 

methods for this purpose undoubtedly has a bearing on defining whether a clean-up operation has 

been successful for not. This study enabled a comparison o f  a number o f  analytical approaches 

and an evaluation o f  the suitability o f  methods for measuring oil spill biodegradation.

Furthermore, an insight into how severe biodegradation effects patterns and parameters that are 

commonly used in fingerprinting and source identification o f  oil spills was an indirect benefit 

from the project.

1.2 PETROLEUM HYDROCARBONS AND UNRESOLVED COMPLEX MIXTURES 

(UCM)

Crude oil is a highly complex mixture o f organic compounds, in which hydrocarbons 

predominate. It is formed, subsurface, at elevated temperature and pressure and derives from 

biological material (often marine algae) (Tissot & Welte 1978, Brooks 1983). Transformation into 

crude oil involves a variety o f processes such as maturation (thermal alteration), migration to 

reservoir (chromatographic effect) and changes in the reservoir (primarily biodegradation and 

water washing). Individual crude oils differ from each other depending on the initial material 

from which they are derived and the processes involved in this derivation. Most crude oils are 

similar in their constituents and can be differentiated based on the relative proportions o f certain 

components. Thus, oils formed in the same basin are more likely to exhibit similarities (W rang & 

Adamsen 1990a).

Crude oil composition can broadly be classed into normal paraffins, isoparaffins, naphthenes, 

aromatics, resins, and asphaltenes (Butt 1986). Normal paraffins are «-alkanes; a homologous 

series o f the simplest straight chain hydrocarbons. They usually account for 10-20% o f  the oil but 

can account for up to 60%. Normally the lower «-alkanes predominate. Isoparaffins are branched 

alkanes, and in particular isoprenoids such as pristane and phytane are important constituents. 

«-Alkanes and to some extent isoprenoids are the components responsible for the typical gas 

chromatographic trace o f environmental crude oil samples. Naphthenes are cyclic aliphatic 

compounds based on cyclohexane and cyclopentane units, often existing as condensed structures.

2



The aromatic component of crude oil consists primarily of substituted benzenes or fused ring 

benzenoids (polyaromatic hydrocarbons, PAH) such as naphthalenes and phenanthrenes.

Alkylated PAH compounds are more prevalent than parent PAH structures. High molecular 

weight fractions usually contain significant naphthenoaromatics, and these are particularly 

abundant in immature crude oils. The polar fraction contains compounds that incorporate 

nitrogen, sulfur or oxygen (NSO) in their molecular structure. Sulfur is the most abundant element 

in crude oil after carbon and hydrogen (Tissot & Welte 1978). Thiophene derivatives are 

generally abundant in high sulfur crude oils. Asphaltenes are high molecular weight NSO- 

compounds with complex structures that are defined by their insolubility in «-hexane.

Crude oil is often divided into four classes of compounds i.e. saturates, aromatics, resins and 

asphaltenes, (SARA), {Table LI) .  A  typical North Sea crude oil is light and contains over 60% 

saturates, primarily «-alkanes.

Saturates Aromatics Resins Asphaltenes

Straight chain 

hydrocarbons 

i.e. n-Alkanes

Monocyclic and 

condensed ring 

(polycyclic aromatic

Nitrogen, oxygen and 

sulfur containing 

compounds

High molecular 

weight components 

(insoluble in hexane)

hydrocarbons - PAH) 

Branched alkanes Substituted mono and

e.g. isoprenoids, polycyclic aromatics

‘T-branched ’ alkanes 

Cyclic alkanes 

e.g. steranes, 

hopanes,

alkylcyclohexanes

Table 1.1. The chemistry o f  crude oil

Biodegraded oils analysed by gas chromatography, (GC), often exhibit a pronounced unresolved 

‘hump’, after biodegradation of simple resolved compounds, («-alkanes). This is the unresolved 

complex mixture, (UCM), and consists of myriad compounds at levels that cannot be resolved by

3



conventional GC. There can be many structural isomers for most hydrocarbons of the same 

molecular weight and isomeric possibilities generally increase with increasing molecular weight. 

This results in complex mixtures that cannot be resolved by gas chromatography. Examples of 

molecular structures of petroleum compounds are presented in figure 1.1. The UCM comprises of 

cyclic compounds, and there is also evidence that ‘T-branched’ alkanes to be a significant 

contributor (Gough & Rowland 1990). Cyclic aliphatic compounds in crude oils include alkyl 

substituted cyclohexanes and polycyclic compounds such as steranes and cyclic triterpanes. 

Aromatic compounds in crude oil include alkyl benzenes, which may have linear or branched 

substitutions and are often multi-substituted, and substituted polyaromatic hydrocarbons (PAH). 

Parent PAH are not prevalent in crude oil and their occurrence in the environment is generally 

associated with fossil fuel combustion products or coal tars/ creosotes.

Killops and Al-Juboori (1990) used a number of analytical techniques for characterisation of 

UCM. They concluded that approximately 10% seems to be typical for the amount of aromatic 

carbon. Aromatic components are primarily alkyl benzenoid compounds. Terminal alkyl chains of 

up to ca. C19 are present. Cycloalkanes form an important component and probably comprise of 

more monocycloalkanes than fused polycyclics. ‘T-branched’ alkanes have been suggested as a 

significant component of UCM (Gough & Rowland 1990, 1991). The biodegradability of model 

compounds, proposed as typical of UCM molecular structures, was examined and these 

compounds were found to be recalcitrant (Killops & Al-Juboori 1990, Gough et al. 1992). UCM 

is a significant part of reservoir biodegraded crude oils, petroleum based lubricating oils, some 

fuel oils, and often environmentally weathered oils. As well as aromatic and aliphatic multi-ring 

compounds, some compounds may have both cycloparaffin and aromatic rings. Although multi­

ring aromatics generally exist as condensed rings, more than one ring system may exist for multi­

ring aliphatics (Melpolder et al. 1956, Hood & O’Neal 1959).

4



A

H

Fig. L I: Molecular structures o f example crude oil compounds. A) C25 n-alkane [pentacosane], B) Pristane [2,6,10,14 tetramethyl 

pentadecanej, C) a ‘T-branched’ alkane [9-heptylheptadecane] D) alkylbenzene [n-decy I benzene], E) T-branched alkyl-cyclohexane [8-(2- 

cyclohexylethyl)-hexadecane], F) a C3-phenanthrene [2,3,5-phenanthrene], G) dibenzothiophene H) steranes. I) hopanes, J) triaromatic steranes

5



1.3 PETROLEUM  HYDROCARBONS IN THE M ARINE ENVIRONM ENT

1.3.1 Inputs and sources

There is a range of mechanisms by which petroleum hydrocarbons enter the marine environment. 

The GESAMP report no. 50 (IMG 1993) on the ‘ impact o f oil and related chemicals and wastes 

on the marine environment’, reviewed estimates of global input of petroleum hydrocarbons into 

the oceans. The main source indicated was urban run-off and discharges, followed by operational 

discharges from tankers at sea and then accidents from tankers at sea. Other sources included 

natural seeps, losses from non-tanker shipping, atmospheric deposition, and coastal refinery loss. 

While the figures quoted are probably outdated (most recent data 1981), they suggest a decrease 

in petroleum hydrocarbon pollution of the world’s oceans, due to international measures taken to 

reduce hydrocarbon inputs. Despite improvements in preventative measures however, the 

occurrence o f major oil spills due to tanker accidents will always be a risk while oil is transported 

in bulk by sea. The catastrophic effects of major oil spills on the marine and coastal environment 

is evident and usually results in public outcry for rapid counter measures to be taken (Miller 1990, 

Maki 1991, Lord Donaldson’s Inquiry 1994, SEEC 1998). As well as severe damage to marine 

ecosystems and wildlife, such as seabirds, impacts on fisheries and tourist amenities can have 

direct economic effects. Some of the more infamous marine oil spills are listed below:

Torrey Canyon (Scilly Isles, UK 1967) ~100,000 tonnes

Amoco Cadiz (Brittany, France 1978) -220,000 tonnes

Ixtoc lb \o w  out (Gulf of Mexico 1979) -350,000 tonnes

Exxon Valdez (Alaska, 1989) -  37,000 tonnes

Gulf war spill, (Persian Gulf, 1991) -800,000 -  1,000,000 tonnes

Braer, (Shetland Islands, UK, 1993), ~ 84,700 tonnes crude & 1,600 tonnes bunker fuel

Sea Empress, (Milford Haven, UK, 1996). -  72,000 tonnes crude & 420 tonnes bunker.

1.3.2 Physical and chem ical fate of oil spills

After an oil is spilled at sea its physical and chemical characteristics begin to alter rapidly by 

processes collectively known as weathering (Butt 1986, Wrang & Adamsen 1990b). Lighter 

components are evaporated and within 48 hours components with a boiling point of less than that 

o f tetradecane («C14) can be completely removed. Furthermore, polar compounds can be 

dispersed in the water column, and photo-oxidation can degrade certain compounds (Garret et al.

6



1998). Oil spreads on the water surface forming slicks. Wave action may aid in dispersing the oil, 

or a water-in-oil emulsion known as ‘chocolate m ousse’ may be formed. M ousse is stable and 

persistent in the environment. Biodegradation o f oil is a slower process (weeks to months after a 

spill) and depends on factors such as availability o f nutrients. The «-alkanes are amenable to 

degradation and are removed in the early stages o f biological weathering but the UCM is more 

resistant due to more complex chemical structures o f  its components. There are many factors that 

influence the rate o f  biodegradation and these will be considered later in this chapter. Substantial 

physical weathering and wave action can result in the formation o f tar balls. These present a small 

surface area to volume ratio with more weathered exteriors and are therefore slow to biodegrade.

After the Sea Empress oil spill in Milford Haven 1996, it was estimated that approximately 40% 

of the oil evaporated, 52% dispersed into the water column and 5-7% was stranded on the 

shoreline. A year later less than 1% o f oil remained on the shoreline; the rest having been 

removed, biodegraded or washed back to sea (SEEC 1998).

Ffioto-or.idation

d MousseSpreading Tar baus

Dispersiory
DissoIuHoy}

Biode^adaiia'i

Fig. 1.2 Simple schematic o f  the primary weathering effects on spilt oil

Fusey and Oudot (1984) investigated the relative influence o f physical removal and 

biodegradation in the removal o f  hydrocarbons from shoreline sediments and concluded that the 

stage o f  biodegradation after experimental oilings was directly related to the residual

7



concentration o f  the oil in the sediments. Oil was found to persist at depths o f 25-50cm on 

iniermittently exposed coarse-grained gravel beaches in Prince W illiam Sound 8 years after the 

Exxon Valdez spill, under the protective cover o f cobble/boulder armour (Hayes & Michel 1999, 

M xhel & Hayes, 1999).

1.3.3 Toxicity of petroleum hydrocarbons and impacts o f marine oil spills

The effects o f major oil spills on marine flora and fauna have been well documented and can 

cause immediate death o f  organisms or delayed death by sublethal effects. Massive bird kills 

occurred after the Torrey Canyon and Amoco Cadiz spills (IMO 1993). In 1978 the Amoco Cadiz 

lost 220,000 tonnes o f  crude oil o ff Brittany. Insignificant effects on fin-fish were observed but 

aquaculture was damaged (NRC 1985). By 1985, the concentration o f the weathered Amoco 

Cadiz in sediments was small when compared to other hydrocarbon inputs. This was due to the 

numerous more recent hydrocarbon inputs due to the high level o f  shipping activity (Page et al. 

1988). Oyster growing areas remained polluted with aromatic hydrocarbons seven years after the 

spill and other detrimental effects on marine biota were noted seven to eight years after the spill 

(Berthou et al. 1987, Dauvin 1987). The Exxon Valdez spill caused major mortalities o f  sea otters 

and sea birds, including bald eagles. Clams mussels and finfish were also contaminated (Maki 

1991). The huge crude oil release into the Persian G ulf during the G ulf W ar in 1991, caused 

extensive environmental damage and one year after the spill most severely impacted areas were 

found to be halophyte marsh/algal mat complexes and mudflats at the head o f  sheltered bays. 

Heavy oiling o f burrows was also observed (Hayes et al. 1993). Subtidal sediments showed that 

contamination largely consisted o f UCM by 1992 and that levels o f  contamination had decreased 

by approximately 50%. The reduction in the 1992-1993 period was much less (Readman et al. 

1996 ). Biological effect studies have indicated that the recovery o f  the subtidal region after major 

oil spills is generally rapid (Lee & Page 1997).

The effect o f oiling o f environmentally sensitive ecosystems such as mangrove swamps, coral 

reefs, and mudflats has also been investigated. The severe damage o f  oil contamination in 

mangrove habitat and coral reef has been described. (Bum s & Knap 1989)

The Sea Empress spill resulted in closures o f fisheries, within a designated area, for up to 8 

months for some shellfish species. Although significant contamination o f  fin-fish was not 

detected, shellfish did show contamination. Negative ‘scope for grow th’ and modulations in cell-



mediated immunity were recorded after the spill. The bird species most effected were razorbills, 

guillemots and especially the common scoter. (SEEC 1998)

The initial impact o f oil spills on marine ecosystems is related to the physical effects, such as 

coating and ingestion by marine wildlife and smothering o f  sediments by oil slicks. The toxic 

effects o f  oil have been primarily attributed to the aromatic component. The toxicity o f 

monoaromatics and PAH has been considerably researched and there is growing evidence o f 

mutagenic and carcinogenic effects o f PAH (NRC 1985). W hile higher m olecular weight PAH 

may have a greater inherent toxicity, the greater solubility o f  lower molecular weight components 

may increase their uptake and relative effect. Aliphatic UCM does not appear to have significant 

toxicological effects, UCM oxidation products have been shown to have an increased toxicity on 

M ytilis edulis probably due to increased solubility (Thomas et al. 1995)

1.3.4 C lean-up o f m arine oil spills

In the event o f a major oil spill a wide variety o f technologies and control agents can be employed 

to clean-up oil spills or at least to mitigate the effects (W estermeyer 1991, IMO 1993, Lunel et al. 

1995). Initial efforts usually involve attempts to contain and recover oil or/and to disperse the oil 

while still at sea. Deployment o f booms and skimmers to corral oil at sea is commonly used but 

such methods are o f limited effectiveness, particularly in bad weather. There has been much 

debate about the use o f dispersants, regarding their effectiveness as well as their environmental 

impact. With modem  oil spill dispersants, toxicological impacts o f dispersant application are 

generally due to the effects o f  dispersed oil rather than dispersants themselves (Singer et al. 1993, 

Burridge 1995). Other approaches include burning, sinking, use o f  recovery agents and use o f 

emulsion preventers and breakers.

On the shoreline, approaches include removal and disposal elsewhere o f  contaminated material, 

hot and cold water washing, use o f surfactant or solvent based washing agents, and 

biodegradation enhancement (bioremediation). In 19 9 1 the Exxon Valdez grounded in Prince 

W illiam Sound in Alaska spilling nearly 11 million gallons o f Prudhoe Bay crude oil in an 

ecologically sensitive area. A massive clean-up operation was launched. Many o f the available 

techniques were used as part o f  the Exxon Valdez spill response. This afforded the opportunity to 

carry out a major assessment o f the clean-up operation and the immediate and long-term effects o f 

the oil spill (M iller 1990, Kelso & Kendziorel 1991).

9



Tie appropriate choice o f clean-up technologies, or combinations o f technologies, depends on the 

ci'cumstances, such as type o f oil, environmental factors, type o f location effected (e.g. nature 

reserve, amenity beach, etc.), and necessity to protect/deflect pollution from a sensitive area. On 

slorelines, techniques such as power washing can drive oil into the subsurface where it may be 

sbwer to degrade, or remove oil from the upper-intertidal zones to mid- and sub-intertidal areas, 

wiere it may cause greater damage. (Houghton et al. 1991) In some circumstances, the best 

ajproach may be “the do-nothing option” . The treatment may cause more damage than the oil 

it:elf, and hydrocarbon degrading micro-organisms are ubiquitous and will naturally biodegrade 

oi in time (Atlas 1984).

Boremediation as a clean-up technology involves accelerating biodegradation rates (Atlas &

Alas 1991, Atlas 1996). It has the advantage o f being a technique which is perceived as a natural 

rrechanism for oil removal from the environment, and that it is a relatively ‘soft’, i.e. low 

inrusion approach, with minimal physical manipulation o f the shoreline. Thus, it may be more 

siitable for ecologically sensitive environments than more aggressive approaches. It results in the 

e\entual transformation o f the pollutants to relatively innocuous substances (carbon dioxide and 

witer) as opposed to removal or relocation in the environment. There are certain instances where 

bbremediation may not be suitable; a quick clean-up may be required, (amenity beaches for 

e>ample), or it may be difficult to overcome the rate limiting factor. Furthermore, it can be 

di ficult to predict how effective bioremediation will be for a given spill situation (H off 1993). 

\\h ile the technology may be relatively inexpensive, the cost o f  monitoring over an extended time 

pciod may be prohibitive.

10



1.4 BIOREMEDIATION OF OIL SPILLS

L4.1 Application and case studies

Bioremediation has been proposed as a technology for clean-up o f many organic contaminants in 

a variety o f environments. Land-farming was one o f the earliest forms o f biotreatment and was 

used for disposal o f waste oils and sludges at refineries, and involved nutrient addition and tilling 

o f contaminated soil (Morgan & Watkinson 1989, Atlas & Bartha 1992). More recently soil 

bioremediation has involved using more engineered systems such as soil banking or composting. 

Soil banking utilises a greater degree o f engineering and containment, and along with nutrient 

addition, other strategies to optimise biodegradation are employed such as aeration/oxygenation, 

amendment o f  soil structure, surfactant additions, irrigation and pH control (W ilson & Jones 

1993). As well as treatment o f  hydrocarbons, investigations into the role o f  bioremediation in 

PCS, pesticide, explosive and metal decontamination have been carried out (Kaplan 1992, 

Summers 1992, W ome & Fortune 1993). Bioremedial treatment o f ground water can involve 

pump and bioreactor treatment. Much effort has also been put into developing in situ 

bioremediation techniques for ground water and soil (M organ & Watkinson 1989, Wilson &

Jones 1993, Ellis & Gorder 1997). A wide range o f  commercial products are available that claim 

to assist bioremediation and many are based on the use o f microbial seed cultures.

Bioremediation has also been used to treat marine oil spills (Swannell et al. 1995) and was used 

as one o f the shoreline clean-up techniques following the Exxon Valdez spill. Oleophilic, (Inipol 

EAP22), slow release (Customblen), and soluble inorganic fertilisers were applied on a number of 

beaches (Pritchard & Costa 1991, Prince 1997). Fertiliser levels were adjudged not likely to 

present toxic effects to marine fish or invertebrates. An increase in hexadecane and phenanthrene 

mineralisation activities was demonstrated for fertilised as opposed to non-fertilised sites on three 

beaches (Lindstrom et al. 1991). Ratios o f  total GC detectable hydrocarbons (TGCDHC), total 

resolved hydrocarbons (TRHC: «-alkanes + pristane & phytane) and sum o f  selected PAH 

(TPAH) to C30-hopane was used to assess biodegradation. Results indicated significant 

enhancement o f biodegradation rates correlated to the amount o f  nitrogen delivered per unit o f oil 

(Prince et al. 1994, Bragg 1996). On one beach the rate enhancement was approximately five­

fold.

Venosa et al. reported on the bioremediation o f  an experimental oil spill on the shoreline o f 

Delaware Bay, USA (Venosa et al. 1996). This experiment was designed on a randomised block

I I



basis to provide statistically sound evidence for the efficacy o f  bioremedial treatments.

Treatments included nutrient supplementation and nutrient/ microbial inocula addition, and were 

assessed with reference to untreated control plots. Although substantial biodegradation was 

observed for the untreated plots, a statistically significant increase in biodegradation rates o f total 

target alkanesihopane ratios (Z (hCIO- «C35):hopane) and total aromatic:hopane (IPA H :hopane) 

was observed for the treated plots. No significant difference was detected between the nutrient 

only and nutrient/inocula treatments.

In 1996 the Sea Empress grounded off Milford Haven releasing 72,000 tons o f  Forties blend 

crude oil and approximately 420 tons o f bunker fuel. An investigation o f bioremediation for 

treating a mixture o f  the oils was carried out on a gravel beach at Bullwell bay. Also using a 

randomised block experiment and hydrocarbon:hopane ratios, it was found that nutrient amended 

plots were biodegraded by, on average, 37% more than untreated plots. There was no evidence 

that bioremediation o f the sediments increased oil toxicity (Swannell et al. 1999).

A shoreline study in the Bay o f Brest, France, measured biodegradation rates for total oil, 

aliphatics, cycloalkanes and aromatics in a bioremediation trial using slow release fertiliser. 

Norhopane rather than hopane was used as conserved internal marker. High background 

degradation was observed with no significant improvement for treated plots. This was attributed 

to the high background nutrient levels (Oudot et al. 1998).

The necessity for operational guidelines has been recognised (Swannell et al. 1996, Lee 1998, Lee 

& de M ora 1999), to assist in

• making the decision to bioremediate or not,

• choice o f  bioremediation product,

• designing monitoring strategies,

• consideration o f toxicity and health factors.

1.4.2 M icrobial biodegradation of petroleum hydrocarbons

The ability o f many microorganisms to utilise hydrocarbons as sole carbon sources was identified 

as early as 1946 (Zobell 1946). The micro flora o f the marine environment includes bacteria, 

microalgae, protozoa, actinomycetes, yeasts, fungi and virus’. The ability to metabolise 

hydrocarbon is common in marine bacteria and is spread through a wide variety o f  genera. The

12



genera most frequently described was listed by Floodgate as Pseudomonas, Achromobacter, 

Flavobacterium, Acinetobacter, Vibrio, Bacillus, Arthrobcater, Nocardia, Corynebacterium, and 

Micrococcus (Floodgate 1984). Hydrocarbonoclastic bacterial numbers have been linked to 

previous exposure to hydrocarbons (Zobell 1969, Atlas 1981). This is important in determining 

the rapidity of degradation of subsequent hydrocarbon inputs (Hoff 1993). Bacteria and yeasts 

apparently are the predominant degraders of hydrocarbons in the marine environment and 

filamentous fungi do not appear to play as significant a role as for hydrocarbon degradation in soil 

(Floodgate 1984, Atlas & Bartha 1992). There is some evidence that plasmids may play a role in 

hydrocarbon biodegradation, as there is an increase in proportion of bacterial populations 

containing hydrocarbon utilising plasmids (Hoff 1993, Atlas 1995).

There has been much attention on the possible role of white rot fungi in remediation of soil 

contaminated with persistent organics such as high molecular weight PAH (e.g. gas works sites). 

White rot fungi have received a lot of research interest as they produce non-specific extracellular 

ligninolytic enzymes (Lamar 1992). As well as degradation o f lignin, these fungi have shown the 

ability to degrade persistent organics such as DDT, PAH and PCBs (Field et al. 1992, Colombo et 

al. 1996). As the enzymes are extracellular they may prove to be useful in the bioremediation of 

insoluble compounds, when bioavailability is a limiting factor for other microorganisms.

In general the alkanes degrade more readily than aromatics with n-alkanes being most 

biodegradable. The resistance to aliphatic biodegradation increases with degree of branching and 

cycloalkanes are slow to biodegrade (Atlas 1981). Oudot (1984) used capillary GC and direct MS 

analysis to determine the relative biodegradability of BAL petroleum (Arabian light crude) in 

quasi-continuous culture using an enriched mixed marine microbial inoculum. He concluded that 

the n- and wo-alkanes exhibited most degradation, followed by 6-, 1-, 5- and 2-ring alkanes, 

sulfur aromatics and monoaromatics. 3- and 4- ring alkanes and 2- and 3- ring aromatics were 

moderately susceptible, while 4-ring aromatics and biomarkers were resistant. He concluded the 

most resistant to be 5-ring aromatics, resins and asphaltenes. Gough and Rowland proposed T- 

branched alkanes as significant contributors to the UCM and demonstrated the recalcitrance of 

model branched UCM alkanes (Gough & Rowland 1990, Gough et al. 1992). The refractory 

nature of the larger condensed cycloalkanes has made them useful in fingerprinting severely 

weathered oils and enabled use of C30 hopane as a conserved internal marker. However, 

biodegradation of steranes in laboratory conditions has been shown to occur (Chosson et al.

1991). The route of attack on structures with long alkyl chains on ring structure is o f interest and

13



C2-C7 alkyl benzenes have been shown in one study to be attacked through the ring structure by a 

Pseudom onas sp.(Smith & Ratledge 1989). It is probable that different populations are 

responsible for degrading hydrocarbon classes or that they are at least incapable o f  biodegrading 

structurally different compounds simultaneously (Prince 1993)

Aerobic biodegradation is undoubtedly the primary mechanism for the biodegradation o f  

hydrocarbons. However, hydrocarbons can also be degraded in anaerobic conditions and this has 

received research interest due to the potential application in oxygen depleted environments, such 

as in situ  treatment o f  groundwater. Sulfate reducing bacteria and nitrate reducing bacteria have 

shown an ability to degrade hydrocarbons. The biodegradation o f  up to 4-ring PAH under 

denitrifying conditions has been observed for environmental isolates (M cN ally et al. 1998). Rates 

o f  anaerobic degradation are generally considerably slower than aerobic biodegradation rates.

1.4.3 M etabolic pathways

Much effort has been spent on the elucidation o f  biodegradation pathways. For «-alkanes, the 

primary aerobic pathway involves mono-terminal oxidation to the primary alcohol, aldehyde and 

finally the carboxylic acid. Carboxylic acid biodegradation takes place by P-oxidation with 

subsequent formation o f  a fatty acid, two carbon units shorter, and eventual liberation o f  CO2. 

Diterminal (om ega) and subterminal oxidation have been reported but are less important 

mechanisms. Branched chain m olecules, such as isoprenoids, undergo om ega oxidation and the 

greater the methyl substitution the more resistant they are to biodegradation (Atlas 1981, Singer & 

Finnerty 1984).

Cycloalkanes are a major component o f  crude oil but are quite resistant to biodegradation. Co­

oxidation appears to play an important part in cycloparaffin degradation and alkyl substitution 

often increases the susceptibility to biodegradation. The presence o f  another substrate is 

som etim es necessary to enable biodegradation o f  cyclic compounds by certain micro-organisms 

(Perry 1979). Although som e micro-organisms can not utilise cyclic hydrocarbons as a sole  

source they may be able to oxidise the compound to a ketone or alcohol, which is a substrate for 

other microorganisms (Beam & Perry 1974, Atlas 1981, Perry 1984).

14



Bacterial degradation o f aromatics usually involves hydroxylation to form m -dihydrodiols with 

subsequent formation o f catechols and ring-cleavage. Fungi tend to produce /ra«5-diol 

intermediates (Cem iglia 1984, Rosenberg et al. 1992, Atlas 1995). Examples o f isoprenoid and 

aromatic degradation pathways are given in figures 1.3 and 1.4 respectively.

Pristane (2, 6, 10, 14 tetramethylpentadecane) 

2, 6, 10, 14 tetramethylpentadecanoic acid

(tiJ-oxidation)
i

2, 6, 10, 14 tetramethylpentadecanedioic acid
i

2, 6, 10 trimethyltridecanedioic acid

2, 6, 10 trimethylundecanedioic acid 
>1

2, 6 dimethylnonanedioic acid 

2, 6 dimethylheptanedioic acid
i

2 methylpentanedioic acid

(P 'oxidation)
i

4, 9, 12 trimethyltridecanoic acid 

2, 6, 10 trimethylundecanoic acid

Fig. 1.3: Pathway fo r  pristane oxidation by Brevibacterium erythrogenes 

(adaptedfrom  Singer & Finnerty 1984 in Petroleum M icrobiology ed. Atlas)

15



K, OH
OH

HO

naphthalene

"H (+)-cis-1,2-dihydroxy-1,2-dihydronaphthalene

,0H

1,2-dihydroxynaphthalene

OH
0=C-C00H

m -o-hydroxy benzal pyruvic acid

•OH

^^^C O O H  

cis, c«-m uconic acid

salicylaldehyde

CHO

salicylic acid

COOH

•OH
catechol

• O H

meta fissionortho fission

OOH 2-hydroxymuconic semialdehyde 

OH

Fig. 1.4 ; Pathway fo r  the bacterial oxidation o f  naphthalene 

(adaptedfrom  Cerniglia 1984 in Petroleum M icrobiology ed. Adas)

16



1.4.4 Factors affecting, and optimisation of liydrocarbon biodegradation

There are many factors that affect and control the environmental rates and these have been 

extensively reviewed (M organ & Watkinson 1989, Leahy & Colwell 1990, Atlas & Bartha 1992, 

Prince 1993, Atlas 1995, Head 1998, Swannell et al. 1996, Prince 1997).

1.4.4.1 Nutrients

Metabolism o f hydrocarbons requires the availability o f  certain nutrients, especially nitrogen (N) 

and phosphate (P) for incorporation into biomass. After an oil spill, the organic carbon source is 

abundant and nutrient availability may then become rate limiting (Prince 1993). Most 

bioremediation treatments involve some form o f nutrient addition. Shoreline and microcosm 

nutrient supplementation strategies have included application o f soluble inorganic nutrients 

(Venosa et al. 1996, Swannell et al. 1999), oleophilic fertilisers (Inipol EAP22™ (Swannell et al. 

1995, Bragg e /a /. 1996)) and slow release fertilisers (Swannell e? a/. 1999) (Customblen™ 

(Bragg e /a /. 1996), Max-Bac™( Swannell e /a /. 1995, Oudot e /a /. 1998)). In many instances the 

application o f nutrients did enhance biodegradation rates and slow release, dissolved, and 

oleophilic fertiliser applications have all been shown to be successful. Economic considerations 

are therefore likely to be important when selecting the fertiliser for use. In some studies the 

substantial biodegradation o f the control sites due to background nutrients is commented on 

(Venosa et al. 1996, Oudot et al. 1998). Oudot et al. suggested that if  background level o f  N in 

the interstitial pore water o f  the sediment is >100 ^m oles litre ' then bioremediation through 

fertilisation may be o f limited use. Recently it has been suggested that both the absolute amount 

o f  inorganic nutrient and the ratio o f N:P can effect hydrocarbon degradation rates (Smith et al. 

1998, Head & Swannell 1999).

Lithium nitrate (W renn et al. 1997a, 1997b) has been used as a conserved tracer to assess the 

effect o f  beach hydraulic dynamics in the intertidal zone on dissolved nutrient concentrations and 

distributions. Movement o f the tracer vertically through the sediment subsurface and horizontally 

in a seaward direction from the bioremediation zone was observed. Rapid washout o f tracer from 

the bioremediation zone suggested that, in this case, frequent application o f dissolved nutrients 

(24 hourly) would be required to maintain elevated nutrient concentrations, and thus by 

implication for continued stimulation o f  biodegradation. Use o f tracers such as lithium provides a 

useful technique for evaluating the role o f shoreline physical dynamics on nutrient application 

during bioremediation thereby assisting the design o f  an effective nutrient addition programme.

17



1.4.4.2 Oxygen

As shoreline bioremediation relies on aerobic metabolism, oxygen is critical to the biodegradation 

rates. Dissolved oxygen concentrations in water are very low (~8mg/l) and therefore could rapidly 

become limiting in subsurface sediments. Wave action can cause supersaturation o f water and 

oxygen is less likely to be limiting in high- energy beaches. W aterlogged sediments will be 

oxygen depleted. It should be noted that subsurface water levels will probably not subside with 

receding tide levels but will form a gradient to open water level dependent on sediment 

permeability (Wrenn et al. 1997a). Adequate sediment drainage to enable O 2 diffusion to 

hydrocarbon degrading microbes is therefore essential for effective bioremediation (Atlas & 

Bartha 1992). Tidal cycles facilitate air exchange in intertidal shoreline sediments with good 

drainage. The low sediment permeability o f mudflats means subsurface O 2 is limiting and nutrient 

addition alone is unlikely to be sufficient for effective bioremediation in these environments.

Unsaturated subsurface soil is normally aerobic, but O 2 utilisation may exhaust supplies faster 

than they can be replenished by diffusion, while oxygen is often limiting during in situ 

bioremediation o f groundwater. Physical techniques to improve aeration include manipulation 

(raising and lowering) o f the ground water table, alteration o f soil characteristics using fillers (in 

soil banking), pumping to improve drainage, and bioventing. Chemical oxygenation through use 

o f hydrogen peroxide (H 2O 2) has also been investigated (M organ & W atkinson 1992).

1.4.4.3 Seeding

The introduction o f cultures into the spill environment is referred to as bioaugmentation, or 

‘seeding’, or inoculation (Pritchard 1992). Many o f the products commercially available and 

marketed for bioremediation include microbial formulations. W hile hydrocarbon degrading 

microrganisms are ubiquitous in the environment there is a broad range o f  hydrocarbon structural 

heterogeneity and some hydrocarbons are resistant to biodegradation. Seeding may introduce a 

broader range o f enzymatic capability than present in the indigenous population. Furthermore, an 

adaptive period is necessary for the indigenous microbial population so there may be an 

advantage in adding pre-adapted culture that have demonstrated hydrocarbon degrading ability. 

While it is theoretically promising, ‘seeding’ has not been convincingly demonstrated in the field 

as significantly affecting biodegradation rates. Some commercial products also contain 

dispersants and/or nutrients and it can be difficult to discern whether it is the introduced 

microorganisms, as opposed to other additives, that have enhanced biodegradation. After the 

Exxon Valdez spill, 10 products were screened for efficacy in stimulating oil biodegradation,

18



using a two-tier system. After laboratory tests only two products were selected for field trials 

(Anon. 1991a, Venosa et al. 1991). Neither product demonstrated enhanced biodegradation when 

compared with nutrient application alone in field trials (Venosa et al. 1992). In the Delaware Bay 

study, a mixed consortium isolated from the experimental beach was grown on the same crude oil 

as that used to contaminate the beach as sole carbon source (Venosa et al. 1996). This was used as 

an inoculum but did not enhance biodegradation relative to the nutrient only plots.

There are potentially many factors that may limit the ability o f introduced cultures to survive and 

proliferate in the environment (Pritchard 1992). Cultures that have been substantially laboratory 

cultured or that were initially isolated from a different environment may not be viable; a 

minimum inoculum concentration may be required; competition may occur from indigenous 

micro-organisms and predation. Environmental influences are paramount and it was shown that 

the hydrocarbon mineralisation potential in Prince W illiam Sound after Exxon Valdez was not 

related to substrate concentration but other factors such as intensity o f  physical mixing 

experienced, treatments received and the availability o f alternative carbon sources.

The use o f genetically engineered microorganisms (GEMs) has also been proposed (de Lorenzo 

1992). The first patent for a recombinant micro-organism was issued to Chakrabarty and General 

Electric for a hydrocarbon degrading bacteria (Atlas 1992). However, the potential use o f 

genetically modified micro-organisms is controversial due to considerations such as safety, 

control, and ethical issues. Therefore restrictions remain on environmental release (Atlas &

Bartha 1992). Furthermore, the concept o f  using genetically engineered ‘superbugs’ to 

bioremediate highly complex mixtures, such as petroleum hydrocarbons, is suspect (Lethbridge et 

al. 1994).

Generally, there are few details available with most commercial products, but it is probable that 

many are based on cocktails o f  individual isolates that demonstrate hydrocarbon mineralisation 

potential on single hydrocarbon substrates in the laboratory. Thus, the complexity and synergistic 

effects present in the environment when a range o f  microbes interact to degrade a range o f 

compounds may not be incorporated. This could be particularly important for biodegradation o f 

cyclic compounds where co-oxidation has been demonstrated to be an important factor. Many o f 

the vendors claims for microbial based bioremediation products remain to be independently 

verified.

19



1.4.4.4 Physical state o f  the oil and bioavailability

The availabihty o f hydrocarbon substrate has been recognised as a key element that may limit 

biodegradation in the environment (Head 1998). Biodegradation takes place at oil interfaces or on 

dissolved hydrocarbons. As hydrocarbon solubility is very low, the biodegradation on true 

dissolved compounds is relatively low. Therefore a high surface area to volume ratio o f  oil is 

necessary. M echanisms o f hydrocarbon uptake are via dissolved phase hydrocarbons, direct 

contact with oil surface, or surfactant solubilised hydrocarbon (Floodgate 1984, Singer & Finnerty 

1984). An area that has attracted much attention in recent years is the role o f biosurfactants, 

particularly rhamnolipids, in solubilising and enabling hydrocarbon transport into the cell (Singh 

& Desai 1988). Rhamnolipids have been shown to enhance biodegradation o f  octadecane (Zhang 

& Miller 1992) and to increase solubilisation and removal o f  soil sorbed phenanthrene from soil 

columns (Noordman et al. 1998). A glycolipid was shown to be more effective than Tween-80 for 

solubilisation o f  naphthalene and methyl-substituted naphthalenes from crude oil. It was also 

shown to be less toxic (Kanga et al. 1997). In some cases the use o f synthetic surfactants has been 

shown to enhance biodegradation (Prince 1993, Ducreux et al. 1994). Liposome encapsulation o f 

rtC18 and «C36 was also shown to enhance microbial uptake (M iller & Bartha 1989).

If surfactant addition is to be used, the possibility o f inhibition o f biodegradation due to 

introduction o f competitive substrate should be considered and careful choice o f surfactant is 

required.

1.4.4.5 Other factors -  pH, temperature, competitive substrates

Seawater is naturally well buffered and seawater pH enables high rates o f  oil biodegradation. 

Generally metabolism rates decrease with decreasing temperature and often biodegradation will 

be substantially reduced during winter. Furthermore, physical effects on the oil such as increased 

viscosity may reduce degradation. However, hydrocarbon degrading microorganisms have been 

isolated for a wide temperature range and oil biodegradation has been observed in polar climates 

(Sexstone & Atlas 1977, Kennicutt 1990, AMAP 1998).

Obligate hydrocarbonoclastic bacteria are rare and most hydrocarbon degraders can metabolise 

other substances. If other substrates are available they may be preferentially degraded rather than 

oil, or alternatively, may actually induce hydrocarbon biodegradation, possibly through

20



biosurfactant production (Zhang & M iller 1992, Prince 1993, Al-Hadrami et al. 1996). 

Bioremediation product formulations are being investigate that include nutrients, alternative 

carbon sources, surfactants etc. (Kjeilen 1997).



1.5 ANALYSIS OF PETROLEUM HYDROCARBONS

Methods o f  analysis o f  hydrocarbons, and in particular UCM, were reviewed, with a view to 

selecting and evaluating suitable methods for development and use in monitoring bioremediation 

trials.

1.5.1 Fractionation techniques

There are numerous techniques that can be applied to hydrocarbons and petroleum oils to separate 

into subgroups or classes. These techniques may lead to direct analytical information pertaining to 

the subgroups or be used as preparative fractionation for further analysis (Philp 1994).

1.5.1.1 Column and thin layer chromatography

The fluorescent indicator adsorption (FIA) method {ASTM D1319-98} (ASTM 1998) is the 

petroleum industry standard for hydrocarbon class (group/type) determination. This involves use 

o f silica gel in a specially designed glass column, and separates hydrocarbons into saturates, 

olefms and aromatics. The method is not suitable for dark coloured samples. Alumina and silica 

gel are widely used to achieve separation into saturates, aromatics, resins, and asphaltenes 

(SARA). Separation o f olefms is not required for crude oils as they are present in insignificant 

proportions in crude oils (Tissot & Welte 1978). Thin layer chromatography (TLC) can also be 

used for class separation (Harvey et al. 1984) and can be used with FID detection to directly 

quantify petroleum subgroups (Selucky 1985). Killops and Al-Juboori (1990) used silica gel 

(Merck 60G, 0.4mm, pre-eluted for 24h with dichloromethane prior to re-activation). Bands were 

isolated corresponding to aliphatics (Rf>tridecylbenzene), monoaromatics (tridecylbenzene > R f 

> naphthalene) and polyaromatics (naphthalene > R f > 0) by reference to appropriate standards 

applied to channels at both edges o f the TLC plate. Standards were visualised under 

UV(254nm)after aerosol application o f Rhodamine 6G in methanol. Quantitation was achieved in 

relation to internal standards (squalane, tridecylbenzene, anthracene) added to each fraction, each 

at a level o f 2|ig/m g total hydrocarbons. Gough and Rowland (1990) used argentation TLC in 

conjunction with column chromatography and triple urea adductions to isolate aliphatic UCM.

1.5.1.2 Molecular sieve and urea and thiourea adductions

Urea adduction (Baron 1961) and 5A molecular sieve (O ’Connor et. al 1962, Murphy 1969) are 

both used in separating «-paraffms from crude oils. Both can form channel complexes that can

22



accommodate straight chain alkane molecules but exclude branched and cyclic molecules due to 

the 5A internal diameter crystal lattice. Urea adduction involves a urea clathrate forming around 

the molecule whereas in a molecular sieve the cavity already exists. A comparison o f  5A  

m olecular sieve with urea adduction (Chukwuemeka & Nwobodo 1994) found urea adduction to 

include a broader range o f adductable «-paraffms and also indicated that urea was more efficient. 

Killops and Al-Juboori (1990) used urea adduction in studying UCM composition. Using 

reference compounds (eicosane, anthracene, trideyclbenzene, pyrene and squalane) they found 

that while «-eicosane was completely urea adducted as expected, all other reference compounds 

were also adducted to some extent. They quantified UCM relative to non-adducted squalane 

(assuming that the same quantities o f non-adductable squalane and UCM were retained in the 

adduct fraction). This suggested that 25% o f UCM was adducted suggesting a significant presence 

o f components containing a terminal «-alkyl group attached to another structure. In the same 

paper, thiourea adduction was also employed. This has a crystal lattice o f  ?A  and should retain 

branched acyclic components, cycloalkanes and benzenoid compounds with suitable alkyl chains. 

«-Alkanes and some o f the larger cyclic components are reported not to be retained but the author 

found all reference components were retained to some degree.

Silicalite has also been used to remove branched and cyclic fractions from the «-alkane fraction 

(W est e /a /. 1990).

1.5.2 Spectroscopic techniques 

1.5.2.1 Infrared spectroscopy (IR)

US EPA method 418.1 (US EPA 1983) for m easuring total petroleum hydrocarbons (TPH) in 

water (and soil - EPA 418.1 modified) is based on IR analysis o f the hydrocarbon extract in a 

CFC 113 solution. The aliphatic C-H stretch at ~2930cm ' is quantified against the same peak for 

a reference solution. This may be a reference oil sample or a mixture o f reference compounds in 

solution. However, the unavailability o f CFC as a solvent is likely to make this method redundant. 

A similar ASTM standard test method for oil and grease and petroleum hydrocarbons in water 

also employs CFC 113 as the solvent (ASTM 1996). ASTM method D 3 I4 I-8 0  (ASTM 1981) 

uses IR spectroscopy for comparison o f waterborne petroleum oils and includes guidelines for 

interpreting weathering o f the oil. FT-IR has been used to qualitatively analyse UCM as a film on 

a sodium chloride cell (Killops & Al-Juboori 1990).

23



Comparing %IR7Gravimetric for a number o f oils measured by EPA method 418.1 using a single 

calibration oil illustrates the bias inherent in such a method (Douglas et al. 1992). Oils, such as 

lube oil, with a higher proportion o f saturates to total oil than the calibration oil are positively 

biased, (%IR/Gravimetric > 100%), and oils such as gasoline, with a lower proportion o f saturates 

than the calibration oil are negatively biased, (%IR/Gravimetric < 100%). M easuring the IR 

absorbance o f the oil directly using a fixed path transmission cell or attenuated total reflectance 

(ATR) can eliminate solvent background absorbance.

1.5.2.2 Fluorescence/Luminescence spectroscopy

Fluorescence spectroscopy has been used for measurement o f hydrocarbons in the marine 

environment (Law et al. 1988) using fixed excitation and emission wavelengths and measuring 

the emission intensity against a crude oil, such as Ekofisk, or a poiyaromatic hydrocarbon mixture 

standard curve. Synchronous scanning fluorescence and fixed excitation scanned emission spectra 

enables fingerprinting o f petroleums (Pharr et al. 1992, Shadle et al. 1994). However, total 

fluorescence spectroscopy gives a much more detailed fingerprint, usually displayed as contour 

plots (Butt 1986). Aliphatic components do not fluoresce and this must be taken into account if 

using fluorescence for quantification o f oil samples, as bias may be introduced if a comparable 

standard is not used. Mason (1987) has obtained total fluorescence spectra for oil contamination 

in mussels and also obtained total fluorescence spectra for individual PAH compounds. This 

enabled information to be gained on the aromatic composition o f the oil based on the contour plot, 

and also preparation o f a suitable standard mixture o f PAH that produced a similar contour plot.

1.5.2.3NM R

Killops and Al-Juboori have used FT-'H  NM R and FT-'^C NM R in the analysis o f  UCM. 

Although some useful structural information was obtained for UCM composition, the complexity 

o f  UCM made interpretation difficult (Killops & Al-Juboori 1990).

24



1.5.3 Chromatography

1.5.3.1 Gas chromatography

Gas chromatography with flame ionisation detection is a well-established technique for the 

analysis o f hydrocarbons and oil samples. There are many capillary columns available suitable for 

crude oil analysis resolving n-alkanes, and pristane and phytane, and recently the advent o f  high 

temperature columns has enabled detailed investigations o f  high m olecular weight hydrocarbons 

present in oils, waxes, bitumens and source rocks (Philp 1994). GC is regularly used in estimating 

the boiling range o f both crude petroleums and refined products (ASTM 1984, Yancey et al.

1994). The term ‘unresolved complex m ixture’ refers to the 'hum p’ o f unresolved or poorly 

resolved components in GC-FID traces. The presence o f  UCM in GC analysis o f  environmental 

samples has been considered to provide a better indication o f  petrogenic contamination than 

resolved compounds, as there is a greater chance o f resolved compounds having a recent biogenic 

origin (Volkman et al. 1992). ASTM prescribe a method for comparison o f waterborne petroleum 

oils by gas chromatography (ASTM 1991b).

1.5.3.2 HPLC

Fractionation o f oils by high performance liquid chromatography (HPLC) has proved useful in 

separating petroleum into component classes. Separation o f  petroleum into saturates, aromatics, 

resins and asphaltenes (SARA) has been achieved by initially deasphalting the sample and then 

performing HPLC analysis using silica, alumina, amino, cyano or aminocyano columns, or 

various combinations o f  these columns (Lundanes & Greibrokk 1994). Two columns are 

commonly used, often amino or cyano as the first column and silica, amino, cyano or amino­

cyano as the second. Usually backfiushing o f the resins from the first column is employed, and 

often backfiushing o f the second column is used to determine the aromatics. Grizzle and Sablotny 

(1986) used two amino columns in series, with hexane to elute saturates, and hexane:methylene 

chloride gradient to elute aromatics, with separation according to ring number. Polar compounds 

were eluted by backflushing with 90% methylene chloride in hexane. Detection was achieved 

with a refractive index (RI) detector and ultraviolet (UV) detector. Pearson and Gharfeh (1986), 

used a cyano and an amino-cyano column and two backfiush valves to achieve separation.

Hexane eluted saturates before the aromatics were backflushed from the second column with 

hexane and then polar compounds eluted from the first column with MTBE. A flame ionisation 

detector was used. Shadle et.al. (1994), used a cyano and a silica column in series to analyse shale

25



oils and crude oils using hexaneimethylene chloride gradients and eluting polar compounds in a 

forward direction. Killops and Readman (1985), used two PAC (amino-cyano) columns in series 

to analyse the aromatic fraction o f sediment extracts. Adsorption o f  water and irreversible binding 

o f heteroatoms on silica can cause reproducibility problems.

1.5.4 Mass spectrometry (MS)

1.5.4.1 Gas chromatography-mass spectrometry (GC-MS)

GC-MS in positive electron ionisation mode is widely used for petroleum analysis. GC-MS 

analysis o f  oils and kerogens is particularly useful in geochemical studies for assessing potential 

o f oil bearing source rocks, correlation o f oils and sources, studies o f  depositional environment, 

assessment o f  maturation, extent o f  biodegradation, and other processes (Peters & Moldowan 

1993, Requejo et al. 1996). These techniques have also been adapted as the preferred 

methodology in identifying the source in oil spill incidents (Albaiges & Albrecht 1979, Anon. 

1991b, ASTM 1995). The power o f  GC-MS lies in the use o f mass spectrometry as a selective 

detector to obtain gas chromatographic ‘fingerprints’ for various groups o f  components that 

would be occluded in oil by more concentrated components when analysed by more conventional 

methods. There are two basic component groups that are used for this purpose (W rang &

Adamsen 1990a, 1990b).

(i) Substituted aromatics - Selected ion recording (SIR) o f  a suitable molecular mass 

can allow a chromatogram to be obtained for a group o f structural substituted aromatic 

isomers e.g. methyl phenanthrenes at m/z 192. As the chemical and structural properties o f 

such isomers are very similar, the isomer ‘pattern’ as depicted by a GC-MS trace is relatively 

resistant to weathering.

(ii) Biomarkers - The biomarkers are ‘molecular fossils’ that can be directly related to a 

biological precursor compound from their molecular structure, for example cyclic alkanes 

such as steranes and hopanes. Although biomarker groups tend to have a number o f 

homologous compounds o f differing molecular weights, they often have common structural 

attributes that can be detected as fragments using GC-M S in EI+ mode. Furthermore, 

stereoisomers o f compounds are often present adding to the complexity o f  the patterns. The 

resulting chromatogram is referred to as a ‘mass fragm entogram ’. The term mass

26



chromatogram will be subsequently used in this report to denote a chromatogram for a given 

mass-charge ratio, including mass fragmentograms. The aliphatic biomarker mass 

chromatogram patterns are relatively resistant to weathering, as the compounds are resistant 

to the primary weathering processes (i.e. evaporation, biodegradation and dissolution).

Standard test procedures such as NORDTEST use diagnostic ions based alky! substituted 2,3 and 

4 ring PAH and dibenzothiophene (e.g. m/z 192, 212, 226 for C1-, C2- and C3-phenanthrenes) 

and biomarker fragments (e.g. m/z 217 steranes and m/z 191 pentacyclic triterpanes). Other ions 

that have a diagnostic use include m/z 57, 71, 85 etc for acyclic alkanes, m/z 55 for alkenes and 

cycloalkanes, m/z 83 for alkylcyclohexanes and m/z 95 for naphthenes (Simoneit 1986). 

Multivariate analysis (principal component analysis) was used to compare sterane and triterpane 

data after artificial weathering o f crude oils from the Cook and Brent formations in the same well 

(no. 9) at the Gullfaks field (Brakstad & Grahl-Nielsen 1988). The steranes were unsuitable for 

identification o f weathered oils after four months, but the pentacyclic triterpanes, and in particular 

the demethylated triterpanes, were deemed a good diagnostic tool for oil identification throughout 

the course o f experimental weathering (1 year).

Prince et al. (Prince et al. 1994), used the resistance o f pentacyclic triterpanes to weathering to 

suggest the use o f 17a(H ),2ip(H )-hopane (C30-hopane) as a conserved internal marker for 

estimating the biodegradation o f crude oil. As quantification o f  total crude oil biodegradation in 

the field is extremely difficult, the use o f an internal m arker is an invaluable tool for measuring 

the reduction o f the biodegradable components relative to the conserved marker. From 

biodegradation experiments, Prince et al. concluded that C30-hopane was neither biodegraded or 

generated during the biodegradation o f crude oil fractions on time scales relevant to the estimation 

o f oil spill cleansing. More recently in the absence o f C30-hopane, the use o f norhopane has been 

used as conserved internal standard (Oudot et al. 1998).

The use o f  17a(H),21p(H)-hopane as an conserved internal standard was further developed 

during the EUROCRUDE project, (Anon. 1995). This was a six laboratory collaboration funded 

by the EU LIFE programme to develop an analytical, (GC-M S), database o f  crude oils that are 

produced and transported in European waters, and to evaluate and develop statistical methods and 

computer software for comparing and matching fingerprints with the database. When spilt oils are 

compared with suspect sources by GC-MS, the samples are run on the same machine within the

27



same sample batch. Thus, the primary problem faced was to treat the data in such a way that 

samples analysed on different instruments (or indeed instrument types e.g. quadrupole or 

magnetic sector instruments), or at different times could be compared. This was achieved firstly 

by using a Brent crude oil as a reference, for quality assurance o f data. Analysis o f  spilt crude oil 

samples for comparison with the database involves analysing the reference Brent crude oil by GC- 

MS, followed by the sample twice, and then again the reference oil. Up to 56 peaks can be used 

and the integration data for the sample and reference oils are entered into the model for 

comparison with the database. Four normalisation routines can be used but the most effective for 

reducing interlaboratory and long term variability is triple normalisation. It involves normalisation 

against C30-hopane for both samples and reference oils. The hopane normalised samples are 

averaged and further normalised with the bracketing hopane normalised reference samples.

GC-MS analysis o f biomarkers and aromatics has also proved a useful tool in investigations o f 

chronic hydrocarbon contamination o f the marine environment, and for reconciling the source 

(Volkman et al. 1992). Tricyclic and pentacyclic terpanes have been considered as indicators o f 

the origin o f diffuse lubricating oil contamination in plankton and sediment (Bieger et al. 1996). 

GC-MS analysis o f  biomarkers and aromatics has also been used in tracking oil spillages and in 

assessing the environmental effect o f such spillages in cases such as the Sivand  spill, Humber 

Estuary J983 (Jones et al. 1986), Exxon Valdez, Alaska 1989 (Krahn et al. 1992, Bence et al. 

1996), The G ulf War spill, Kuwait 1991 (Sauer et al. 1993), and the Braer spill, Shetland 1993 

(W olff et al. 1993, Glegg & Rowland 1996).

1.5.4.2 Direct sample introduction mass spectrometry

A current ASTM method is available for estimation for 0-6 ring saturates and monoaromatics in 

non-olefmic saturate petroleum fractions o f average carbon number between 1 6 - 3 2  (ASTM 

1991b). This is based on the fragment peak method devised by Hood and O ’Neal (1959). A 

similar ASTM method enables aromatic type analysis for non-olefmic aromatic fractions 

(Robinson & Cook 1969, ASTM 1991c).

In their study o f UCM Killops and Al-Juboori (1990) used probe MS and used ion abundance of 

m ajor fragments in the UCM spectrum in their interpretation o f  UCM composition.

28



1.5.4.3 Chemical ionisation mass spectrometry (CI-MS)

C I-M S has been less frequently applied to the characterisation o f  crude o ils  and heavy  o ils . C l 

m ass spectra, u sing  N 2O as a chem ical ion isation  reagent gas, for total crude o ils  from a range o f  

fields w ithin  four geographical areas w as tested as a p otentia lly  u sefu l fingerprinting tool for 

d istingu ish in g the geographical origin o f  the oil (Burke e t  al. 1982). Principal com ponent analysis  

and discrim inant analysis dem onstrate the clear d ifferentiation in the o il com p osition  for the 

d iagnostic  ions ch osen . The ions used as d iagnostic ions w ere o f  m /z  > 1 8 0  to exclu de vo latile  

com p onents and m /z < 259. C I-G C -M S has also b een  u sed  for an alysis o f  prefractionated heavy  

o il contam inated so il sam ple extracts (Pollard et al. 1994), although few  details are supplied.

1.5.4.4 Chromic acid oxidation

Interpretation o f  the products o f  UCM  oxidation has proved useful in provid ing an insight into the 

com p osition  o f  the U C M  and indeed surprisingly resolved  chrom atogram s can be obtained. 

U su ally  chrom ium  (V I) oxidation is used, and in particular CrOs, (chrom ium  trioxide). A  general 

review  o f  chrom ium  oxidations in organic chem istry is available (C ainelli &  C ardillo 1984). 

C hrom ium  oxidation  o f  saturated hydrocarbons is not esp ecia lly  se lec tiv e  and considerable  

secon d  stage oxidation  does occur. Thus com p lex  m ixtures o f  reaction products are produced. The 

C -H  attacked in chrom ic oxidation  depends on the attached structures. M ethyl groups are 

relatively  resistant to oxidation  and are eventu ally  ox id ised  to acetic acid  w h ile  tertiary 

hydrocarbons are m uch m ore prone to attack and this sen sitiv ity  is further increased w ith a phenyl 

group attached.

A lthough oxidation  o f  relatively sim ple hydrocarbons can g iv e  rise to com p lex  m ixtures o f  

product, oxidation  o f  the U C M  can produce surprisingly reso lved  chrom atogram s as there is a 

lim ited  class o f  structures that are ox id ised  and sim ilar com pounds tend to be ox id ised  to the sam e 

products. C hrom ic acid  oxidation has been used to study the U C M  com p osition  (G ough  & 

R ow land  1990, 1991) and has been proposed as a potential m echanism  for fingerprinting o ils  

(R e v ille ^ a /. 1992).

1.5.4.5 Stereochemistry and nomenclature o f  petroleum hydrocarbons

S om e understanding o f  the structural configurations o f  petroleum  hydrocarbon m olecu les in 

three-d im ensional space is required w hen interpreting G C -M S an alysis o f  biom arkers

29



(MacKenzie 1984, Peters & Moldowan 1993). Asymmetric or ""chiral" carbons, that is carbons 

with four differing substituents, are common in petroleum hydrocarbons. Two forms of such 

molecules, known as enantiomers, may exist, each being a mirror image of the other. The two 

possible configurations around an asymmetric carbon are called “R” and “S” in conventional 

nomenclature. If more than one asymmetric centre is present, inversion of all the asymmetric 

centres results in enantiomers. However, if less than all the asymmetric centres are inverted the 

different molecular configurations are referred to as diastereomers or epimers. As an example 

pristane contains two asymmetric carbons (fig. 1.5 a). The 6R, lOR and 6S, lOS configuration are 

mirror images and thus are enantiomers. However, note that the 6S, lOR and the 6R, lOS 

configurations can be superimposed on one another and they are therefore identical. These 

configurations are known as meso-pristane and their relationship with 6R, lOR or 6S, lOS pristane 

is diastereomeric.

A property of enantiomers in solution is that they exhibit optical activity, i.e. the extent and 

direction of rotation of plane polarised light is equal and opposite for each enantiomer. The 

physical and chemical properties of enantiomers are very similar, and their chromatographic 

separation requires special chiral stationary phases. However, diastereomers can have different 

chemical and physical properties and can generally be separated on conventional stationary 

phases. Asymmetric configurations in ring systems are distinguished using the nomenclature a  

and p. Subtituents that are above the plane of the molecule are designated p and those below a.

As an example, 5a(H), 14a(H), 17a(H), 20R cholestane, [denoted a a a R ], indicates that 

hydrogens at the 5, 14 and 17 positions of the ring structure face below the plane, and that the 

acyclic asymmetric carbon in the 20 position has an R configuration (fig 1.5 b&c).

Abiotic production of chiral molecules will usually result in a racemic (50:50) mix of 

enantiomers. In biological production of asymmetric molecules, usually only one configuration 

occurs. The fact that most petroleums are optically active is further evidence of the biological 

origin of petroleum. The stereochemical configurations present in biological precursor molecules 

are not necessarily stable during petroleum formation and stereoisomerisation can produce 

different isomeric patterns in petroleum. The a a a R  configuration described above is the natural 

biological configuration of sterane precursors. During petroleum formation stereoisomerisation of 

the biological configurations results in other configurations occurring in petroleum contributing 

to, for example the complex sterane pattern observed in GC-MS analysis.

30



6(R), 10(S) Pristane 6(R), 10(S) Pristane

6(R), 10(R) Pristane
6(S), 10(S) Pristane

a) Stereoisomeric configurations o f  pristane

b) 5a(H ), 14a(H), 17a(H), 20R cholestane 

SoHd circles are used to represent p configurations on asymmetric ring carbons and R 

configurations on acyclic asymmetric configurations. Hollow circles represent the equivalent a

and S configurations.

c) 3-d representation o f  cholestane showing the 5a  (H) and the 3p(H) (adapted from Peters and 

Moldowan 1993).

Fig. 1.5: Stereochemistry and nomenclature o f  petroleum hydrocarbons.

31



1.6 OBJECTIVES OF THE PROJECT

There are many factors involved in successfully bioremediating oils spills on the shoreline. The 

preparation o f bioremediation ‘guidelines’ to aid personnel engaged in an oil spill clean-up has 

been proposed by researchers in this field (Swannell et al. 1996, Lee 1998, Lee & de Mora, 1999). 

This would assist a clean-up team to decide whether bioremediation is an appropriate tool for a 

given set o f  circumstances, and to effectively implement the technology. This project provides 

additional information to the current knowledge base that may prove useful in the preparation o f 

guidelines.

The objectives o f the project were to

• Achieve an understanding into the alteration o f the UCM during bioremediation.

• Establish any order o f biodegradation for certain compound groups characterised by 

m olecular structural attributes.

• Assess whether bioaugmentation o f the microbial population with laboratory enriched 

cultures from a spill site can enable enhanced biodegradation rates o f  UCM.

• Determine whether specifically selected microbial cultures may have a role to play in 

bioremediation o f UCM.

• Determine if  bioremediation is a suitable technique for treating oil spills with high UCM 

content such as lubricating oils and some heavily weathered oils.

• Achieve an insight as to whether recalcitrance o f UCM components can be attributed to 

physical parameters (bioavailability) or biological limitations (metabolism).

• Comparison o f analytical tools for measuring and assessing biodegradation o f  petroleum 

hydrocarbons.

• Insight into the effect o f  severe biodegradation on patterns and parameters used in 

fingerprinting and source identification o f oil spills.

32



1.7 OUTLINE OF THE WORK PROGRAMME

The project was divided into two phases. Phase I included a comprehensive literature survey. This 

was followed by analytical method development to